thesis in artificial intelligence

• Onsite training

3,000,000+ delegates

15,000+ clients

1,000+ locations

• KnowledgePass
• Log a ticket

01344203999 Available 24/7

12 Best Artificial Intelligence Topics for Research in 2024

Explore the "12 Best Artificial Intelligence Topics for Research in 2024." Dive into the top AI research areas, including Natural Language Processing, Computer Vision, Reinforcement Learning, Explainable AI (XAI), AI in Healthcare, Autonomous Vehicles, and AI Ethics and Bias. Stay ahead of the curve and make informed choices for your AI research endeavours.

Exclusive 40% OFF

Training Outcomes Within Your Budget!

We ensure quality, budget-alignment, and timely delivery by our expert instructors.

Share this Resource

• AI Tools in Performance Marketing Training
• Deep Learning Course
• Natural Language Processing (NLP) Fundamentals with Python
• Machine Learning Course
• Duet AI for Workspace Training

Table of Contents  

1) Top Artificial Intelligence Topics for Research 

     a) Natural Language Processing 

     b) Computer vision 

     c) Reinforcement Learning 

     d) Explainable AI (XAI) 

     e) Generative Adversarial Networks (GANs) 

     f) Robotics and AI 

     g) AI in healthcare 

     h) AI for social good 

     i) Autonomous vehicles 

     j) AI ethics and bias 

2) Conclusion 

Top Artificial Intelligence Topics for Research   

This section of the blog will expand on some of the best Artificial Intelligence Topics for research.

Natural Language Processing   

Natural Language Processing (NLP) is centred around empowering machines to comprehend, interpret, and even generate human language. Within this domain, three distinctive research avenues beckon: 

1) Sentiment analysis: This entails the study of methodologies to decipher and discern emotions encapsulated within textual content. Understanding sentiments is pivotal in applications ranging from brand perception analysis to social media insights. 

2) Language generation: Generating coherent and contextually apt text is an ongoing pursuit. Investigating mechanisms that allow machines to produce human-like narratives and responses holds immense potential across sectors. 

3) Question answering systems: Constructing systems that can grasp the nuances of natural language questions and provide accurate, coherent responses is a cornerstone of NLP research. This facet has implications for knowledge dissemination, customer support, and more. 

Computer Vision   

Computer Vision, a discipline that bestows machines with the ability to interpret visual data, is replete with intriguing avenues for research: 

1) Object detection and tracking: The development of algorithms capable of identifying and tracking objects within images and videos finds relevance in surveillance, automotive safety, and beyond. 

2) Image captioning: Bridging the gap between visual and textual comprehension, this research area focuses on generating descriptive captions for images, catering to visually impaired individuals and enhancing multimedia indexing. 

3) Facial recognition: Advancements in facial recognition technology hold implications for security, personalisation, and accessibility, necessitating ongoing research into accuracy and ethical considerations. 

Reinforcement Learning   

Reinforcement Learning revolves around training agents to make sequential decisions in order to maximise rewards. Within this realm, three prominent Artificial Intelligence Topics emerge: 

1) Autonomous agents: Crafting AI agents that exhibit decision-making prowess in dynamic environments paves the way for applications like autonomous robotics and adaptive systems. 

2) Deep Q-Networks (DQN): Deep Q-Networks, a class of reinforcement learning algorithms, remain under active research for refining value-based decision-making in complex scenarios. 

3) Policy gradient methods: These methods, aiming to optimise policies directly, play a crucial role in fine-tuning decision-making processes across domains like gaming, finance, and robotics.  

Explainable AI (XAI)   

The pursuit of Explainable AI seeks to demystify the decision-making processes of AI systems. This area comprises Artificial Intelligence Topics such as: 

1) Model interpretability: Unravelling the inner workings of complex models to elucidate the factors influencing their outputs, thus fostering transparency and accountability. 

2) Visualising neural networks: Transforming abstract neural network structures into visual representations aids in comprehending their functionality and behaviour. 

3) Rule-based systems: Augmenting AI decision-making with interpretable, rule-based systems holds promise in domains requiring logical explanations for actions taken. 

Generative Adversarial Networks (GANs)   

The captivating world of Generative Adversarial Networks (GANs) unfolds through the interplay of generator and discriminator networks, birthing remarkable research avenues: 

1) Image generation: Crafting realistic images from random noise showcases the creative potential of GANs, with applications spanning art, design, and data augmentation. 

2) Style transfer: Enabling the transfer of artistic styles between images, merging creativity and technology to yield visually captivating results. 

3) Anomaly detection: GANs find utility in identifying anomalies within datasets, bolstering fraud detection, quality control, and anomaly-sensitive industries. 

Robotics and AI   

The synergy between Robotics and AI is a fertile ground for exploration, with Artificial Intelligence Topics such as: 

1) Human-robot collaboration: Research in this arena strives to establish harmonious collaboration between humans and robots, augmenting industry productivity and efficiency. 

2) Robot learning: By enabling robots to learn and adapt from their experiences, Researchers foster robots' autonomy and the ability to handle diverse tasks. 

3) Ethical considerations: Delving into the ethical implications surrounding AI-powered robots helps establish responsible guidelines for their deployment. 

AI in healthcare   

AI presents a transformative potential within healthcare, spurring research into: 

1) Medical diagnosis: AI aids in accurately diagnosing medical conditions, revolutionising early detection and patient care. 

2) Drug discovery: Leveraging AI for drug discovery expedites the identification of potential candidates, accelerating the development of new treatments. 

3) Personalised treatment: Tailoring medical interventions to individual patient profiles enhances treatment outcomes and patient well-being. 

AI for social good   

Harnessing the prowess of AI for Social Good entails addressing pressing global challenges: 

1) Environmental monitoring: AI-powered solutions facilitate real-time monitoring of ecological changes, supporting conservation and sustainable practices. 

2) Disaster response: Research in this area bolsters disaster response efforts by employing AI to analyse data and optimise resource allocation. 

3) Poverty alleviation: Researchers contribute to humanitarian efforts and socioeconomic equality by devising AI solutions to tackle poverty. 

Unlock the potential of Artificial Intelligence for effective Project Management with our Artificial Intelligence (AI) for Project Managers Course . Sign up now!  

Autonomous vehicles   

Autonomous Vehicles represent a realm brimming with potential and complexities, necessitating research in Artificial Intelligence Topics such as: 

1) Sensor fusion: Integrating data from diverse sensors enhances perception accuracy, which is essential for safe autonomous navigation. 

2) Path planning: Developing advanced algorithms for path planning ensures optimal routes while adhering to safety protocols. 

3) Safety and ethics: Ethical considerations, such as programming vehicles to make difficult decisions in potential accident scenarios, require meticulous research and deliberation. 

AI ethics and bias   

Ethical underpinnings in AI drive research efforts in these directions: 

1) Fairness in AI: Ensuring AI systems remain impartial and unbiased across diverse demographic groups. 

2) Bias detection and mitigation: Identifying and rectifying biases present within AI models guarantees equitable outcomes. 

3) Ethical decision-making: Developing frameworks that imbue AI with ethical decision-making capabilities aligns technology with societal values. 

Future of AI  

The vanguard of AI beckons Researchers to explore these horizons: 

1) Artificial General Intelligence (AGI): Speculating on the potential emergence of AI systems capable of emulating human-like intelligence opens dialogues on the implications and challenges. 

2) AI and creativity: Probing the interface between AI and creative domains, such as art and music, unveils the coalescence of human ingenuity and technological prowess. 

3) Ethical and regulatory challenges: Researching the ethical dilemmas and regulatory frameworks underpinning AI's evolution fortifies responsible innovation. 

AI and education   

The intersection of AI and Education opens doors to innovative learning paradigms: 

1) Personalised learning: Developing AI systems that adapt educational content to individual learning styles and paces. 

2) Intelligent tutoring systems: Creating AI-driven tutoring systems that provide targeted support to students. 

3) Educational data mining: Applying AI to analyse educational data for insights into learning patterns and trends. 

Unleash the full potential of AI with our comprehensive Introduction to Artificial Intelligence Training . Join now!  

Conclusion  

The domain of AI is ever-expanding, rich with intriguing topics about Artificial Intelligence that beckon Researchers to explore, question, and innovate. Through the pursuit of these twelve diverse Artificial Intelligence Topics, we pave the way for not only technological advancement but also a deeper understanding of the societal impact of AI. By delving into these realms, Researchers stand poised to shape the trajectory of AI, ensuring it remains a force for progress, empowerment, and positive transformation in our world. 

Unlock your full potential with our extensive Personal Development Training Courses. Join today!  

Frequently Asked Questions

Upcoming data, analytics & ai resources batches & dates.

Fri 26th Apr 2024

Fri 2nd Aug 2024

Fri 15th Nov 2024

Get A Quote

WHO WILL BE FUNDING THE COURSE?

My employer

By submitting your details you agree to be contacted in order to respond to your enquiry

• Business Analysis
• Lean Six Sigma Certification

Share this course

Our biggest spring sale.

We cannot process your enquiry without contacting you, please tick to confirm your consent to us for contacting you about your enquiry.

By submitting your details you agree to be contacted in order to respond to your enquiry.

We may not have the course you’re looking for. If you enquire or give us a call on 01344203999 and speak to our training experts, we may still be able to help with your training requirements.

Or select from our popular topics

• ITIL® Certification
• Scrum Certification
• Change Management Certification
• Business Analysis Courses
• Microsoft Azure Certification
• Microsoft Excel & Certification Course
• Microsoft Project
• Explore more courses

Press esc to close

Fill out your  contact details  below and our training experts will be in touch.

Fill out your   contact details   below

Thank you for your enquiry!

One of our training experts will be in touch shortly to go over your training requirements.

Back to Course Information

Fill out your contact details below so we can get in touch with you regarding your training requirements.

* WHO WILL BE FUNDING THE COURSE?

Preferred Contact Method

No preference

Back to course information

Fill out your  training details  below

Fill out your training details below so we have a better idea of what your training requirements are.

HOW MANY DELEGATES NEED TRAINING?

HOW DO YOU WANT THE COURSE DELIVERED?

Online Instructor-led

Online Self-paced

WHEN WOULD YOU LIKE TO TAKE THIS COURSE?

Next 2 - 4 months

WHAT IS YOUR REASON FOR ENQUIRING?

Looking for some information

Looking for a discount

I want to book but have questions

One of our training experts will be in touch shortly to go overy your training requirements.

Your privacy & cookies!

Like many websites we use cookies. We care about your data and experience, so to give you the best possible experience using our site, we store a very limited amount of your data. Continuing to use this site or clicking “Accept & close” means that you agree to our use of cookies. Learn more about our privacy policy and cookie policy cookie policy .

We use cookies that are essential for our site to work. Please visit our cookie policy for more information. To accept all cookies click 'Accept & close'.

• Free Python 3 Tutorial
• Control Flow
• Exception Handling
• Python Programs
• Python Projects
• Python Interview Questions
• Python Database
• Data Science With Python
• Machine Learning with Python
• How Amazon Uses Machine Learning?
• Artificial Intelligence Could be a Better Doctor
• 10 Most Interesting Chatbots in the World
• What Are DeepFakes And How Dangerous Are They?
• Impact of AI and ML On Warfare Techniques
• What is the Role of Artificial Intelligence in Fighting Coronavirus?
• Top 7 Artificial Intelligence and Machine Learning Trends For 2022
• 5 Algorithms that Demonstrate Artificial Intelligence Bias
• What is Artificial Intelligence as a Service (AIaaS) in the Tech Industry?
• Top 10 Business Intelligence Platforms in 2020
• What is IBM Watson and Its Services?
• 8 Best Artificial Intelligence Books For Beginners in 2024
• Can Artificial Intelligence Help in Curing Cancer?
• Is AI Really a Threat to Cybersecurity?
• 10 Best Artificial Intelligence Project Ideas To Kick-Start Your Career
• 5 Best Humanoid Robots in The World
• Propositional Logic based Agent
• Resolution Algorithm in Artificial Intelligence
• Face recognition using Artificial Intelligence

8 Best Topics for Research and Thesis in Artificial Intelligence

Imagine a future in which intelligence is not restricted to humans!!! A future where machines can think as well as humans and work with them to create an even more exciting universe. While this future is still far away, Artificial Intelligence has still made a lot of advancement in these times. There is a lot of research being conducted in almost all fields of AI like Quantum Computing, Healthcare, Autonomous Vehicles, Internet of Things , Robotics , etc. So much so that there is an increase of 90% in the number of annually published research papers on Artificial Intelligence since 1996. Keeping this in mind, if you want to research and write a thesis based on Artificial Intelligence, there are many sub-topics that you can focus on. Some of these topics along with a brief introduction are provided in this article. We have also mentioned some published research papers related to each of these topics so that you can better understand the research process.

So without further ado, let’s see the different Topics for Research and Thesis in Artificial Intelligence!

1. Machine Learning

Machine Learning involves the use of Artificial Intelligence to enable machines to learn a task from experience without programming them specifically about that task. (In short, Machines learn automatically without human hand holding!!!) This process starts with feeding them good quality data and then training the machines by building various machine learning models using the data and different algorithms. The choice of algorithms depends on what type of data do we have and what kind of task we are trying to automate. However, generally speaking, Machine Learning Algorithms are divided into 3 types i.e. Supervised Machine Learning Algorithms, Unsupervised Machine Learning Algorithms , and Reinforcement Machine Learning Algorithms.

2. Deep Learning

Deep Learning is a subset of Machine Learning that learns by imitating the inner working of the human brain in order to process data and implement decisions based on that data. Basically, Deep Learning uses artificial neural networks to implement machine learning. These neural networks are connected in a web-like structure like the networks in the human brain (Basically a simplified version of our brain!). This web-like structure of artificial neural networks means that they are able to process data in a nonlinear approach which is a significant advantage over traditional algorithms that can only process data in a linear approach. An example of a deep neural network is RankBrain which is one of the factors in the Google Search algorithm.

3. Reinforcement Learning

Reinforcement Learning is a part of Artificial Intelligence in which the machine learns something in a way that is similar to how humans learn. As an example, assume that the machine is a student. Here the hypothetical student learns from its own mistakes over time (like we had to!!). So the Reinforcement Machine Learning Algorithms learn optimal actions through trial and error. This means that the algorithm decides the next action by learning behaviors that are based on its current state and that will maximize the reward in the future. And like humans, this works for machines as well! For example, Google’s AlphaGo computer program was able to beat the world champion in the game of Go (that’s a human!) in 2017 using Reinforcement Learning.

4. Robotics

Robotics is a field that deals with creating humanoid machines that can behave like humans and perform some actions like human beings. Now, robots can act like humans in certain situations but can they think like humans as well? This is where artificial intelligence comes in! AI allows robots to act intelligently in certain situations. These robots may be able to solve problems in a limited sphere or even learn in controlled environments. An example of this is Kismet , which is a social interaction robot developed at M.I.T’s Artificial Intelligence Lab. It recognizes the human body language and also our voice and interacts with humans accordingly. Another example is Robonaut , which was developed by NASA to work alongside the astronauts in space.

5. Natural Language Processing

It’s obvious that humans can converse with each other using speech but now machines can too! This is known as Natural Language Processing where machines analyze and understand language and speech as it is spoken (Now if you talk to a machine it may just talk back!). There are many subparts of NLP that deal with language such as speech recognition, natural language generation, natural language translation , etc. NLP is currently extremely popular for customer support applications, particularly the chatbot . These chatbots use ML and NLP to interact with the users in textual form and solve their queries. So you get the human touch in your customer support interactions without ever directly interacting with a human.

Some Research Papers published in the field of Natural Language Processing are provided here. You can study them to get more ideas about research and thesis on this topic.

6. Computer Vision

The internet is full of images! This is the selfie age, where taking an image and sharing it has never been easier. In fact, millions of images are uploaded and viewed every day on the internet. To make the most use of this huge amount of images online, it’s important that computers can see and understand images. And while humans can do this easily without a thought, it’s not so easy for computers! This is where Computer Vision comes in. Computer Vision uses Artificial Intelligence to extract information from images. This information can be object detection in the image, identification of image content to group various images together, etc. An application of computer vision is navigation for autonomous vehicles by analyzing images of surroundings such as AutoNav used in the Spirit and Opportunity rovers which landed on Mars.

7. Recommender Systems

When you are using Netflix, do you get a recommendation of movies and series based on your past choices or genres you like? This is done by Recommender Systems that provide you some guidance on what to choose next among the vast choices available online. A Recommender System can be based on Content-based Recommendation or even Collaborative Filtering. Content-Based Recommendation is done by analyzing the content of all the items. For example, you can be recommended books you might like based on Natural Language Processing done on the books. On the other hand, Collaborative Filtering is done by analyzing your past reading behavior and then recommending books based on that.

8. Internet of Things

Artificial Intelligence deals with the creation of systems that can learn to emulate human tasks using their prior experience and without any manual intervention. Internet of Things , on the other hand, is a network of various devices that are connected over the internet and they can collect and exchange data with each other. Now, all these IoT devices generate a lot of data that needs to be collected and mined for actionable results. This is where Artificial Intelligence comes into the picture. Internet of Things is used to collect and handle the huge amount of data that is required by the Artificial Intelligence algorithms. In turn, these algorithms convert the data into useful actionable results that can be implemented by the IoT devices.

Please Login to comment...

Similar reads.

• AI-ML-DS Blogs
• 10 Best Slack Integrations to Enhance Your Team's Productivity
• 10 Best Zendesk Alternatives and Competitors
• 10 Best Trello Power-Ups for Maximizing Project Management
• Google Rolls Out Gemini In Android Studio For Coding Assistance
• 30 OOPs Interview Questions and Answers (2024)

Improve your Coding Skills with Practice

What kind of Experience do you want to share?

•   LibGuides
•   A-Z List
•   Help

Artificial Intelligence

• Background Information
• Getting started
• Browse Journals
• Dissertations & Theses
• Datasets and Repositories
• Research Data Management 101
• Scientific Writing
• Find Videos
• Related Topics
• Quick Links
• Ask Us/Contact Us

FIU dissertations

Non-FIU dissertations

Many   universities   provide full-text access to their dissertations via a digital repository.  If you know the title of a particular dissertation or thesis, try doing a Google search.  

Aims to be the best possible resource for finding open access graduate theses and dissertations published around the world with metadata from over 800 colleges, universities, and research institutions. Currently, indexes over 1 million theses and dissertations.

This is a discovery service for open access research theses awarded by European universities.

A union catalog of Canadian theses and dissertations, in both electronic and analog formats, is available through the search interface on this portal.

There are currently more than 90 countries and over 1200 institutions represented. CRL has catalog records for over 800,000 foreign doctoral dissertations.

An international collaborative resource, the NDLTD Union Catalog contains more than one million records of electronic theses and dissertations. Use BASE, the VTLS Visualizer or any of the geographically specific search engines noted lower on their webpage.

Indexes doctoral dissertations and masters' theses in all areas of academic research includes international coverage.

ProQuest Dissertations & Theses global

Related Sites

• << Previous: Browse Journals
• Next: Datasets and Repositories >>
• Last Updated: Apr 4, 2024 8:33 AM
• URL: https://library.fiu.edu/artificial-intelligence

Information

Fiu libraries floorplans, green library, modesto a. maidique campus, hubert library, biscayne bay campus.

Directions: Green Library, MMC

Directions: Hubert Library, BBC

Completed Theses

State space search solves navigation tasks and many other real world problems. Heuristic search, especially greedy best-first search, is one of the most successful algorithms for state space search. We improve the state of the art in heuristic search in three directions.

In Part I, we present methods to train neural networks as powerful heuristics for a given state space. We present a universal approach to generate training data using random walks from a (partial) state. We demonstrate that our heuristics trained for a specific task are often better than heuristics trained for a whole domain. We show that the performance of all trained heuristics is highly complementary. There is no clear pattern, which trained heuristic to prefer for a specific task. In general, model-based planners still outperform planners with trained heuristics. But our approaches exceed the model-based algorithms in the Storage domain. To our knowledge, only once before in the Spanner domain, a learning-based planner exceeded the state-of-the-art model-based planners.

A priori, it is unknown whether a heuristic, or in the more general case a planner, performs well on a task. Hence, we trained online portfolios to select the best planner for a task. Today, all online portfolios are based on handcrafted features. In Part II, we present new online portfolios based on neural networks, which receive the complete task as input, and not just a few handcrafted features. Additionally, our portfolios can reconsider their choices. Both extensions greatly improve the state-of-the-art of online portfolios. Finally, we show that explainable machine learning techniques, as the alternative to neural networks, are also good online portfolios. Additionally, we present methods to improve our trust in their predictions.

Even if we select the best search algorithm, we cannot solve some tasks in reasonable time. We can speed up the search if we know how it behaves in the future. In Part III, we inspect the behavior of greedy best-first search with a fixed heuristic on simple tasks of a domain to learn its behavior for any task of the same domain. Once greedy best-first search expanded a progress state, it expands only states with lower heuristic values. We learn to identify progress states and present two methods to exploit this knowledge. Building upon this, we extract the bench transition system of a task and generalize it in such a way that we can apply it to any task of the same domain. We can use this generalized bench transition system to split a task into a sequence of simpler searches.

In all three research directions, we contribute new approaches and insights to the state of the art, and we indicate interesting topics for future work.

Greedy best-first search (GBFS) is a sibling of A* in the family of best-first state-space search algorithms. While A* is guaranteed to find optimal solutions of search problems, GBFS does not provide any guarantees but typically finds satisficing solutions more quickly than A*. A classical result of optimal best-first search shows that A* with admissible and consistent heuristic expands every state whose f-value is below the optimal solution cost and no state whose f-value is above the optimal solution cost. Theoretical results of this kind are useful for the analysis of heuristics in different search domains and for the improvement of algorithms. For satisficing algorithms a similarly clear understanding is currently lacking. We examine the search behavior of GBFS in order to make progress towards such an understanding.

We introduce the concept of high-water mark benches, which separate the search space into areas that are searched by GBFS in sequence. High-water mark benches allow us to exactly determine the set of states that GBFS expands under at least one tie-breaking strategy. We show that benches contain craters. Once GBFS enters a crater, it has to expand every state in the crater before being able to escape.

Benches and craters allow us to characterize the best-case and worst-case behavior of GBFS in given search instances. We show that computing the best-case or worst-case behavior of GBFS is NP-complete in general but can be computed in polynomial time for undirected state spaces.

We present algorithms for extracting the set of states that GBFS potentially expands and for computing the best-case and worst-case behavior. We use the algorithms to analyze GBFS on benchmark tasks from planning competitions under a state-of-the-art heuristic. Experimental results reveal interesting characteristics of the heuristic on the given tasks and demonstrate the importance of tie-breaking in GBFS.

Classical planning tackles the problem of finding a sequence of actions that leads from an initial state to a goal. Over the last decades, planning systems have become significantly better at answering the question whether such a sequence exists by applying a variety of techniques which have become more and more complex. As a result, it has become nearly impossible to formally analyze whether a planning system is actually correct in its answers, and we need to rely on experimental evidence.

One way to increase trust is the concept of certifying algorithms, which provide a witness which justifies their answer and can be verified independently. When a planning system finds a solution to a problem, the solution itself is a witness, and we can verify it by simply applying it. But what if the planning system claims the task is unsolvable? So far there was no principled way of verifying this claim.

This thesis contributes two approaches to create witnesses for unsolvable planning tasks. Inductive certificates are based on the idea of invariants. They argue that the initial state is part of a set of states that we cannot leave and that contains no goal state. In our second approach, we define a proof system that proves in an incremental fashion that certain states cannot be part of a solution until it has proven that either the initial state or all goal states are such states.

Both approaches are complete in the sense that a witness exists for every unsolvable planning task, and can be verified efficiently (in respect to the size of the witness) by an independent verifier if certain criteria are met. To show their applicability to state-of-the-art planning techniques, we provide an extensive overview how these approaches can cover several search algorithms, heuristics and other techniques. Finally, we show with an experimental study that generating and verifying these explanations is not only theoretically possible but also practically feasible, thus making a first step towards fully certifying planning systems.

Heuristic search with an admissible heuristic is one of the most prominent approaches to solving classical planning tasks optimally. In the first part of this thesis, we introduce a new family of admissible heuristics for classical planning, based on Cartesian abstractions, which we derive by counterexample-guided abstraction refinement. Since one abstraction usually is not informative enough for challenging planning tasks, we present several ways of creating diverse abstractions. To combine them admissibly, we introduce a new cost partitioning algorithm, which we call saturated cost partitioning. It considers the heuristics sequentially and uses the minimum amount of costs that preserves all heuristic estimates for the current heuristic before passing the remaining costs to subsequent heuristics until all heuristics have been served this way.

In the second part, we show that saturated cost partitioning is strongly influenced by the order in which it considers the heuristics. To find good orders, we present a greedy algorithm for creating an initial order and a hill-climbing search for optimizing a given order. Both algorithms make the resulting heuristics significantly more accurate. However, we obtain the strongest heuristics by maximizing over saturated cost partitioning heuristics computed for multiple orders, especially if we actively search for diverse orders.

The third part provides a theoretical and experimental comparison of saturated cost partitioning and other cost partitioning algorithms. Theoretically, we show that saturated cost partitioning dominates greedy zero-one cost partitioning. The difference between the two algorithms is that saturated cost partitioning opportunistically reuses unconsumed costs for subsequent heuristics. By applying this idea to uniform cost partitioning we obtain an opportunistic variant that dominates the original. We also prove that the maximum over suitable greedy zero-one cost partitioning heuristics dominates the canonical heuristic and show several non-dominance results for cost partitioning algorithms. The experimental analysis shows that saturated cost partitioning is the cost partitioning algorithm of choice in all evaluated settings and it even outperforms the previous state of the art in optimal classical planning.

Classical planning is the problem of finding a sequence of deterministic actions in a state space that lead from an initial state to a state satisfying some goal condition. The dominant approach to optimally solve planning tasks is heuristic search, in particular A* search combined with an admissible heuristic. While there exist many different admissible heuristics, we focus on abstraction heuristics in this thesis, and in particular, on the well-established merge-and-shrink heuristics.

Our main theoretical contribution is to provide a comprehensive description of the merge-and-shrink framework in terms of transformations of transition systems. Unlike previous accounts, our description is fully compositional, i.e. can be understood by understanding each transformation in isolation. In particular, in addition to the name-giving merge and shrink transformations, we also describe pruning and label reduction as such transformations. The latter is based on generalized label reduction, a new theory that removes all of the restrictions of the previous definition of label reduction. We study the four types of transformations in terms of desirable formal properties and explain how these properties transfer to heuristics being admissible and consistent or even perfect. We also describe an optimized implementation of the merge-and-shrink framework that substantially improves the efficiency compared to previous implementations.

Furthermore, we investigate the expressive power of merge-and-shrink abstractions by analyzing factored mappings, the data structure they use for representing functions. In particular, we show that there exist certain families of functions that can be compactly represented by so-called non-linear factored mappings but not by linear ones.

On the practical side, we contribute several non-linear merge strategies to the merge-and-shrink toolbox. In particular, we adapt a merge strategy from model checking to planning, provide a framework to enhance existing merge strategies based on symmetries, devise a simple score-based merge strategy that minimizes the maximum size of transition systems of the merge-and-shrink computation, and describe another framework to enhance merge strategies based on an analysis of causal dependencies of the planning task.

In a large experimental study, we show the evolution of the performance of merge-and-shrink heuristics on planning benchmarks. Starting with the state of the art before the contributions of this thesis, we subsequently evaluate all of our techniques and show that state-of-the-art non-linear merge-and-shrink heuristics improve significantly over the previous state of the art.

Admissible heuristics are the main ingredient when solving classical planning tasks optimally with heuristic search. Higher admissible heuristic values are more accurate, so combining them in a way that dominates their maximum and remains admissible is an important problem.

The thesis makes three contributions in this area. Extensions to cost partitioning (a well-known heuristic combination framework) allow to produce higher estimates from the same set of heuristics. The new heuristic family called operator-counting heuristics unifies many existing heuristics and offers a new way to combine them. Another new family of heuristics called potential heuristics allows to cast the problem of finding a good heuristic as an optimization problem.

Both operator-counting and potential heuristics are closely related to cost partitioning. They offer a new look on cost partitioned heuristics and already sparked research beyond their use as classical planning heuristics.

Master's theses

Classical planning tasks are typically formulated in PDDL. Some of them can be described more concisely using derived variables. Contrary to basic variables, their values cannot be changed by operators and are instead determined by axioms which specify conditions under which they take a certain value. Planning systems often support axioms in their search component, but their heuristics’ support is limited or nonexistent. This leads to decreased search performance with tasks that use axioms. We compile axioms away using our implementation of a known algorithm in the Fast Downward planner. Our results show that the compilation has a negative impact on search performance with its only benefit being the ability to use heuristics that have no axiom support. As a compromise between performance and expressivity, we identify axioms of a simple form and devise a compilation for them. We compile away all axioms in several of the tested domains without a decline in search performance.

The International Planning Competitions (IPCs) serve as a testing suite for planning sys- tems. These domains are well-motivated as they are derived from, or possess characteristics analogous to real-life applications. In this thesis, we study the computational complexity of the plan existence and bounded plan existence decision problems of the following grid- based IPC domains: VisitAll, TERMES, Tidybot, Floortile, and Nurikabe. In all of these domains, there are one or more agents moving through a rectangular grid (potentially with obstacles) performing actions along the way. In many cases, we engineer instances that can be solved only if the movement of the agent or agents follows a Hamiltonian path or cycle in a grid graph. This gives rise to many NP-hardness reductions from Hamiltonian path/cycle problems on grid graphs. In the case of VisitAll and Floortile, we give necessary and suffi- cient conditions for deciding the plan existence problem in polynomial time. We also show that Tidybot has the game Push -1F as a special case, and its plan existence problem is thus PSPACE-complete. The hardness proofs in this thesis highlight hard instances of these domains. Moreover, by assigning a complexity class to each domain, researchers and practitioners can better assess the strengths and limitations of new and existing algorithms in these domains.

Planning tasks can be used to describe many real world problems of interest. Solving those tasks optimally is thus an avenue of great interest. One established and successful approach for optimal planning is the merge-and-shrink framework, which decomposes the task into a factored transition system. The factors initially represent the behaviour of one state variable and are repeatedly combined and abstracted. The solutions of these abstract states is then used as a heuristic to guide search in the original planning task. Existing merge-and-shrink transformations keep the factored transition system orthogonal, meaning that the variables of the planning task are represented in no more than one factor at any point. In this thesis we introduce the clone transformation, which duplicates a factor of the factored transition system, making it non-orthogonal. We test two classes of clone strategies, which we introduce and implement in the Fast Downward planning system and conclude that, while theoretically promising, our clone strategies are practically inefficient as their performance was worse than state-of-the-art methods for merge-and-shrink.

This thesis aims to present a novel approach for improving the performance of classical planning algorithms by integrating cost partitioning with merge-and-shrink techniques. Cost partitioning is a well-known technique for admissibly adding multiple heuristic values. Merge-and-shrink, on the other hand, is a technique to generate well-informed abstractions. The "merge” part of the technique is based on creating an abstract representation of the original problem by replacing two transition systems with their synchronised product. In contrast, the ”shrink” part refers to reducing the size of the factor. By combining these two approaches, we aim to leverage the strengths of both methods to achieve better scalability and efficiency in solving classical planning problems. Considering a range of benchmark domains and the Fast Downward planning system, the experimental results show that the proposed method achieves the goal of fusing merge and shrink with cost partitioning towards better outcomes in classical planning.

Planning is the process of finding a path in a planning task from the initial state to a goal state. Multiple algorithms have been implemented to solve such planning tasks, one of them being the Property-Directed Reachability algorithm. Property-Directed Reachability utilizes a series of propositional formulas called layers to represent a super-set of states with a goal distance of at most the layer index. The algorithm iteratively improves the layers such that they represent a minimum number of states. This happens by strengthening the layer formulas and therefore excluding states with a goal distance higher than the layer index. The goal of this thesis is to implement a pre-processing step to seed the layers with a formula that already excludes as many states as possible, to potentially improve the run-time performance. We use the pattern database heuristic and its associated pattern generators to make use of the planning task structure for the seeding algorithm. We found that seeding does not consistently improve the performance of the Property-Directed Reachability algorithm. Although we observed a significant reduction in planning time for some tasks, it significantly increased for others.

Certifying algorithms is a concept developed to increase trust by demanding affirmation of the computed result in form of a certificate. By inspecting the certificate, it is possible to determine correctness of the produced output. Modern planning systems have been certifying for long time in the case of solvable instances, where a generated plan acts as a certificate.

Only recently there have been the first steps towards certifying unsolvability judgments in the form of inductive certificates which represent certain sets of states. Inductive certificates are expressed with the help of propositional formulas in a specific formalism.

In this thesis, we investigate the use of propositional formulas in conjunctive normal form (CNF) as a formalism for inductive certificates. At first, we look into an approach that allows us to construct formulas representing inductive certificates in CNF. To show general applicability of this approach, we extend this to the family of delete relaxation heuristics. Furthermore, we present how a planning system is able to generate an inductive validation formula, a single formula that can be used to validate if the set found by the planner is indeed an inductive certificate. At last, we show with an experimental evaluation that the CNF formalism can be feasible in practice for the generation and validation of inductive validation formulas.

In generalized planning the aim is to solve whole classes of planning tasks instead of single tasks one at a time. Generalized representations provide information or knowledge about such classes to help solving them. This work compares the expressiveness of three generalized representations, generalized potential heuristics, policy sketches and action schema networks, in terms of compilability. We use a notion of equivalence that requires two generalized representations to decompose the tasks of a class into the same subtasks. We present compilations between pairs of equivalent generalized representations and proofs where a compilation is impossible.

A Digital Microfluidic Biochip (DMFB) is a digitally controllable lab-on-a-chip. Droplets of fluids are moved, merged and mixed on a grid. Routing these droplets efficiently has been tackled by various different approaches. We try to use temporal planning to do droplet routing, inspired by the use of it in quantum circuit compilation. We test a model for droplet routing in both classical and temporal planning and compare both versions. We show that our classical planning model is an efficient method to find droplet routes on DMFBs. Then we extend our model and include spawning, disposing, merging, splitting and mixing of droplets. The results of these extensions show that we are able to find plans for simple experiments. When scaling the problem size to real life experiments our model fails to find plans.

Cost partitioning is a technique used to calculate heuristics in classical optimal planning. It involves solving a linear program. This linear program can be decomposed into a master and pricing problems. In this thesis we combine Fourier-Motzkin elimination and the double description method in different ways to precompute the generating rays of the pricing problems. We further empirically evaluate these approaches and propose a new method that replaces the Fourier-Motzkin elimination. Our new method improves the performance of our approaches with respect to runtime and peak memory usage.

The increasing number of data nowadays has contributed to new scheduling approaches. Aviation is one of the domains concerned the most, as the aircraft engine implies millions of maintenance events operated by staff worldwide. In this thesis we present a constraint programming-based algorithm to solve the aircraft maintenance scheduling problem. We want to find the best time to do the maintenance by determining which employee will perform the work and when. Here we report how the scheduling process in aviation can be automatized.

To solve stochastic state-space tasks, the research field of artificial intelligence is mainly used. PROST2014 is state of the art when determining good actions in an MDP environment. In this thesis, we aimed to provide a heuristic by using neural networks to outperform the dominating planning system PROST2014. For this purpose, we introduced two variants of neural networks that allow to estimate the respective Q-value for a pair of state and action. Since we envisaged the learning method of supervised learning, in addition to the architecture as well as the components of the neural networks, the generation of training data was also one of the main tasks. To determine the most suitable network parameters, we performed a sequential parameter search, from which we expected a local optimum of the model settings. In the end, the PROST2014 planning system could not be surpassed in the total rating evaluation. Nevertheless, in individual domains, we could establish increased final scores on the side of the neural networks. The result shows the potential of this approach and points to eventual adaptations in future work pursuing this procedure furthermore.

In classical planning, there are tasks that are hard and tasks that are easy. We can measure the complexity of a task with the correlation complexity, the improvability width, and the novelty width. In this work, we compare these measures.

We investigate what causes a correlation complexity of at least 2. To do so we translate the state space into a vector space which allows us to make use of linear algebra and convex cones.

Additionally, we introduce the Basel measure, a new measure that is based on potential heuristics and therefore similar to the correlation complexity but also comparable to the novelty width. We show that the Basel measure is a lower bound for the correlation complexity and that the novelty width +1 is an upper bound for the Basel measure.

Furthermore, we compute the Basel measure for some tasks of the International Planning Competitions and show that the translation of a task can increase the Basel measure by removing seemingly irrelevant state variables.

Unsolvability is an important result in classical planning and has seen increased interest in recent years. This thesis explores unsolvability detection by automatically generating parity arguments, a well-known way of proving unsolvability. The argument requires an invariant measure, whose parity remains constant across all reachable states, while all goal states are of the opposite parity. We express parity arguments using potential functions in the field F 2 . We develop a set of constraints that describes potential functions with the necessary separating property, and show that the constraints can be represented efficiently for up to two-dimensional features. Enhanced with mutex information, an algorithm is formed that tests whether a parity function exists for a given planning task. The existence of such a function proves the task unsolvable. To determine its practical use, we empirically evaluate our approach on a benchmark of unsolvable problems and compare its performance to a state of the art unsolvability planner. We lastly analyze the arguments found by our algorithm to confirm their validity, and understand their expressive power.

We implemented the invariant synthesis algorithm proposed by Rintanen and experimentally compared it against Helmert’s mutex group synthesis algorithm as implemented in Fast Downward.

The context for the comparison is the translation of propositional STRIPS tasks to FDR tasks, which requires the identification of mutex groups.

Because of its dominating lead in translation speed, combined with few and marginal advantages in performance during search, Helmert’s algorithm is clearly better for most uses. Meanwhile Rintanen’s algorithm is capable of finding invariants other than mutexes, which Helmert’s algorithm per design cannot do.

The International Planning Competition (IPC) is a competition of state-of-the-art planning systems. The evaluation of these planning systems is done by measuring them with different problems. It focuses on the challenges of AI planning by analyzing classical, probabilistic and temporal planning and by presenting new problems for future research. Some of the probabilistic domains introduced in IPC 2018 are Academic Advising, Chromatic Dice, Cooperative Recon, Manufacturer, Push Your Luck, Red-finned Blue-eyes, etc.

This thesis aims to solve (near)-optimally two probabilistic IPC 2018 domains, Academic Advising and Chromatic Dice. We use different techniques to solve these two domains. In Academic Advising, we use a relevance analysis to remove irrelevant actions and state variables from the planning task. We then convert the problem from probabilistic to classical planning, which helped us solve it efficiently. In Chromatic Dice, we implement backtracking search to solve the smaller instances optimally. More complex instances are partitioned into several smaller planning tasks, and a near-optimal policy is derived as a combination of the optimal solutions to the small instances.

The motivation for finding (near)-optimal policies is related to the IPC score, which measures the quality of the planners. By providing the optimal upper bound of the domains, we contribute to the stabilization of the IPC score evaluation metric for these domains.

Most well-known and traditional online planners for probabilistic planning are in some way based on Monte-Carlo Tree Search. SOGBOFA, symbolic online gradient-based optimization for factored action MDPs, offers a new perspective on this: it constructs a function graph encoding the expected reward for a given input state using independence assumptions for states and actions. On this function, they use gradient ascent to perform a symbolic search optimizing the actions for the current state. This unique approach to probabilistic planning has shown very strong results and even more potential. In this thesis, we attempt to integrate the new ideas SOGBOFA presents into the traditionally successful Trial-based Heuristic Tree Search framework. Specifically, we design and evaluate two heuristics based on the aforementioned graph and its Q value estimations, but also the search using gradient ascent. We implement and evaluate these heuristics in the Prost planner, along with a version of the current standalone planner.

In this thesis, we consider cyclical dependencies between landmarks for cost-optimal planning. Landmarks denote properties that must hold at least once in all plans. However, if the orderings between them induce cyclical dependencies, one of the landmarks in each cycle must be achieved an additional time. We propose the generalized cycle-covering heuristic which considers this in addition to the cost for achieving all landmarks once.

Our research is motivated by recent applications of cycle-covering in the Freecell and logistics domain where it yields near-optimal results. We carry it over to domain-independent planning using a linear programming approach. The relaxed version of a minimum hitting set problem for the landmarks is enhanced by constraints concerned with cyclical dependencies between them. In theory, this approach surpasses a heuristic that only considers landmarks.

We apply the cycle-covering heuristic in practice where its theoretical dominance is confirmed; Many planning tasks contain cyclical dependencies and considering them affects the heuristic estimates favorably. However, the number of tasks solved using the improved heuristic is virtually unaffected. We still believe that considering this feature of landmarks offers great potential for future work.

Potential heuristics are a class of heuristics used in classical planning to guide a search algorithm towards a goal state. Most of the existing research on potential heuristics is focused on finding heuristics that are admissible, such that they can be used by an algorithm such as A* to arrive at an optimal solution. In this thesis, we focus on the computation of potential heuristics for satisficing planning, where plan optimality is not required and the objective is to find any solution. Specifically, our focus is on the computation of potential heuristics that are descending and dead-end avoiding (DDA), since these prop- erties guarantee favorable search behavior when used with greedy search algorithms such as hillclimbing. We formally prove that the computation of DDA heuristics is a PSPACE-complete problem and propose several approximation algorithms. Our evaluation shows that the resulting heuristics are competitive with established approaches such as Pattern Databases in terms of heuristic quality but suffer from several performance bottlenecks.

Most automated planners use heuristic search to solve the tasks. Usually, the planners get as input a lifted representation of the task in PDDL, a compact formalism describing the task using a fragment of first-order logic. The planners then transform this task description into a grounded representation where the task is described in propositional logic. This new grounded format can be exponentially larger than the lifted one, but many planners use this grounded representation because it is easier to implement and reason about.

However, sometimes this transformation between lifted and grounded representations is not tractable. When this is the case, there is not much that planners based on heuristic search can do. Since this transformation is a required preprocess, when this fails, the whole planner fails.

To solve the grounding problem, we introduce new methods to deal with tasks that cannot be grounded. Our work aims to find good ways to perform heuristic search while using a lifted representation of planning problems. We use the point-of-view of planning as a database progression problem and borrow solutions from the areas of relational algebra and database theory.

Our theoretical and empirical results are motivating: several instances that were never solved by any planner in the literature are now solved by our new lifted planner. For example, our planner can solve the challenging Organic Synthesis domain using a breadth-first search, while state-of-the-art planners cannot solve more than 60% of the instances. Furthermore, our results offer a new perspective and a deep theoretical study of lifted representations for planning tasks.

The generation of independently verifiable proofs for the unsolvability of planning tasks using different heuristics, including linear Merge-and-Shrink heuristics, is possible by usage of a proof system framework. Proof generation in the case of non-linear Merge-and-Shrink heuristic, however, is currently not supported. This is due to the lack of a suitable state set representation formalism that allows to compactly represent states mapped to a certain value in the belonging Merge-and-Shrink representation (MSR). In this thesis, we overcome this shortcoming using Sentential Decision Diagrams (SDDs) as set representations. We describe an algorithm that constructs the desired SDD from the MSR, and show that efficient proof verification is possible with SDDs as representation formalism. Aditionally, we use a proof of concept implementation to analyze the overhead occurred by the proof generation functionality and the runtime of the proof verification.

The operator-counting framework is a framework in classical planning for heuristics that are based on linear programming. The operator-counting framework covers several kinds of state-of-the-art linear programming heuristics, among them the post-hoc optimization heuristic. In this thesis we will use post-hoc optimization constraints and evaluate them under altered cost functions instead of the original cost function of the planning task. We show that such cost-altered post-hoc optimization constraints are also covered by the operator-counting framework and that it is possible to achieve improved heuristic estimates with them, compared with post-hoc optimization constraints under the original cost function. In our experiments we have not been able to achieve improved problem coverage, as we were not able to find a method for generating favorable cost functions that work well in all domains.

Heuristic forward search is the state-of-the-art approach to solve classical planning problems. On the other hand, bidirectional heuristic search has a lot of potential but was never able to deliver on those expectations in practice. Only recently the near-optimal bidirectional search algorithm (NBS) was introduces by Chen et al. and as the name suggests, NBS expands nearly the optimal number of states to solve any search problem. This is a novel achievement and makes the NBS algorithm a very promising and efficient algorithm in search. With this premise in mind, we raise the question of how applicable NBS is to planning. In this thesis, we inquire this very question by implementing NBS in the state- of-the-art planner Fast-Downward and analyse its performance on the benchmark of the latest international planning competition. We additionally implement fractional meet-in- the-middle and computeWVC to analyse NBS’ performance more thoroughly in regards to the structure of the problem task.

The conducted experiments show that NBS can successfully be applied to planning as it was able to consistently outperform A*. Especially good results were achieved on the domains: blocks, driverlog, floortile-opt11-strips, get-opt14-strips, logistics00, and termes- opt18-strips. Analysing these results, we deduce that the efficiency of forward and backward search depends heavily upon the underlying implicit structure of the transition system which is induced by the problem task. This suggests that bidirectional search is inherently more suited for certain problems. Furthermore, we find that this aptitude for a certain search direction correlates with the domain, thereby providing a powerful analytic tool to a priori derive the effectiveness of certain search approaches.

In conclusion, even without intricate improvements the NBS algorithm is able to compete with A*. It therefore has further potential for future research. Additionally, the underlying transition system of a problem instance is shown to be an important factor which influences the efficiency of certain search approaches. This knowledge could be valuable for devising portfolio planners.

Multiple Sequence Alignment (MSA) is the problem of aligning multiple biological sequences in the evoluationary most plausible way. It can be viewed as a shortest path problem through an n-dimensional lattice. Because of its large branching factor of 2^n − 1, it has found broad attention in the artificial intelligence community. Finding a globally optimal solution for more than a few sequences requires sophisticated heuristics and bounding techniques in order to solve the problem in acceptable time and within memory limitations. In this thesis, we show how existing heuristics fall into the category of combining certain pattern databases. We combine arbitrary pattern collections that can be used as heuristic estimates and apply cost partitioning techniques from classical planning for MSA. We implement two of those heuristics for MSA and compare their estimates to the existing heuristics.

Increasing Cost Tree Search is a promising approach to multi-agent pathfinding problems, but like all approaches it has to deal with a huge number of possible joint paths, growing exponentially with the number of agents. We explore the possibility of reducing this by introducing a value abstraction to the Multi-valued Decision Diagrams used to represent sets of joint paths. To that end we introduce a heat map to heuristically judge how collisionprone agent positions are and present how to use and possible refine abstract positions in order to still find valid paths.

Estimating cheapest plan costs with the help of network flows is an established technique. Plans and network flows are already very similar, however network flows can differ from plans in the presence of cycles. If a transition system contains cycles, flows might be composed of multiple disconnected parts. This discrepancy can make the cheapest plan estimation worse. One idea to get rid of the cycles works by introducing time steps. For every time step the states of a transition system are copied. Transitions will be changed, so that they connect states only with states of the next time step, which ensures that there are no cycles. It turned out, that by applying this idea to multiple transitions systems, network flows of the individual transition systems can be synchronized via the time steps to get a new kind of heuristic, that will also be discussed in this thesis.

Probabilistic planning is a research field that has become popular in the early 1990s. It aims at finding an optimal policy which maximizes the outcome of applying actions to states in an environment that feature unpredictable events. Such environments can consist of a large number of states and actions which make finding an optimal policy intractable using classical methods. Using a heuristic function for a guided search allows for tackling such problems. Designing a domain-independent heuristic function requires complex algorithms which may be expensive when it comes to time and memory consumption.

In this thesis, we are applying the supervised learning techniques for learning two domain-independent heuristic functions. We use three types of gradient descent methods: stochastic, batch and mini-batch gradient descent and their improved versions using momen- tum, learning decay rate and early stopping. Furthermore, we apply the concept of feature combination in order to better learn the heuristic functions. The learned functions are pro- vided to Prost, a domain-independent probabilistic planner, and benchmarked against the winning algorithms of the International Probabilistic Planning Competition held in 2014. The experiments show that learning an offline heuristic improves the overall score of the search for some of the domains used in aforementioned competition.

The merge-and-shrink heuristic is a state-of-the-art admissible heuristic that is often used for optimal planning. Recent studies showed that the merge strategy is an important factor for the performance of the merge-and-shrink algorithm. There are many different merge strategies and improvements for merge strategies described in the literature. One out of these merge strategies is MIASM by Fan et al. MIASM tries to merge transition systems that produce unnecessary states in their product which can be pruned. Another merge strategy is the symmetry-based merge-and-shrink framework by Sievers et al. This strategy tries to merge transition systems that cause factored symmetries in their product. This strategy can be combined with other merge strategies and it often improves the performance for many merge strategy. However, the current combination of MIASM with factored symmetries performs worse than MIASM. We implement a different combination of MIASM that uses factored symmetries during the subset search of MIASM. Our experimental evaluation shows that our new combination of MIASM with factored symmetries solves more tasks than the existing MIASM and the previously implemented combination of MIASM with factored symmetries. We also evaluate different combinations of existing merge strategies and find combinations that perform better than their basic version that were not evaluated before.

Tree Cache is a pathfinding algorithm that selects one vertex as a root and constructs a tree with cheapest paths to all other vertices. A path is found by traversing up the tree from both the start and goal vertices to the root and concatenating the two parts. This is fast, but as all paths constructed this way pass through the root vertex they can be highly suboptimal.

To improve this algorithm, we consider two simple approaches. The first is to construct multiple trees, and save the distance to each root in each vertex. To find a path, the algorithm first selects the root with the lowest total distance. The second approach is to remove redundant vertices, i.e. vertices that are between the root and the lowest common ancestor (LCA) of the start and goal vertices. The performance and space requirements of the resulting algorithm are then compared to the conceptually similar hub labels and differential heuristics.

Greedy Best-First Search (GBFS) is a prominent search algorithm to find solutions for planning tasks. GBFS chooses nodes for further expansion based on a distance-to-goal estimator, the heuristic. This makes GBFS highly dependent on the quality of the heuristic. Heuristics often face the problem of producing Uninformed Heuristic Regions (UHRs). GBFS additionally suffers the possibility of simultaneously expanding nodes in multiple UHRs. In this thesis we change the heuristic approach in UHRs. The heuristic was unable to guide the search and so we try to expand novel states to escape the UHRs. The novelty measures how “new” a state is in the search. The result is a combination of heuristic and novelty guided search, which is indeed able to escape UHRs quicker and solve more problems in reasonable time.

In classical AI planning, the state explosion problem is a reoccurring subject: although the problem descriptions are compact, often a huge number of states needs to be considered. One way to tackle this problem is to use static pruning methods which reduce the number of variables and operators in the problem description before planning.

In this work, we discuss the properties and limitations of three existing static pruning techniques with a focus on satisficing planning. We analyse these pruning techniques and their combinations, and identify synergy effects between them and the domains and problem structures in which they occur. We implement the three methods into an existing propositional planner, and evaluate the performance of different configurations and combinations in a set of experiments on IPC benchmarks. We observe that static pruning techniques can increase the number of solved problems, and that the synergy effects of the combinations also occur on IPC benchmarks, although they do not lead to a major performance increase.

The goal of classical domain-independent planning is to find a sequence of actions which lead from a given initial state to a goal state that satisfies some goal criteria. Most planning systems use heuristic search algorithms to find such a sequence of actions. A critical part of heuristic search is the heuristic function. In order to find a sequence of actions from an initial state to a goal state efficiently this heuristic function has to guide the search towards the goal. It is difficult to create such an efficient heuristic function. Arfaee et al. show that it is possible to improve a given heuristic function by applying machine learning techniques on a single domain in the context of heuristic search. To achieve this improvement of the heuristic function, they propose a bootstrap learning approach which subsequently improves the heuristic function.

In this thesis we will introduce a technique to learn heuristic functions that can be used in classical domain-independent planning based on the bootstrap-learning approach introduced by Arfaee et al. In order to evaluate the performance of the learned heuristic functions, we have implemented a learning algorithm for the Fast Downward planning system. The experiments have shown that a learned heuristic function generally decreases the number of explored states compared to blind-search . The total time to solve a single problem increases because the heuristic function has to be learned before it can be applied.

Essential for the estimation of the performance of an algorithm in satisficing planning is its ability to solve benchmark problems. Those results can not be compared directly as they originate from different implementations and different machines. We implemented some of the most promising algorithms for greedy best-first search, published in the last years, and evaluated them on the same set of benchmarks. All algorithms are either based on randomised search, localised search or a combination of both. Our evaluation proves the potential of those algorithms.

Heuristic search with admissible heuristics is the leading approach to cost-optimal, domain-independent planning. Pattern database heuristics - a type of abstraction heuristics - are state-of-the-art admissible heuristics. Two recent pattern database heuristics are the iPDB heuristic by Haslum et al. and the PhO heuristic by Pommerening et al.

The iPDB procedure performs a hill climbing search in the space of pattern collections and evaluates selected patterns using the canonical heuristic. We apply different techniques to the iPDB procedure, improving its hill climbing algorithm as well as the quality of the resulting heuristic. The second recent heuristic - the PhO heuristic - obtains strong heuristic values through linear programming. We present different techniques to influence and improve on the PhO heuristic.

We evaluate the modified iPDB and PhO heuristics on the IPC benchmark suite and show that these abstraction heuristics can compete with other state-of-the-art heuristics in cost-optimal, domain-independent planning.

Greedy best-first search (GBFS) is a prominent search algorithm for satisficing planning - finding good enough solutions to a planning task in reasonable time. GBFS selects the next node to consider based on the most promising node estimated by a heuristic function. However, this behaviour makes GBFS heavily depend on the quality of the heuristic estimator. Inaccurate heuristics can lead GBFS into regions far away from a goal. Additionally, if the heuristic ranks several nodes the same, GBFS has no information on which node it shall follow. Diverse best-first search (DBFS) is a new algorithm by Imai and Kishimoto [2011] which has a local search component to emphasis exploitation. To enable exploration, DBFS deploys probabilities to select the next node.

In two problem domains, we analyse GBFS' search behaviour and present theoretical results. We evaluate these results empirically and compare DBFS and GBFS on constructed as well as on provided problem instances.

State-of-the-art planning systems use a variety of control knowledge in order to enhance the performance of heuristic search. Unfortunately most forms of control knowledge use a specific formalism which makes them hard to combine. There have been several approaches which describe control knowledge in Linear Temporal Logic (LTL). We build upon this work and propose a general framework for encoding control knowledge in LTL formulas. The framework includes a criterion that any LTL formula used in it must fulfill in order to preserve optimal plans when used for pruning the search space; this way the validity of new LTL formulas describing control knowledge can be checked. The framework is implemented on top of the Fast Downward planning system and is tested with a pruning technique called Unnecessary Action Application, which detects if a previously applied action achieved no useful progress.

Landmarks are known to be useable for powerful heuristics for informed search. In this thesis, we explain and evaluate a novel algorithm to find ordered landmarks of delete free tasks by intersecting solutions in the relaxation. The proposed algorithm efficiently finds landmarks and natural orders of delete free tasks, such as delete relaxations or Pi-m compilations.

Planning as heuristic search is the prevalent technique to solve planning problems of any kind of domains. Heuristics estimate distances to goal states in order to guide a search through large state spaces. However, this guidance is sometimes moderate, since still a lot of states lie on plateaus of equally prioritized states in the search space topology. Additional techniques that ignore or prefer some actions for solving a problem are successful to support the search in such situations. Nevertheless, some action pruning techniques lead to incomplete searches.

We propose an under-approximation refinement framework for adding actions to under-approximations of planning tasks during a search in order to find a plan. For this framework, we develop a refinement strategy. Starting a search on an initial under-approximation of a planning task, the strategy adds actions determined at states close to a goal, whenever the search does not progress towards a goal, until a plan is found. Key elements of this strategy consider helpful actions and relaxed plans for refinements. We have implemented the under-approximation refinement framework into the greedy best first search algorithm. Our results show considerable speedups for many classical planning problems. Moreover, we are able to plan with fewer actions than standard greedy best first search.

The main approach for classical planning is heuristic search. Many cost heuristics are based on the delete relaxation. The optimal heuristic of a delete free planning problem is called h + . This thesis explores two new ways to compute h + . Both approaches use factored planning, which decomposes the original planning problem to work on each subproblem separately. The algorithm reuses the subsolutions and combines them to a global solution.

The two algorithms are used to compute a cost heuristic for an A* search. As both approaches compute the optimal heuristic for delete free planning tasks, the algorithms can also be used to find a solution for relaxed planning tasks.

Multi-Agent-Path-Finding (MAPF) is a common problem in robotics and memory management. Pebbles in Motion is an implementation of a problem solver for MAPF in polynomial time, based on a work by Daniel Kornhauser from 1984. Recently a lot of research papers have been published on MAPF in the research community of Artificial Intelligence, but the work by Kornhauser seems hardly to be taken into account. We assumed that this might be related to the fact that said paper was more mathematically and hardly describing algorithms intuitively. This work aims at filling this gap, by providing an easy understandable approach of implementation steps for programmers and a new detailed description for researchers in Computer Science.

Bachelor's theses

Fast Downward is a classical planner using heuristical search. The planner uses many advanced planning techniques that are not easy to teach, since they usually rely on complex data structures. To introduce planning techniques to the user an interactive application is created. This application uses an illustrative example to showcase planning techniques: Blocksworld

Blocksworld is an easy understandable planning problem which allows a simple representation of a state space. It is implemented in the Unreal Engine and provides an interface to the Fast Downward planner. Users can explore a state space themselves or have Fast Downward generate plans for them. The concept of heuristics as well as the state space are explained and made accessible to the user. The user experiences how the planner explores a state space and which techniques the planner uses.

This thesis is about implementing Jussi Rintanen’s algorithm for schematic invariants. The algo- rithm is implemented in the planning tool Fast Downward and refers to Rintanen’s paper Schematic Invariants by Reduction to Ground Invariants. The thesis describes all necessary definitions to under- stand the algorithm and draws a comparison between the original task and a reduced task in terms of runtime and number of grounded actions.

Planning is a field of Artificial Intelligence. Planners are used to find a sequence of actions, to get from the initial state to a goal state. Many planning algorithms use heuristics, which allow the planner to focus on more promising paths. Pattern database heuristics allow us to construct such a heuristic, by solving a simplified version of the problem, and saving the associated costs in a pattern database. These pattern databases can be computed and stored by using symbolic data structures.

In this paper we will look at how pattern databases using symbolic data structures using binary decision diagrams and algebraic decision diagrams can be implemented. We will extend fast down- ward (Helmert [2006]) with it, and compare the performance of this implementation with the already implemented explicit pattern database.

In the field of automated planning and scheduling, a planning task is essentially a state space which can be defined rigorously using one of several different formalisms (e.g. STRIPS, SAS+, PDDL etc.). A planning algorithm tries to determine a sequence of actions that lead to a goal state for a given planning task. In recent years, attempts have been made to group certain planners together into so called planner portfolios, to try and leverage their effectiveness on different specific problem classes. In our project, we create an online planner which in contrast to its offline counterparts, makes use of task specific information when allocating a planner to a task. One idea that has recently gained interest, is to apply machine learning methods to planner portfolios.

In previous work such as Delfi (Katz et al., 2018; Sievers et al., 2019a) supervised learning techniques were used, which made it necessary to train multiple networks to be able to attempt multiple, potentially different, planners for a given task. The reason for this being that, if we used the same network, the output would always be the same, as the input to the network would remain unchanged. In this project we make use of techniques from rein- forcement learning such as DQNs (Mnih et al., 2013). Using RL approaches such as DQNs, allows us to extend the input to the network to include information on things, such as which planners were previously attempted and for how long. As a result multiple attempts can be made after only having trained a single network.

Unfortunately the results show that current reinforcement learning agents are, amongst other reasons, too sample inefficient to be able to deliver viable results given the size of the currently available data sets.

Planning tasks are important and difficult problems in computer science. A widely used approach is the use of delete relaxation heuristics to which the additive and FF heuristic belong. Those two heuristics use a graph in their calculation, which only has to be constructed once for a planning task but then can be used repeatedly. To solve such a problem efficiently it is important that the calculation of the heuristics are fast. In this thesis the idea to achieve a faster calculation is to combine redundant parts of the graph when building it to reduce the number of edges and therefore speed up the calculation. Here the reduction of the redundancies is done for each action within a planning task individually, but further ideas to simplify over all actions are also discussed.

Monte Carlo search methods are widely known, mostly for their success in game domains, although they are also applied to many non-game domains. In previous work done by Schulte and Keller, it was established that best-first searches could adapt to the action selection functionality which make Monte Carlo methods so formidable. In practice however, the trial-based best first search, without exploration, was shown to be slightly slower than its explicit open list counterpart. In this thesis we examine the non-trial and trial-based searches and how they can address the exploitation exploration dilemma. Lastly, we will see how trial-based BFS can rectify a slower search by allowing occasional random action selection, by comparing it to regular open list searches in a line of experiments.

Sudoku has become one of the world’s most popular logic puzzles, arousing interest in the general public and the science community. Although the rules of Sudoku may seem simple, they allow for nearly countless puzzle instances, some of which are very hard to solve. SAT-solvers have proven to be a suitable option to solve Sudokus automatically. However, they demand the puzzles to be encoded as logical formulae in Conjunctive Normal Form. In earlier work, such encodings have been successfully demonstrated for original Sudoku Puzzles. In this thesis, we present encodings for rather unconventional Sudoku Variants, developed by the puzzle community to create even more challenging solving experiences. Furthermore, we demonstrate how Pseudo-Boolean Constraints can be utilized to encode Sudoku Variants that follow rules involving sums. To implement an encoding of Pseudo-Boolean Constraints, we use Binary Decision Diagrams and Adder Networks and study how they compare to each other.

In optimal classical planning, informed search algorithms like A* need admissible heuristics to find optimal solutions. Counterexample-guided abstraction refinement (CEGAR) is a method used to generate abstractions that yield suitable abstraction heuristics iteratively. In this thesis, we propose a class of CEGAR algorithms for the generation of domain abstractions, which are a class of abstractions that rank in between projections and Cartesian abstractions regarding the grade of refinement they allow. As no known algorithm constructs domain abstractions, we show that our algorithm is competitive with CEGAR algorithms that generate one projection or Cartesian abstraction.

This thesis will look at Single-Player Chess as a planning domain using two approaches: one where we look at how we can encode the Single-Player Chess problem as a domain-independent (general-purpose AI) approach and one where we encode the problem as a domain-specific solver. Lastly, we will compare the two approaches by doing some experiments and comparing the results of the two approaches. Both the domain-independent implementation and the domain-specific implementation differ from traditional chess engines because the task of the agent is not to find the best move for a given position and colour, but the agent’s task is to check if a given chess problem has a solution or not. If the agent can find a solution, the given chess puzzle is valid. The results of both approaches were measured in experiments, and we found out that the domain-independent implementation is too slow and that the domain-specific implementation, on the other hand, can solve the given puzzles reliably, but it has a memory bottleneck rooted in the search method that was used.

Carcassonne is a tile-based board game with a large state space and a high branching factor and therefore poses a challenge to artificial intelligence. In the past, Monte Carlo Tree Search (MCTS), a search algorithm for sequential decision-making processes, has been shown to find good solutions in large state spaces. MCTS works by iteratively building a game tree according to a tree policy. The profitability of paths within that tree is evaluated using a default policy, which influences in what directions the game tree is expanded. The functionality of these two policies, as well as other factors, can be implemented in many different ways. In consequence, many different variants of MCTS exist. In this thesis, we applied MCTS to the domain of two-player Carcassonne and evaluated different variants in regard to their performance and runtime. We found significant differences in performance for various variable aspects of MCTS and could thereby evaluate a configuration which performs best on the domain of Carcassonne. This variant consistently outperformed an average human player with a feasible runtime.

In general, it is important to verify software as it is prone to error. This also holds for solving tasks in classical planning. So far, plans in general as well as the fact that there is no plan for a given planning task can be proven and independently verified. However, no such proof for the optimality of a solution of a task exists. Our aim is to introduce two methods with which optimality can be proven and independently verified. We first reduce unit cost tasks to unsolvable tasks, which enables us to make use of the already existing certificates for unsolvability. In a second approach, we propose a proof system for optimality, which enables us to infer that the determined cost of a task is optimal. This permits the direct generation of optimality certificates.

Pattern databases are one of the most powerful heuristics in classical planning. They evaluate the perfect cost for a simplified sub-problem. The post-hoc optimization heuristic is a technique on how to optimally combine a set of pattern databases. In this thesis, we will adapt the post-hoc optimization heuristic for the sliding tile puzzle. The sliding tile puzzle serves as a benchmark to compare the post-hoc optimization heuristic to already established methods, which also deal with the combining of pattern databases. We will then show how the post-hoc optimization heuristic is an improvement over the already established methods.

In this thesis, we generate landmarks for a logistics-specific task. Landmarks are actions that need to occur at least once in every plan. A landmark graph denotes a structure with landmarks and their edges called orderings. If there are cycles in a landmark graph, one of those landmarks needs to be achieved at least twice for every cycle. The generation of the logistics-specific landmarks and their orderings calculate the cyclic landmark heuristic. The task is to pick up on related work, the evaluation of the cyclic landmark heuristic. We compare the generation of landmark graphs from a domain-independent landmark generator to a domain-specific landmark generator, the latter being the focus. We aim to bridge the gap between domain-specific and domain-independent landmark generators. In this thesis, we compare one domain-specific approach for the logistics domain with results from a domain- independent landmark generator. We devise a unit to pre-process data for other domain- specific tasks as well. We will show that specificity is better suited than independence.

Lineare Programmierung ist eine mathematische Modellierungstechnik, bei der eine lineare Funktion, unter der Berücksichtigung verschiedenen Beschränkungen, maximiert oder minimiert werden soll. Diese Technik ist besonders nützlich, falls Entscheidungen für Optimierungsprobleme getroffen werden sollen. Ziel dieser Arbeit war es ein Tool für das Spiel Factory Town zu entwickeln, mithilfe man Optimierungsanfragen bearbeiten kann. Dabei ist es möglich wahlweise zwischen diversen Fragestellungen zu wählen und anhand von LP-\ IP-Solvern diese zu beantworten. Zudem wurden die mathematischen Formulierungen, sowie die Unterschiede beider Methoden angegangen. Schlussendlich unterstrichen die generierten Resultate, dass LP Lösungen mindestens genauso gut oder sogar besser seien als die Lösungen eines IP.

Symbolic search is an important approach to classical planning. Symbolic search uses search algorithms that process sets of states at a time. For this we need states to be represented by a compact data structure called knowledge compilations. Merge-and-shrink representations come a different field of planning, where they have been used to derive heuristic functions for state-space search. More generally they represent functions that map variable assignments to a set of values, as such we can regard them as a data structure we will call Factored Mappings. In this thesis, we will investigate Factored Mappings (FMs) as a knowledge compilation language with the hope of using them for symbolic search. We will analyse the necessary transformations and queries for FMs, by defining the needed operations and a canonical representation of FMs, and showing that they run in polynomial time. We will then show that it is possible to use Factored Mappings as a knowledge compilation for symbolic search by defining a symbolic search algorithm for a finite-domain plannings task that works with FMs.

Version control systems use a graph data structure to track revisions of files. Those graphs are mutated with various commands by the respective version control system. The goal of this thesis is to formally define a model of a subset of Git commands which mutate the revision graph, and to model those mutations as a planning task in the Planning Domain Definition Language. Multiple ways to model those graphs will be explored and those models will be compared by testing them using a set of planners.

Pattern Databases are admissible abstraction heuristics for classical planning. In this thesis we are introducing the Boosting processes, which consists of enlarging the pattern of a Pattern Database P, calculating a more informed Pattern Database P' and then min-compress P' to the size of P resulting in a compressed and still admissible Pattern Database P''. We design and implement two boosting algorithms, Hillclimbing and Randomwalk.

We combine pattern database heuristics using five different cost partitioning methods. The experiments compare computing cost partitionings over regular and boosted pattern databases. The experiments, performed on IPC (optimal track) tasks, show promising results which increased the coverage (number of solved tasks) by 9 for canonical cost partitioning using our Randomwalk boosting variant.

One dimensional potential heuristics assign a numerical value, the potential, to each fact of a classical planning problem. The heuristic value of a state is the sum over the poten- tials belonging to the facts contained in the state. Fišer et al. (2020) recently proposed to strengthen potential heuristics utilizing mutexes and disambiguations. In this thesis, we embed the same enhancements in the planning system Fast Downward. The experi- mental evaluation shows that the strengthened potential heuristics are a refinement, but too computationally expensive to solve more problems than the non-strengthened potential heuristics.

The potentials are obtained with a Linear Program. Fišer et al. (2020) introduced an additional constraint on the initial state and we propose additional constraints on random states. The additional constraints improve the amount of solved problems by up to 5%.

This thesis discusses the PINCH heuristic, a specific implementation of the additive heuristic. PINCH intends to combine the strengths of existing implementations of the additive heuristic. The goal of this thesis is to really dig into the PINCH heuristic. I want to provide the most accessible resource for understanding PINCH and I want to analyze the performance of PINCH by comparing it to the algorithm on which it is based, Generalized Dijkstra.

Suboptimal search algorithms can offer attractive benefits compared to optimal search, namely increased coverage of larger search problems and quicker search times. Improving on such algorithms, such as reducing costs further towards optimal solutions and reducing the number of node expansions, is therefore a compelling area for further research. This paper explores the utility and scalability of recently developed priority functions, XDP, XUP, and PWXDP, and the Improved Optimistic Search algorithm, compared to Weighted A*, in the Fast Downward planner. Analyses focus on the cost, total time, coverage, and node expansion parameters, with experimental evidence suggesting preferable performance if strict optimality is not desired. The implementation of priorityb functions in eager best-first search showed marked improvements compared to A* search on coverage, total time, and number of expansions, without significant cost penalties. Following previous suboptimal search research, experimental evidence even seems to indicate that these cost penalties do not reach the designated bound, even in larger search spaces.

In the Automated Planning field, algorithms and systems are developed for exploring state spaces and ultimately finding an action sequence leading from a task’s initial state to its goal. Such planning systems may sometimes show unexpected behavior, caused by a planning task or a bug in the planner itself. Generally speaking, finding the source of a bug tends to be easier when the cause can be isolated or simplified. In this thesis, we tackle this problem by making PDDL and SAS+ tasks smaller while ensuring they still invoke a certain characteristic when executed with a planner. We implement a system that successively removes elements, such as objects, from a task and checks whether the transformed task still fails on the planner. Elements are removed in a syntactically consistent way, however, no semantic integrity is enforced. Our system’s design is centered around the Fast Downward Planning System, as we re-use some of its translator modules and all test runs are performed with Fast Downward. At the core of our system, first-choice hill-climbing is used for optimization. Our “minimizer” takes (1) a failing planner execution command, (2) a description of the failing characteristic and (3) the type of element to be deleted as arguments. We evaluate our system’s functionality on the basis of three use-cases. In our most successful test runs, (1) a SAS+ task with initially 1536 operators and 184 variables is reduced to 2 operators and 2 variables and (2)a PDDL task with initially 46 actions, 62 objects and 29 predicate symbols is reduced to 2 actions, 6 objects and 4 predicates.

Fast Downward is a classical planning system based on heuristic search. Its successor generator is an efficient and intelligent tool to process state spaces and generate their successor states. In this thesis we implement different successor generators in the Fast Downward planning system and compare them against each other. Apart from the given fast downward successor generator we implement four other successor generators: a naive successor generator, one based on the marking of delete relaxed heuristics, one based on the PSVN planning system and one based on watched literals as used in modern SAT solvers. These successor generators are tested in a variety of different planning benchmarks to see how well they compete against each other. We verified that there is a trade-off between precomputation and faster successor generation and showed that all of the implemented successor generators have a use case and it is advisable to switch to a successor generator that fits the style of the planning task.

Verifying whether a planning algorithm came to the correct result for a given planning task is easy if a plan is emitted which solves the problem. But if a task is unsolvable most planners just state this fact without any explanation or even proof. In this thesis we present extended versions of the symbolic search algorithms SymPA and symbolic bidirectional uniform-cost search which, if a given planning task is unsolvable, provide certificates which prove unsolvability. We also discuss a concrete implementation of this version of SymPA.

Classical planning is an attractive approach to solving problems because of its generality and its relative ease of use. Domain-specific algorithms are appealing because of their performance, but require a lot of resources to be implemented. In this thesis we evaluate concepts languages as a possible input language for expert domain knowledge into a planning system. We also explore mixed integer programming as a way to use this knowledge to improve search efficiency and to help the user find and refine useful domain knowledge.

Classical Planning is a branch of artificial intelligence that studies single agent, static, deterministic, fully observable, discrete search problems. A common challenge in this field is the explosion of states to be considered when searching for the goal. One technique that has been developed to mitigate this is Strong Stubborn Set based pruning, where on each state expansion, the considered successors are restricted to Strong Stubborn Sets, which exploit the properties of independent operators to cut down the tree or graph search. We adopt the definitions of the theory of Strong Stubborn Sets from the SAS+ setting to transition systems and validate a central theorem about the correctness of Strong Stubborn Set based pruning for transition systems in the interactive theorem prover Isabelle/HOL.

Ein wichtiges Feld in der Wissenschaft der künstliche Intelligenz sind Planungsprobleme. Man hat das Ziel, eine künstliche intelligente Maschine zu bauen, die mit so vielen ver- schiedenen Probleme umgehen und zuverlässig lösen kann, indem sie ein optimaler Plan herstellt.

Der Trial-based Heuristic Tree Search(THTS) ist ein mächtiges Werkzeug um Multi-Armed- Bandit-ähnliche Probleme, Marcow Decsision Processe mit verändernden Rewards, zu lösen. Beim momentanen THTS können explorierte gefundene gute Rewards auf Grund von der grossen Anzahl der Rewards nicht beachtet werden. Ebenso können beim explorieren schlech- te Rewards, gute Knoten im Suchbaum, verschlechtern. Diese Arbeit führt eine Methodik ein, die von der stückweise stationären MABs Problematik stammt, um den THTS weiter zu optimieren.

Abstractions are a simple yet powerful method of creating a heuristic to solve classical planning problems optimally. In this thesis we make use of Cartesian abstractions generated with Counterexample-Guided Abstraction Refinement (CEGAR). This method refines abstractions incrementally by finding flaws and then resolving them until the abstraction is sufficiently evolved. The goal of this thesis is to implement and evaluate algorithms which select solutions of such flaws, in a way which results in the best abstraction (that is, the abstraction which causes the problem to then be solved most efficiently by the planner). We measure the performance of a refinement strategy by running the Fast Downward planner on a problem and measuring how long it takes to generate the abstraction, as well as how many expansions the planner requires to find a goal using the abstraction as a heuristic. We use a suite of various benchmark problems for evaluation, and we perform this experiment for a single abstraction and on abstractions for multiple subtasks. Finally, we attempt to predict which refinement strategy should be used based on parameters of the task, potentially allowing the planner to automatically select the best strategy at runtime.

Heuristic search is a powerful paradigm in classical planning. The information generated by heuristic functions to guide the search towards a goal is a key component of many modern search algorithms. The paper “Using Backwards Generated Goals for Heuristic Planning” by Alcázar et al. proposes a way to make additional use of this information. They take the last actions of a relaxed plan as a basis to generate intermediate goals with a known path to the original goal. A plan is found when the forward search reaches an intermediate goal.

The premise of this thesis is to modify their approach by focusing on a single sequence of intermediate goals. The aim is to improve efficiency while preserving the benefits of backwards goal expansion. We propose different variations of our approach by introducing multiple ways to make decisions concerning the construction of intermediate goals. We evaluate these variations by comparing their performance and illustrate the challenges posed by this approach.

Counterexample-guided abstraction refinement (CEGAR) is a way to incrementally compute abstractions of transition systems. It starts with a coarse abstraction and then iteratively finds an abstract plan, checks where the plan fails in the concrete transition system and refines the abstraction such that the same failure cannot happen in subsequent iterations. As the abstraction grows in size, finding a solution for the abstract system becomes more and more costly. Because the abstraction grows incrementally, however, it is possible to maintain heuristic information about the abstract state space, allowing the use of informed search algorithms like A*. As the quality of the heuristic is crucial to the performance of informed search, the method for maintaining the heuristic has a significant impact on the performance of the abstraction refinement as a whole. In this thesis, we investigate different methods for maintaining the value of the perfect heuristic h* at all times and evaluate their performance.

Pattern Databases are a powerful class of abstraction heuristics which provide admissible path cost estimates by computing exact solution costs for all states of a smaller task. Said task is obtained by abstracting away variables of the original problem. Abstractions with few variables offer weak estimates, while introduction of additional variables is guaranteed to at least double the amount of memory needed for the pattern database. In this thesis, we present a class of algorithms based on counterexample-guided abstraction refinement (CEGAR), which exploit additivity relations of patterns to produce pattern collections from which we can derive heuristics that are both informative and computationally tractable. We show that our algorithms are competitive with already existing pattern generators by comparing their performance on a variety of planning tasks.

We consider the problem of Rubik’s Cube to evaluate modern abstraction heuristics. In order to find feasible abstractions of the enormous state space spanned by Rubik’s Cube, we apply projection in the form of pattern databases, Cartesian abstraction by doing counterexample guided abstraction refinement as well as merge-and-shrink strategies. While previous publications on Cartesian abstractions have not covered applicability for planning tasks with conditional effects, we introduce factorized effect tasks and show that Cartesian abstraction can be applied to them. In order to evaluate the performance of the chosen heuristics, we run experiments on different problem instances of Rubik’s Cube. We compare them by the initial h-value found for all problems and analyze the number of expanded states up to the last f-layer. These criteria provide insights about the informativeness of the considered heuristics. Cartesian Abstraction yields perfect heuristic values for problem instances close to the goal, however it is outperformed by pattern databases for more complex instances. Even though merge-and-shrink is the most general abstraction among the considered, it does not show better performance than the others.

Probabilistic planning expands on classical planning by tying probabilities to the effects of actions. Due to the exponential size of the states, probabilistic planners have to come up with a strong policy in a very limited time. One approach to optimising the policy that can be found in the available time is called metareasoning, a technique aiming to allocate more deliberation time to steps where more time to plan results in an improvement of the policy and less deliberation time to steps where an improvement of the policy with more time to plan is unlikely.

This thesis aims to adapt a recent proposal of a formal metareasoning procedure from Lin. et al. for the search algorithm BRTDP to work with the UCT algorithm in the Prost planner and compare its viability to the current standard and a number of less informed time management methods in order to find a potential improvement to the current uniform deliberation time distribution.

A planner tries to produce a policy that leads to a desired goal given the available range of actions and an initial state. A traditional approach for an algorithm is to use abstraction. In this thesis we implement the algorithm described in the ASAP-UCT paper: Abstraction of State-Action Pairs in UCT by Ankit Anand, Aditya Grover, Mausam and Parag Singla.

The algorithm combines state and state-action abstraction with a UCT-algorithm. We come to the conclusion that the algorithm needs to be improved because the abstraction of action-state often cannot detect a similarity that a reasonable action abstraction could find.

The notion of adding a form of exploration to guide a search has been proven to be an effective method of combating heuristical plateaus and improving the performance of greedy best-first search. The goal of this thesis is to take the same approach and introduce exploration in a bounded suboptimal search problem. Explicit estimation search (EES), established by Thayer and Ruml, consults potentially inadmissible information to determine the search order. Admissible heuristics are then used to guarantee the cost bound. In this work we replace the distance-to-go estimator used in EES with an approach based on the concept of novelty.

Classical domain-independent planning is about finding a sequence of actions which lead from an initial state to a goal state. A popular approach for solving planning problems efficiently is to utilize heuristic functions. A possible heuristic function is the perfect heuristic of a delete relaxed planning problem denoted as h+. Delete relaxation simplifies the planning problem thus making it easier to find a perfect heuristic. However computing h+ is still NP-hard problem.

In this thesis we discuss a promising looking approach to compute h+ in practice. Inspired by the paper from Gnad, Hoffmann and Domshlak about star-shaped planning problems, we implemented the Flow-Cut algorithm. The basic idea behind flow-cut to divide a problem that is unsolvable in practice, into smaller sub problems that can be solved. We further tested the flow-cut algorithm on the domains provided by the International Planning Competition benchmarks, resulting in the following conclusion: Using a divide and conquer approach can successfully be used to solve classical planning problems, however it is not trivial to design such an algorithm to be more efficient than state-of-the-art search algorithm.

This thesis deals with the algorithm presented in the paper "Landmark-based Meta Best-First Search Algorithm: First Parallelization Attempt and Evaluation" by Simon Vernhes, Guillaume Infantes and Vincent Vidal. Their idea was to reconsider the approach to landmarks as a tool in automated planning, but in a markedly different way than previous work had done. Their result is a meta-search algorithm which explores landmark orderings to find a series of subproblems that reliably lead to an effective solution. Any complete planner may be used to solve the subproblems. While the referenced paper also deals with an attempt to effectively parallelize the Landmark-based Meta Best-First Search Algorithm, this thesis is concerned mainly with the sequential implementation and evaluation of the algorithm in the Fast Downward planning system.

Heuristics play an important role in classical planning. Using heuristics during state space search often reduces the time required to find a solution, but constructing heuristics and using them to calculate heuristic values takes time, reducing this benefit. Constructing heuristics and calculating heuristic values as quickly as possible is very important to the effectiveness of a heuristic. In this thesis we introduce methods to bound the construction of merge-and-shrink to reduce its construction time and increase its accuracy for small problems and to bound the heuris- tic calculation of landmark cut to reduce heuristic value calculation time. To evaluate the performance of these depth-bound heuristics we have implemented them in the Fast Down- ward planning system together with three iterative-deepening heuristic search algorithms: iterative-deepening A* search, a new breadth-first iterative-deepening version of A* search and iterative-deepening breadth-first heuristic search.

Greedy best-first search has proven to be a very efficient approach to satisficing planning but can potentially lose some of its effectiveness due to the used heuristic function misleading it to a local minimum or plateau. This is where exploration with additional open lists comes in, to assist greedy best-first search with solving satisficing planning tasks more effectively. Building on the idea of exploration by clustering similar states together as described by Xie et al. [2014], where states are clustered according to heuristic values, we propose in this paper to instead cluster states based on the Hamming distance of the binary representation of states [Hamming, 1950]. The resulting open list maintains k buckets and inserts each given state into the bucket with the smallest average hamming distance between the already clustered states and the new state. Additionally, our open list is capable of reclustering all states periodically with the use of the k-means algorithm. We were able to achieve promising results concerning the amount of expansions necessary to reach a goal state, despite not achieving a higher coverage than fully random exploration due to slow performance. This was caused by the amount of calculations required to identify the most fitting cluster when inserting a new state.

Monte Carlo Tree Search Algorithms are an efficient method of solving probabilistic planning tasks that are modeled by Markov Decision Problems. MCTS uses two policies, a tree policy for iterating through the known part of the decission tree and a default policy to simulate the actions and their reward after leaving the tree. MCTS algorithms have been applied with great success to computer Go. To make the two policies fast many enhancements based on online knowledge have been developed. The goal of All Moves as First enhancements is to improve the quality of a reward estimate in the tree policy. In the context of this thesis the, in the field of computer Go very efficient, α-AMAF, Cutoff-AMAF as well as Rapid Action Value Estimation enhancements are implemented in the probabilistic planner PROST. To obtain a better default policy, Move Average Sampling is implemented into PROST and benchmarked against it’s current default policies.

In classical planning the objective is to find a sequence of applicable actions that lead from the initial state to a goal state. In many cases the given problem can be of enormous size. To deal with these cases, a prominent method is to use heuristic search, which uses a heuristic function to evaluate states and can focus on the most promising ones. In addition to applying heuristics, the search algorithm can apply additional pruning techniques that exclude applicable actions in a state because applying them at a later point in the path would result in a path consisting of the same actions but in a different order. The question remains as to how these actions can be selected without generating too much additional work to still be useful for the overall search. In this thesis we implement and evaluate the partition-based path pruning method, proposed by Nissim et al. [1], which tries to decompose the set of all actions into partitions. Based on this decomposition, actions can be pruned with very little additional information. The partition-based pruning method guarantees with some alterations to the A* search algorithm to preserve it’s optimality. The evaluation confirms that in several standard planning domains, the pruning method can reduce the size of the explored state space.

Validating real-time systems is an important and complex task which becomes exponentially harder with increasing sizes of systems. Therefore finding an automated approach to check real-time systems for possible errors is crucial. The behaviour of such real-time systems can be modelled with timed automata. This thesis adapts and implements the under-approximation refinement algorithm developed for search based planners proposed by Heusner et al. to find error states in timed automata via the directed model checking approach. The evaluation compares the algorithm to already existing search methods and shows that a basic under-approximation refinement algorithm yields a competitive search method for directed model checking which is both fast and memory efficient. Additionally we illustrate that with the introduction of some minor alterations the proposed under- approximation refinement algorithm can be further improved.

In dieser Arbeit wird versucht eine Heuristik zu lernen. Damit eine Heuristik erlernbar ist, muss sie über Parameter verfügen, die die Heuristik bestimmen. Eine solche Möglichkeit bieten Potential-Heuristiken und ihre Parameter werden Potentiale genannt. Pattern-Databases können mit vergleichsweise wenig Aufwand Eigenschaften eines Zustandsraumes erkennen und können somit eingesetzt werden als Grundlage um Potentiale zu lernen. Diese Arbeit untersucht zwei verschiedene Ansätze zum Erlernen der Potentiale aufgrund der Information aus Pattern-Databases. In Experimenten werden die beiden Ansätze genauer untersucht und schliesslich mit der FF-Heuristik verglichen.

We consider real-time strategy (RTS) games which have temporal and numerical aspects and pose challenges which have to be solved within limited search time. These games are interesting for AI research because they are more complex than board games. Current AI agents cannot consistently defeat average human players, while even the best players make mistakes we think an AI could avoid. In this thesis, we will focus on StarCraft Brood War. We will introduce a formal definition of the model Churchill and Buro proposed for StarCraft. This allows us to focus on Build Order optimization only. We have implemented a base version of the algorithm Churchill and Buro used for their agent. Using the implementation we are able to find solutions for Build Order Problems in StarCraft Brood War.

Auf dem Gebiet der Handlungsplanung stellt die symbolische Suche eine der erfolgversprechendsten angewandten Techniken dar. Um eine symbolische Suche auf endlichen Zustandsräumen zu implementieren bedarf es einer geeigneten Datenstruktur für logische Formeln. Diese Arbeit erprobt die Nutzung von Sentential Decision Diagrams (SDDs) anstelle der gängigen Binary Decision Diagrams (BDDs) zu diesem Zweck. SDDs sind eine Generalisierung von BDDs. Es wird empirisch getestet wie eine Implementierung der symbolischen Suche mit SDDs im FastDownward-Planer sich mit verschiedenen vtrees unterscheidet. Insbesondere wird die Performance von balancierten vtrees, mit welchen die Stärken von SDDs oft gut zur Geltung kommen, mit rechtsseitig linearen vtrees verglichen, bei welchen sich SDDs wie BDDs verhalten.

Die Frage ob es gültige Sudokus - d.h. Sudokus mit nur einer Lösung - gibt, die nur 16 Vorgaben haben, konnte im Dezember 2011 mithilfe einer erschöpfenden Brute-Force-Methode von McGuire et al. verneint werden. Die Schwierigkeit dieser Aufgabe liegt in dem ausufernden Suchraum des Problems und der dadurch entstehenden Erforderlichkeit einer effizienten Beweisidee sowie schnellerer Algorithmen. In dieser Arbeit wird die Beweismethode von McGuire et al. bestätigt werden und für 2 2 × 2 2 und 3 2 × 3 2 Sudokus in C++ implementiert.

Das Finden eines kürzesten Pfades zwischen zwei Punkten ist ein fundamentales Problem in der Graphentheorie. In der Praxis ist es oft wichtig, den Ressourcenverbrauch für das Ermitteln eines solchen Pfades minimal zu halten, was mithilfe einer komprimierten Pfaddatenbank erreicht werden kann. Im Rahmen dieser Arbeit bestimmen wir drei Verfahren, mit denen eine Pfaddatenbank möglichst platzsparend aufgestellt werden kann, und evaluieren die Effektivität dieser Verfahren anhand von Probleminstanzen verschiedener Grösse und Komplexität.

In planning what we want to do is to get from an initial state into a goal state. A state can be described by a finite number of boolean valued variables. If we want to transition from one state to the other we have to apply an action and this, at least in probabilistic planning, leads to a probability distribution over a set of possible successor states. From each transition the agent gains a reward dependent on the current state and his action. In this setting the growth of the number of possible states is exponential with the number of variables. We assume that the value of these variables is determined for each variable independently in a probabilistic fashion. So these variables influence the number of possible successor states in the same way as they did the state space. In consequence it is almost impossible to obtain an optimal amount of reward approaching this problem with a brute force technique. One way to get past this problem is to abstract the problem and then solve a simplified version of the aforementioned. That’s in general the idea proposed by Boutilier and Dearden [1]. They have introduced a method to create an abstraction which depends on the reward formula and the dependencies contained in the problem. With this idea as a basis we’ll create a heuristic for a trial-based heuristic tree search (THTS) algorithm [5] and a standalone planner using the framework PROST (Keller and Eyerich, 2012). These will then be tested on all the domains of the International Probabilistic Planning Competition (IPPC).

In einer Planungsaufgabe geht es darum einen gegebenen Wertezustand durch sequentielles Anwenden von Aktionen in einen Wertezustand zu überführen, welcher geforderte Zieleigenschaften erfüllt. Beim Lösen von Planungsaufgaben zählt Effizienz. Um Zeit und Speicher zu sparen verwenden viele Planer heuristische Suche. Dabei wird mittels einer Heuristik abgeschätzt, welche Aktion als nächstes angewendet werden soll um möglichst schnell in einen gewünschten Zustand zu gelangen.

In dieser Arbeit geht es darum, die von Haslum vorgeschlagene P m -Kompilierung für Planungsaufgaben zu implementieren und die h max -Heuristik auf dem kompilierten Problem gegen die h m -Heuristik auf dem originalen Problem zu testen. Die Implementation geschieht als Ergänzung zum Fast-Downward-Planungssystem. Die Resultate der Tests zeigen, dass mittels der Kompilierung die Zahl der gelösten Probleme erhöht werden kann. Das Lösen eines kompilierten Problems mit der h max -Heuristik geschieht im allgemeinen mit selbiger Informationstiefe schneller als das Lösen des originalen Problems mit der h m -Heuristik. Diesen Zeitgewinn erkauft man sich mit einem höheren Speicherbedarf.

The objective of classical planning is to find a sequence of actions which begins in a given initial state and ends in a state that satisfies a given goal condition. A popular approach to solve classical planning problems is based on heuristic forward search algorithms. In contrast, regression search algorithms apply actions “backwards” in order to find a plan from a goal state to the initial state. Currently, regression search algorithms are somewhat unpopular, as the generation of partial states in a basic regression search often leads to a significant growth of the explored search space. To tackle this problem, state subsumption is a pruning technique that additionally discards newly generated partial states for which a more general partial state has already been explored.

In this thesis, we discuss and evaluate techniques of regression and state subsumption. In order to evaluate their performance, we have implemented a regression search algorithm for the planning system Fast Downward, supporting both a simple subsumption technique as well as a refined subsumption technique using a trie data structure. The experiments have shown that a basic regression search algorithm generally increases the number of explored states compared to uniform-cost forward search. Regression with pruning based on state subsumption with a trie data structure significantly reduces the number of explored states compared to basic regression.

This thesis discusses the Traveling Tournament Problem and how it can be solved with heuristic search. The Traveling Tournament problem is a sports scheduling problem where one tries to find a schedule for a league that meets certain constraints while minimizing the overall distance traveled by the teams in this league. It is hard to solve for leagues with many teams involved since its complexity grows exponentially in the number of teams. The largest instances solved up to date, are instances with leagues of up to 10 teams.

Previous related work has shown that it is a reasonable approach to solve the Traveling Tournament Problem with an IDA*-based tree search. In this thesis I implemented such a search and extended it with several enhancements to examine whether they improve performance of the search. The heuristic that I used in my implementation is the Independent Lower Bound heuristic. It tries to find lower bounds to the traveling costs of each team in the considered league. With my implementation I was able to solve problem instances with up to 8 teams. The results of my evaluation have mostly been consistent with the expected impact of the implemented enhancements on the overall performance.

One huge topic in Artificial Intelligence is the classical planning. It is the process of finding a plan, therefore a sequence of actions that leads from an initial state to a goal state for a specified problem. In problems with a huge amount of states it is very difficult and time consuming to find a plan. There are different pruning methods that attempt to lower the amount of time needed to find a plan by trying to reduce the number of states to explore. In this work we take a closer look at two of these pruning methods. Both of these methods rely on the last action that led to the current state. The first one is the so called tunnel pruning that is a generalisation of the tunnel macros that are used to solve Sokoban problems. The idea is to find actions that allow a tunnel and then prune all actions that are not in the tunnel of this action. The second method is the partition-based path pruning. In this method all actions are distributed into different partitions. These partitions then can be used to prune actions that do not belong to the current partition.

The evaluation of these two pruning methods show, that they can reduce the number of explored states for some problem domains, however the difference between pruned search and normal search gets smaller when we use heuristic functions. It also shows that the two pruning rules effect different problem domains.

Ziel klassischer Handlungsplanung ist es auf eine möglichst effiziente Weise gegebene Planungsprobleme zu lösen. Die Lösung bzw. der Plan eines Planungsproblems ist eine Sequenz von Operatoren mit denen man von einem Anfangszustand in einen Zielzustand gelangt. Um einen Zielzustand gezielter zu finden, verwenden einige Suchalgorithmen eine zusätzliche Information über den Zustandsraum - die Heuristik. Sie schätzt, ausgehend von einem Zustand den Abstand zum Zielzustand. Demnach wäre es ideal, wenn jeder neue besuchte Zustand einen kleineren heuristischen Wert aufweisen würde als der bisher besuchte Zustand. Es gibt allerdings Suchszenarien bei denen die Heuristik nicht weiterhilft um einem Ziel näher zu kommen. Dies ist insbesondere dann der Fall, wenn sich der heuristische Wert von benachbarten Zuständen nicht ändert. Für die gierige Bestensuche würde das bedeuten, dass die Suche auf Plateaus und somit blind verläuft, weil sich dieser Suchalgorithmus ausschliesslich auf die Heuristik stützt. Algorithmen, die die Heuristik als Wegweiser verwenden, gehören zur Klasse der heuristischen Suchalgorithmen.

In dieser Arbeit geht es darum, in Fällen wie den Plateaus trotzdem eine Orientierung im Zustandsraum zu haben, indem Zustände neben der Heuristik einer weiteren Priorisierung unterliegen. Die hier vorgestellte Methode nutzt Abhängigkeiten zwischen Operatoren aus und erweitert die gierige Bestensuche. Wie stark Operatoren voneinander abhängen, betrachten wir anhand eines Abstandsmasses, welches vor der eigentlichen Suche berechnet wird. Die grundlegende Idee ist, Zustände zu bevorzugen, deren Operatoren im Vorfeld voneinander profitierten. Die Heuristik fungiert hierbei erst im Nachhinein als Tie-Breaker, sodass wir einem vielversprechenden Pfad zunächst folgen können, ohne dass uns die Heuristik an einer anderen, weniger vielversprechenden Stelle suchen lässt.

Die Ergebnisse zeigen, dass unser Ansatz in der reinen Suchzeit je nach Heuristik performanter sein kann, als wenn man sich ausschliesslich auf die Heuristik stützt. Bei sehr informationsreichen Heuristiken kann es jedoch passieren, dass die Suche durch unseren Ansatz eher gestört wird. Zudem werden viele Probleme nicht gelöst, weil die Berechnung der Abstände zu zeitaufwändig ist.

In classical planning, heuristic search is a popular approach to solving problems very efficiently. The objective of planning is to find a sequence of actions that can be applied to a given problem and that leads to a goal state. For this purpose, there are many heuristics. They are often a big help if a problem has a solution, but what happens if a problem does not have one? Which heuristics can help proving unsolvability without exploring the whole state space? How efficient are they? Admissible heuristics can be used for this purpose because they never overestimate the distance to a goal state and are therefore able to safely cut off parts of the search space. This makes it potentially easier to prove unsolvability

In this project we developed a problem generator to automatically create unsolvable problem instances and used those generated instances to see how different admissible heuristics perform on them. We used the Japanese puzzle game Sokoban as the first problem because it has a high complexity but is still easy to understand and to imagine for humans. As second problem, we used a logistical problem called NoMystery because unlike Sokoban it is a resource constrained problem and therefore a good supplement to our experiments. Furthermore, unsolvability occurs rather 'naturally' in these two domains and does not seem forced.

Sokoban is a computer game where each level consists of a two-dimensional grid of fields. There are walls as obstacles, moveable boxes and goal fields. The player controls the warehouse worker (Sokoban in Japanese) to push the boxes to the goal fields. The problem is very complex and that is why Sokoban has become a domain in planning.

Phase transitions mark a sudden change in solvability when traversing through the problem space. They occur in the region of hard instances and have been found for many domains. In this thesis we investigate phase transitions in the Sokoban puzzle. For our investigation we generate and evaluate random instances. We declare the defining parameters for Sokoban and measure their influence on the solvability. We show that phase transitions in the solvability of Sokoban can be found and their occurrence is measured. We attempt to unify the parameters of Sokoban to get a prediction on the solvability and hardness of specific instances.

In planning, we address the problem of automatically finding a sequence of actions that leads from a given initial state to a state that satisfies some goal condition. In satisficing planning, our objective is to find plans with preferably low, but not necessarily the lowest possible costs while keeping in mind our limited resources like time or memory. A prominent approach for satisficing planning is based on heuristic search with inadmissible heuristics. However, depending on the applied heuristic, plans found with heuristic search might be of low quality, and hence, improving the quality of such plans is often desirable. In this thesis, we adapt and apply iterative tunneling search with A* (ITSA*) to planning. ITSA* is an algorithm for plan improvement which has been originally proposed by Furcy et al. for search problems. ITSA* intends to search the local space of a given solution path in order to find "short cuts" which allow us to improve our solution. In this thesis, we provide an implementation and systematic evaluation of this algorithm on the standard IPC benchmarks. Our results show that ITSA* also successfully works in the planning area.

In action planning, greedy best-first search (GBFS) is one of the standard techniques if suboptimal plans are accepted. GBFS uses a heuristic function to guide the search towards a goal state. To achieve generality, in domain-independant planning the heuristic function is generated automatically. A well-known problem of GBFS are search plateaus, i.e., regions in the search space where all states have equal heuristic values. In such regions, heuristic search can degenerate to uninformed search. Hence, techniques to escape from such plateaus are desired to improve the efficiency of the search. A recent approach to avoid plateaus is based on diverse best-first search (DBFS) proposed by Imai and Kishimoto. However, this approach relies on several parameters. This thesis presents an implementation of DBFS into the Fast Downward planner. Furthermore, this thesis presents a systematic evaluation of DBFS for several parameter settings, leading to a better understanding of the impact of the parameter choices to the search performance.

Risk is a popular board game where players conquer each other's countries. In this project, I created an AI that plays Risk and is capable of learning. For each decision it makes, it performs a simple search one step ahead, looking at the outcomes of all possible moves it could make, and picks the most beneficial. It judges the desirability of outcomes by a series of parameters, which are modified after each game using the TD(λ)-Algorithm, allowing the AI to learn.

The Canadian Traveler's Problem ( ctp ) is a path finding problem where due to unfavorable weather, some of the roads are impassable. At the beginning, the agent does not know which roads are traversable and which are not. Instead, it can observe the status of roads adjacent to its current location. We consider the stochastic variant of the problem, where the blocking status of a connection is randomly defined with known probabilities. The goal is to find a policy which minimizes the expected travel costs of the agent.

We discuss several properties of the stochastic ctp and present an efficient way to calculate state probabilities. With the aid of these theoretical results, we introduce an uninformed algorithm to find optimal policies.

Finding optimal solutions for general search problems is a challenging task. A powerful approach for solving such problems is based on heuristic search with pattern database heuristics. In this thesis, we present a domain specific solver for the TopSpin Puzzle problem. This solver is based on the above-mentioned pattern database approach. We investigate several pattern databases, and evaluate them on problem instances of different size.

Merge-and-shrink abstractions are a popular approach to generate abstraction heuristics for planning. The computation of merge-and-shrink abstractions relies on a merging and a shrinking strategy. A recently investigated shrinking strategy is based on using bisimulations. Bisimulations are guaranteed to produce perfect heuristics. In this thesis, we investigate an efficient algorithm proposed by Dovier et al. for computing coarsest bisimulations. The algorithm, however, cannot directly be applied to planning and needs some adjustments. We show how this algorithm can be reduced to work with planning problems. In particular, we show how an edge labelled state space can be translated to a state labelled one and what other changes are necessary for the algorithm to be usable for planning problems. This includes a custom data structure to fulfil all requirements to meet the worst case complexity. Furthermore, the implementation will be evaluated on planning problems from the International Planning Competitions. We will see that the resulting algorithm can often not compete with the currently implemented algorithm in Fast Downward. We discuss the reasons why this is the case and propose possible solutions to resolve this issue.

In order to understand an algorithm, it is always helpful to have a visualization that shows step for step what the algorithm is doing. Under this presumption this Bachelor project will explain and visualize two AI techniques, Constraint Satisfaction Processing and SAT Backbones, using the game Gnomine as an example.

CSP techniques build up a network of constraints and infer information by propagating through a single or several constraints at a time, reducing the domain of the variables in the constraint(s). SAT Backbone Computations find literals in a propositional formula, which are true in every model of the given formula.

By showing how to apply these algorithms on the problem of solving a Gnomine game I hope to give a better insight on the nature of how the chosen algorithms work.

Planning as heuristic search is a powerful approach to solve domain-independent planning problems. An important class of heuristics is based on abstractions of the original planning task. However, abstraction heuristics usually come with loss in precision. The contribution of this thesis is the investigation of constrained abstraction heuristics in general, and the application of this concept to pattern database and merge and shrink abstractions in particular. The idea is to use a subclass of mutexes which represent sets of variable-value-pairs so that only one of these pairs can be true at any given time, to regain some of the precision which is lost in the abstraction without increasing its size. By removing states and operators in the abstraction which conflict with such a mutex, the abstraction is refined and hence, the corresponding abstraction heuristic can get more informed. We have implemented the refinements of these heuristics in the Fast Downward planner and evaluated the different approaches using standard IPC benchmarks. The results show that the concept of constrained abstraction heuristics can improve planning as heuristic search in terms of time and coverage.

A permutation problem considers the task where an initial order of objects (ie, an initial mapping of objects to locations) must be reordered into a given goal order by using permutation operators. Permutation operators are 1:1 mappings of the objects from their locations to (possibly other) locations. An example for permutation problems are the wellknown Rubik's Cube and TopSpin Puzzle. Permutation problems have been a research area for a while, and several methods for solving such problems have been proposed in the last two centuries. Most of these methods focused on finding optimal solutions, causing an exponential runtime in the worst case.

In this work, we consider an algorithm for solving permutation problems that has been originally proposed by M. Furst, J. Hopcroft and E. Luks in 1980. This algorithm has been introduced on a theoretical level within a proof for "Testing Membership and Determining the Order of a Group", but has not been implemented and evaluated on practical problems so far. In contrast to the other abovementioned solving algorithms, it only finds suboptimal solutions, but is guaranteed to run in polynomial time. The basic idea is to iteratively reach subgoals, and then to let them fix when we go further to reach the next goals. We have implemented this algorithm and evaluated it on different models, as the Pancake Problem and the TopSpin Puzzle .

Pattern databases (Culberson & Schaeffer, 1998) or PDBs, have been proven very effective in creating admissible Heuristics for single-agent search, such as the A*-algorithm. Haslum et. al proposed, a hill-climbing algorithm can be used to construct the PDBs, using the canonical heuristic. A different approach would be to change action-costs in the pattern-related abstractions, in order to obtain the admissible heuristic. This the so called Cost-Partitioning.

The aim of this project was to implement a cost-partitioning inside the hill-climbing algorithm by Haslum, and compare the results with the standard way which uses the canonical heuristic.

UCT ("upper confidence bounds applied to trees") is a state-of-the-art algorithm for acting under uncertainty, e.g. in probabilistic environments. In the last years it has been very successfully applied in numerous contexts, including two-player board games like Go and Mancala and stochastic single-agent optimization problems such as path planning under uncertainty and probabilistic action planning.

In this project the UCT algorithm was implemented, adapted and evaluated for the classical arcade game "Ms Pac-Man". The thesis introduces Ms Pac-Man and the UCT algorithm, discusses some critical design decisions for developing a strong UCT-based algorithm for playing Ms Pac-Man, and experimentally evaluates the implementation.

• How it works

Useful Links

How much will your dissertation cost?

Have an expert academic write your dissertation paper!

Dissertation Services

Get unlimited topic ideas and a dissertation plan for just £45.00

Order topics and plan

Get 1 free topic in your area of study with aim and justification

Yes I want the free topic

Artificial Intelligence Topics for Dissertations

Published by Carmen Troy at January 6th, 2023 , Revised On August 16, 2023

Introduction

Artificial intelligence (AI) is the process of building machines, robots, and software that are intelligent enough to act like humans. With artificial intelligence, the world will move into a future where machines will be as knowledgeable as humans, and they will be able to work and act as humans.

When completely developed, AI-powered machines will replace a lot of humans in a lot of fields. But would that take away power from the humans? Would it cause humans to suffer as these machines will be intelligent enough to carry out daily tasks and perform routine work? Will AI wreak havoc in the coming days? Well, these are questions that can only be answered after thorough research.

To understand how powerful AI machines will be in the future and what sort of a world we will witness, here are the best AI topics you can choose for your dissertation.

You may also want to start your dissertation by requesting  a brief research proposal  from our writers on any of these topics, which includes an  introduction  to the topic,  research question ,  aim and objectives ,  literature review  along with the proposed  methodology  of research to be conducted.  Let us know  if you need any help in getting started.

Check our  dissertation examples  to get an idea of  how to structure your dissertation .

Review the full list of  dissertation topics for 2022 here.

You may also be interested in technology dissertation topics , computer engineering dissertation topics , networking dissertation topics , and data security dissertation topics .

2022 Artificial Intelligence Topics for Dissertations

Topic 1: artificial intelligence (ai) and supply chain management- an assessment of the present and future role played by ai in supply chain process: a case of ibm corporation in the us.

Research Aim: This research aims to find the present, and future role AI plays in supply chain management. It will analyze how AI affects various components of the supply chain process, such as procurement, distribution, etc. It will use the case study of IBM Corporation, which uses AI in the US to make the supply chain process more efficient and reduce losses. Moreover, through various technological and business frameworks, it will recommend changes in the current AI-based supply chain models to improve their efficiency.

Topic 2: Artificial Intelligence (AI) and Blockchain Technology a Transition Towards Decentralized and Automated Finance- A Study to Find the Role of AI and Blockchains in Making Various Segments of Financial Sector Automated and Decentralized

Research Aim: This study will analyze the role of AI and blockchains in making various segments of financial markets (banking, insurance, investment, stock market, etc.) automated and decentralized. It will find how AI and blockchains can eliminate the part of intimidators and commission charging players such as large banks and corporations to make the economy and financial system more efficient and cheaper. Therefore, it will study applications of various AI and blockchain models to show how they can affect economic governance.

Topic 3: AI and Healthcare- A Comparative Analysis of the Machine Learning (ML) and Deep Learning Models for Cancer Diagnosis

Research Aim: This study aims to identify the role of AI in modern healthcare. It will analyze the efficacy of the contemporary ML and DL models for cancer diagnosis. It will find how these models diagnose cancer, which technology ML or DL does it better, and how much better efficient. Moreover, it will also discuss criticism of these models and ways to improve them for better results.

Topic 4: Are AI and Big Data Analytics New Tools for Digital Innovation? An Assessment of Available Blockchain and Data Analytics Tools for Startups Development

Research Aim: This study aims to assess the role of present AI and data analytics tools for startups development. It will identify how modern startups use these technologies in their development stages to innovate and increase their effectiveness. Moreover, it will analyze its macroeconomic effects by examining its role in speeding up the startup culture, creating more employment, and rising incomes.

Topic 5: The Role of AI and Robotics in Economic Growth and Development- A Case of Emerging Economies

Research Aim: This study aims to find the impact of AI and Robotics on economic growth and development in emerging economies. It will identify how AI and Robotics speed up production and other business-related processes in emerging economies, create more employment, and raise aggregate income levels. Moreover, it will see how it leads to innovation and increasing attention towards learning modern skills such as web development, data analytics, data science, etc. Lastly, it will use two or three emerging countries as a case study to show the analysis.

Artificial Intelligence Topics

Topic 1: machine learning and artificial intelligence in the next generation wearable devices.

Research Aim: This study will aim to understand the role of machine learning and big data in the future of wearables. The research will focus on how an individual’s health and wellbeing can be improved with devices that are powered by AI. The study will first focus on the concept of ML and its implications in various fields. Then, it will be narrowed down to the role of machine learning in the future of wearable devices and how it can help individuals improve their daily routine and lifestyle and move towards a better and healthier life. The research will then conclude how ML will play its role in the future of wearables and help people improve their well-being.

Topic 2: Automation, machine learning and artificial intelligence in the field of medicine

Research Aim: Machine learning and artificial intelligence play a huge role in the field of medicine. From diagnosis to treatment, artificial intelligence is playing a crucial role in the healthcare industry today. This study will highlight how machine learning and automation can help doctors provide the right treatment to patients at the right time. With AI-powered machines, advanced diagnostic tests are being introduced to track diseases much before their occurrence. Moreover, AI is also helping in developing drugs at a faster pace and personalised treatment. All these aspects will be discussed in this study with relevant case studies.

Topic 3: Robotics and artificial intelligence – Assessing the Impact on business and economics

Research Aim: Businesses are changing the way they work due to technological advancements. Robotics and artificial intelligence have paved the way for new technologies and new methods of working. Many people argue that the introduction of robotics and AI will adversely impact humans as most of them might be replaced by AI-powered machines. While this cannot be denied, this artificial intelligence research topic will aim to understand how much the business will be impacted by these new technologies and assess the future of robotics and artificial intelligence in different businesses.

Topic 4: Artificial intelligence governance: Ethical, legal and social challenges

Research Aim: With artificial intelligence taking over the world, many people have reservations over the technology tracking people and their activities 24/7. They have called for strict governance for these intelligent systems and demanded that this technology be fair and transparent. This research will address these issues and present the ethical, legal, and social challenges governing AI-powered systems. The study will be qualitative in nature and will talk about the various ways through which artificial intelligence systems can be governed. It will also address the challenges that will hinder fair and transparent governance.

Topic 5: Will quantum computing improve artificial intelligence? An analysis

Research Aim: Quantum computing (QC) is set to revolutionize the field of artificial intelligence. According to experts, quantum computing combined with artificial intelligence will change medicine, business, and the economy. This research will first introduce the concept of quantum computing and will explain how powerful it is. The study will then talk about how quantum computing will change and help increase the efficiency of artificially intelligent systems. Examples of algorithms that quantum computing utilises will also be presented to help explain how this field of computer science will help improve artificial intelligence.

Topic 6: The role of deep learning in building intelligent systems

Research Aim: Deep learning, an essential branch of artificial intelligence, utilizes neural networks to assess the various factors similar to a human neural system. This research will introduce the concept of deep learning and discuss how it works in artificial intelligence. Deep learning algorithms will also be explored in this study to have a deeper understanding of this artificial intelligence topic. Using case examples and evidence, the research will explore how deep learning assists in creating machines that are intelligent and how they can process information like a human being. The various applications of deep learning will also be discussed in this study.

Topic 7: Evaluating the role of natural language processing in artificial intelligence

Research Aim: Natural language processing (NLP) is an essential element of artificial intelligence. It provides systems and machines with the ability to read, understand and interpret the human language. With the help of natural language processing, systems can even measure sentiments and predict which parts of human language are important. This research will aim to evaluate the role of this language in the field of artificial intelligence. It will further assist in understanding how natural language processing helps build intelligent systems that various organizations can use. Furthermore, the various applications of NLP will also be discussed.

Topic 8: Application of computer vision in building intelligent systems

Research Aim: Computer vision in the field of artificial intelligence makes systems so smart that they can analyze and understand images and pictures. These machines then derive some intelligence from the image that has been fed to the system. This research will first aim to understand computer vision and its role in artificial intelligence. A framework will be presented that will explain the working of computer vision in artificial intelligence. This study will present the applications of computer vision to clarify further how artificial intelligence uses computer vision to build smart systems.

Topic 9: Analysing the use of the IoT in artificial intelligence

Research Aim: The Internet of things and artificial intelligence are two separate, powerful tools. IoT can connect devices wirelessly, which can perform a set of actions without human intervention. When this powerful tool is combined with artificial intelligence, systems become extremely powerful to simulate human behaviour and make decisions without human interference. This artificial intelligence topic will aim to analyze the use of the internet of things in artificial intelligence. Machines that use IoT and AI will be analyzed, and the study will present how human behaviour is simulated so accurately.

Topic 10: Recommender systems – exploring its power in e-commerce

Research Aim: Recommender systems use algorithms to offer relevant suggestions to users. Be it a product, a service, a search result, or a movie/TV show/series. Users receive tons of recommendations after searching for a particular product or browsing their favourite TV shows list. With the help of AI, recommender systems can offer relevant and accurate suggestions to users. The main aim of this research will be to explore the use of recommender systems in e-commerce. Industry giants use this tool to help customers find the product or service they are looking for and make the right decision. This research will discuss where recommender systems are used, how they are implemented, and their results for e-commerce businesses.

Free Dissertation Topic

Phone Number

Academic Level Select Academic Level Undergraduate Graduate PHD

Academic Subject

Area of Research

Frequently Asked Questions

How to find artificial intelligence dissertation topics.

To find artificial intelligence dissertation topics:

• Study recent AI advancements.
• Explore ethical concerns.
• Investigate AI in specific industries.
• Analyze AI’s societal impact.
• Consider human-AI interaction.
• Select a topic aligning with your expertise and passion.

You May Also Like

Need interesting and manageable Mental Health dissertation topics or thesis? Here are the trending Mental Health dissertation titles so you can choose the most suitable one.

Find the most unique and interesting dissertation topic ideas for translation studies to help you in your translation dissertation/ thesis.

Waste disposal is an important part of our everyday lives that often goes unnoticed. Proper waste disposal ensures that our environment and public health remain safe and healthy.

USEFUL LINKS

LEARNING RESOURCES

COMPANY DETAILS

• How It Works

The Future of AI Research: 20 Thesis Ideas for Undergraduate Students in Machine Learning and Deep Learning for 2023!

A comprehensive guide for crafting an original and innovative thesis in the field of ai..

By Aarafat Islam on 2023-01-11

“The beauty of machine learning is that it can be applied to any problem you want to solve, as long as you can provide the computer with enough examples.” — Andrew Ng

This article provides a list of 20 potential thesis ideas for an undergraduate program in machine learning and deep learning in 2023. Each thesis idea includes an  introduction , which presents a brief overview of the topic and the  research objectives . The ideas provided are related to different areas of machine learning and deep learning, such as computer vision, natural language processing, robotics, finance, drug discovery, and more. The article also includes explanations, examples, and conclusions for each thesis idea, which can help guide the research and provide a clear understanding of the potential contributions and outcomes of the proposed research. The article also emphasized the importance of originality and the need for proper citation in order to avoid plagiarism.

1. Investigating the use of Generative Adversarial Networks (GANs) in medical imaging:  A deep learning approach to improve the accuracy of medical diagnoses.

Introduction:  Medical imaging is an important tool in the diagnosis and treatment of various medical conditions. However, accurately interpreting medical images can be challenging, especially for less experienced doctors. This thesis aims to explore the use of GANs in medical imaging, in order to improve the accuracy of medical diagnoses.

2. Exploring the use of deep learning in natural language generation (NLG): An analysis of the current state-of-the-art and future potential.

Introduction:  Natural language generation is an important field in natural language processing (NLP) that deals with creating human-like text automatically. Deep learning has shown promising results in NLP tasks such as machine translation, sentiment analysis, and question-answering. This thesis aims to explore the use of deep learning in NLG and analyze the current state-of-the-art models, as well as potential future developments.

3. Development and evaluation of deep reinforcement learning (RL) for robotic navigation and control.

Introduction:  Robotic navigation and control are challenging tasks, which require a high degree of intelligence and adaptability. Deep RL has shown promising results in various robotics tasks, such as robotic arm control, autonomous navigation, and manipulation. This thesis aims to develop and evaluate a deep RL-based approach for robotic navigation and control and evaluate its performance in various environments and tasks.

4. Investigating the use of deep learning for drug discovery and development.

Introduction:  Drug discovery and development is a time-consuming and expensive process, which often involves high failure rates. Deep learning has been used to improve various tasks in bioinformatics and biotechnology, such as protein structure prediction and gene expression analysis. This thesis aims to investigate the use of deep learning for drug discovery and development and examine its potential to improve the efficiency and accuracy of the drug development process.

5. Comparison of deep learning and traditional machine learning methods for anomaly detection in time series data.

Introduction:  Anomaly detection in time series data is a challenging task, which is important in various fields such as finance, healthcare, and manufacturing. Deep learning methods have been used to improve anomaly detection in time series data, while traditional machine learning methods have been widely used as well. This thesis aims to compare deep learning and traditional machine learning methods for anomaly detection in time series data and examine their respective strengths and weaknesses.

Photo by  Joanna Kosinska  on  Unsplash

6. Use of deep transfer learning in speech recognition and synthesis.

Introduction:  Speech recognition and synthesis are areas of natural language processing that focus on converting spoken language to text and vice versa. Transfer learning has been widely used in deep learning-based speech recognition and synthesis systems to improve their performance by reusing the features learned from other tasks. This thesis aims to investigate the use of transfer learning in speech recognition and synthesis and how it improves the performance of the system in comparison to traditional methods.

7. The use of deep learning for financial prediction.

Introduction:  Financial prediction is a challenging task that requires a high degree of intelligence and adaptability, especially in the field of stock market prediction. Deep learning has shown promising results in various financial prediction tasks, such as stock price prediction and credit risk analysis. This thesis aims to investigate the use of deep learning for financial prediction and examine its potential to improve the accuracy of financial forecasting.

8. Investigating the use of deep learning for computer vision in agriculture.

Introduction:  Computer vision has the potential to revolutionize the field of agriculture by improving crop monitoring, precision farming, and yield prediction. Deep learning has been used to improve various computer vision tasks, such as object detection, semantic segmentation, and image classification. This thesis aims to investigate the use of deep learning for computer vision in agriculture and examine its potential to improve the efficiency and accuracy of crop monitoring and precision farming.

9. Development and evaluation of deep learning models for generative design in engineering and architecture.

Introduction:  Generative design is a powerful tool in engineering and architecture that can help optimize designs and reduce human error. Deep learning has been used to improve various generative design tasks, such as design optimization and form generation. This thesis aims to develop and evaluate deep learning models for generative design in engineering and architecture and examine their potential to improve the efficiency and accuracy of the design process.

10. Investigating the use of deep learning for natural language understanding.

Introduction:  Natural language understanding is a complex task of natural language processing that involves extracting meaning from text. Deep learning has been used to improve various NLP tasks, such as machine translation, sentiment analysis, and question-answering. This thesis aims to investigate the use of deep learning for natural language understanding and examine its potential to improve the efficiency and accuracy of natural language understanding systems.

Photo by  UX Indonesia  on  Unsplash

11. Comparing deep learning and traditional machine learning methods for image compression.

Introduction:  Image compression is an important task in image processing and computer vision. It enables faster data transmission and storage of image files. Deep learning methods have been used to improve image compression, while traditional machine learning methods have been widely used as well. This thesis aims to compare deep learning and traditional machine learning methods for image compression and examine their respective strengths and weaknesses.

12. Using deep learning for sentiment analysis in social media.

Introduction:  Sentiment analysis in social media is an important task that can help businesses and organizations understand their customers’ opinions and feedback. Deep learning has been used to improve sentiment analysis in social media, by training models on large datasets of social media text. This thesis aims to use deep learning for sentiment analysis in social media, and evaluate its performance against traditional machine learning methods.

13. Investigating the use of deep learning for image generation.

Introduction:  Image generation is a task in computer vision that involves creating new images from scratch or modifying existing images. Deep learning has been used to improve various image generation tasks, such as super-resolution, style transfer, and face generation. This thesis aims to investigate the use of deep learning for image generation and examine its potential to improve the quality and diversity of generated images.

14. Development and evaluation of deep learning models for anomaly detection in cybersecurity.

Introduction:  Anomaly detection in cybersecurity is an important task that can help detect and prevent cyber-attacks. Deep learning has been used to improve various anomaly detection tasks, such as intrusion detection and malware detection. This thesis aims to develop and evaluate deep learning models for anomaly detection in cybersecurity and examine their potential to improve the efficiency and accuracy of cybersecurity systems.

15. Investigating the use of deep learning for natural language summarization.

Introduction:  Natural language summarization is an important task in natural language processing that involves creating a condensed version of a text that preserves its main meaning. Deep learning has been used to improve various natural language summarization tasks, such as document summarization and headline generation. This thesis aims to investigate the use of deep learning for natural language summarization and examine its potential to improve the efficiency and accuracy of natural language summarization systems.

Photo by  Windows  on  Unsplash

16. Development and evaluation of deep learning models for facial expression recognition.

Introduction:  Facial expression recognition is an important task in computer vision and has many practical applications, such as human-computer interaction, emotion recognition, and psychological studies. Deep learning has been used to improve facial expression recognition, by training models on large datasets of images. This thesis aims to develop and evaluate deep learning models for facial expression recognition and examine their performance against traditional machine learning methods.

17. Investigating the use of deep learning for generative models in music and audio.

Introduction:  Music and audio synthesis is an important task in audio processing, which has many practical applications, such as music generation and speech synthesis. Deep learning has been used to improve generative models for music and audio, by training models on large datasets of audio data. This thesis aims to investigate the use of deep learning for generative models in music and audio and examine its potential to improve the quality and diversity of generated audio.

18. Study the comparison of deep learning models with traditional algorithms for anomaly detection in network traffic.

Introduction:  Anomaly detection in network traffic is an important task that can help detect and prevent cyber-attacks. Deep learning models have been used for this task, and traditional methods such as clustering and rule-based systems are widely used as well. This thesis aims to compare deep learning models with traditional algorithms for anomaly detection in network traffic and analyze the trade-offs between the models in terms of accuracy and scalability.

19. Investigating the use of deep learning for improving recommender systems.

Introduction:  Recommender systems are widely used in many applications such as online shopping, music streaming, and movie streaming. Deep learning has been used to improve the performance of recommender systems, by training models on large datasets of user-item interactions. This thesis aims to investigate the use of deep learning for improving recommender systems and compare its performance with traditional content-based and collaborative filtering approaches.

20. Development and evaluation of deep learning models for multi-modal data analysis.

Introduction:  Multi-modal data analysis is the task of analyzing and understanding data from multiple sources such as text, images, and audio. Deep learning has been used to improve multi-modal data analysis, by training models on large datasets of multi-modal data. This thesis aims to develop and evaluate deep learning models for multi-modal data analysis and analyze their potential to improve performance in comparison to single-modal models.

I hope that this article has provided you with a useful guide for your thesis research in machine learning and deep learning. Remember to conduct a thorough literature review and to include proper citations in your work, as well as to be original in your research to avoid plagiarism. I wish you all the best of luck with your thesis and your research endeavors!

Continue Learning

Intelligent interactions are transforming app development using chatgpt, innovating responsibly with ai: an ethical perspective, are we building ai systems that learned to lie to us.

DeepFakes = DeepLearning + Fake

How to Use Llama 2 with an API on AWS to Power Your AI Apps

Role of artificial intelligence in metaverse.

Exploring the Saga of Metaverse with AI

Beginner’s Guide to OpenAI’s GPT-3.5-Turbo Model

From GPT-3 to GPT-3.5-Turbo: Understanding the Latest Upgrades in OpenAI’s Language Model API.

Topics for Master Theses at the Chair for Artificial Intelligence

Smart city / smart mobility.

• Traffic Forecasting with Graph Attention Networks
• Learning Traffic Simulation Parameters with Reinforcement Learning
• Extending the Mannheim Mobility Model with Individual Bike Traffic

AI for Business Process Management

• Applications of deep neural networks in Online Conformance Checking
• Accurate Business Process Simulation (BPS) models based on deep learning
• How to tackle concept drift in Predictive Process Monitoring (PPM)

Explainable and Fair Machine Learning

• Extracting Causal Models from Module Handbooks for Explainable Student Success Prediction
• Investigating Different Techniques to Improve Fairness for Tabular Data
• Data-induced Bias in Social Simulations
• Learing Causal Models from Tabular Data

Human Activity and Goal Recognition

• Reinforcement Learning for Goal Recognition
• Investigating the Difficulty of Goal Recognition Problems
• Enhancing Audio-Based Activity Recognition through Autoencoder Encoded Representations
• Activity Recognition from Audio Data in a Kitchen Scenario
• Speaker Diarization and Identification in a Meeting Scenario

Machine Learning for Supply Chain Optimization

• Time Series Analysis & Forecasting of Events (Sales, Demand, etc.)
• Integrated vs. separated optimization: theory and practice
• Leveraging deep learning to build a versatile end-to-end inventory management model
• Reinforcement learning for the vehicle routing problem
• Metaheuristics in SCM: Overview and benchmark study
• Finetuning parametrized inventory management system

Anomaly Detection on Server Logs

• Analyse real-life server logs stored in an existing opensearch library (Graylog)
• Learning values describing normal behavior of servers and detect anomalies in logged messages
• Implement simple alert system (existing systems like Icinga can be used)
• Prepare results in a (Web-)Gui
• Creating eLearning Recommender Systems using NLP
• Hyperparameter Optimization for Symbolic Knowledge Graph Completion
• Applying Symbolic Knowledge Graph Completion to Inductive Link Prediction
• Data Augmentation via Generative Adversarial Networks (GANs)
• Autoencoders for Sparse, Irregularly Spaced Time Series Sequences

Tracking cookies are currently allowed.

Tracking cookies are currently not allowed.

The Main Topics for Coursework or a Thesis Statement in Artificial Intelligence

Artificial Intelligence (AI) is changing the world, from machine learning and the Internet of Things to Robotics and Natural Language processing.

Research is needed to understand more about AI and how it will affect the future. 

AI-powered machines are likely to replace humans in many fields and the consequences of this are still largely unknown.

There are many topics of vital importance to choose from if you’re a student trying to decide on a topic involving AI for your thesis.

Image source:  Freepik.com

Machine learning (ML) as a Thesis Topic

Artificial intelligence enables machines to automatically learn a task from experience and improve performance without any human intervention.

Machines need high-quality data to start with. They are trained by building machine learning models using the data and different algorithms.

The algorithms depend on the type of data and the tasks that need automation. 

A topic for your research could involve discussing wearable devices. They are powered by machine learning and are becoming increasingly popular.

You could discuss their relevance in fields like health and insurance as well as how they can help individuals to improve their daily routines and move towards a more healthy lifestyle.  

Deep learning (DL) as a Thesis Topic

Deep Learning is a subset of ML where learning imitates the inner workings of the human brain. It uses artificial neural networks to process data and make decisions.

The web-like networks take a non-linear approach to processing data which is superior to traditional algorithms that take a linear approach.  

Google’s RankBrain is an example of an artificial neural network.

Deep learning is driving many AI applications such as object recognition, playing computer games, controlling self-driving cars and language translation.

A research topic could involve discussing deep learning and its various applications. 

Reinforcement learning (RL) as a Thesis Topic

Reinforcement learning is the closest form of learning to the way human beings learn. For instance, students learn from their mistakes and a process of trial-and-error.

There are many different ways to use AI in education to help students, such as using AI-powered tutors, customized learning and smart content.

RL works on a similar principle to learning from a process of trial-and-error. Google’s AlphaGo program beat the world champion of Go in 2017 by using RL. 

Students who don’t yet have the skills to handle complex assignments can make use of various tools, writing apps and professional writers.

To find help with your student papers when you’re conducting research for a university, EduBirdie has free plagiarism checker and citations tools but professional writers who can take the pressure off you.

At U.K. EduBirdie , a professional  thesis writer will finish your paper  for you. It also offers editing and proofreading services at very reasonable prices.

Image source: Freepik.com

Natural language processing (NLP) as a Thesis Topic

This area of AI relates to how machines can learn to recognize and analyze human speech. Speech recognition, natural language translation and natural language generation are some of the areas of NLP.

With the help of NLP, systems can even read sentiment and predict which parts of the language are important. Revolutionary tools like IBM Watson, Google Translate, Speech Recognition and sentiment analysis show the importance of NLP in the daily lives of individuals. 

NLP helps build intelligent systems, such as customer support applications like chatbots and  AI in education  is also a great example.

Chatbots use NLP and machine learning to interact with customers and solve their queries. Your research topic could relate to chatbots and their interaction with humans.

Computer vision (CV) as a Thesis Topic

Millions of images are uploaded daily on the internet. Computers are very good at certain tasks but they can struggle with simple tasks like being able to recognize and identify objects.

Computer vision is a field of AI that makes systems so smart that they can analyze and understand images. CV systems can even outperform humans now in some tasks like classifying visual objects.  

One of the applications of computer vision is in autonomous vehicles that need to analyze images of surroundings in order to navigate.

A study topic could involve discussing computer vision and how using it allows smart systems to be built. Applications of computer vision could then be presented.  

Recommender systems (RS) as a Thesis Topic

Recommender systems  use algorithms  to offer relevant suggestions to users. These may be suggestions on a TV show, a product, a service or even who to date.

You will receive many recommendations after you search for a particular product or browse a list of favorite movies. RS can base suggestions on your past behavior and past preferences, trends and the preferences of your peers. 

A very relevant topic would be to explore the use of recommender systems in the field of e-commerce. Industry giants like Amazon are currently using recommender systems to help customers find the right products or services.

You could discuss their implementation and the type of results they bring to ecommerce businesses. 

Robotics as a Thesis Topic

Robots can behave and perform the same actions as human beings, thanks to AI. They can act intelligently and even solve problems and learn in controlled environments.

For example, Kismet is a social interaction robot developed by MIT’s AI lab that can recognize human language and interact with humans. 

Robots and AI are changing the way businesses work. Some people argue that this will have an adverse effect on humans as they are replaced by AI-powered machines.

A research topic could aim to understand to what extent businesses will be impacted by  AI-powered machines  and assess their future in different businesses.

There is an increase in the number of research papers being published in different areas of AI. If you’re a student wanting to come up with a topic involving artificial intelligence for your thesis, there are many vitally important sub-topics to choose from.

Each of these sub-topics provides plenty of opportunities for meaningful research into AI and new ideas on its application in the future as machines keep growing in intelligence. 

About The Author

Paul Calderon

Paul Calderon is data security specialist working with a tech startup based in Silicon Valley. After work hours, he helps students studying for their computer science degrees or programming courses with essays, dissertations and term papers. When he isn’t doing any work, he likes playing tennis, cycling, and creating vlogs on local travel.

Leave a Reply Cancel Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed .

The present and future of AI

Finale doshi-velez on how ai is shaping our lives and how we can shape ai.

Finale Doshi-Velez, the John L. Loeb Professor of Engineering and Applied Sciences. (Photo courtesy of Eliza Grinnell/Harvard SEAS)

How has artificial intelligence changed and shaped our world over the last five years? How will AI continue to impact our lives in the coming years? Those were the questions addressed in the most recent report from the One Hundred Year Study on Artificial Intelligence (AI100), an ongoing project hosted at Stanford University, that will study the status of AI technology and its impacts on the world over the next 100 years.

The 2021 report is the second in a series that will be released every five years until 2116. Titled “Gathering Strength, Gathering Storms,” the report explores the various ways AI is  increasingly touching people’s lives in settings that range from  movie recommendations  and  voice assistants  to  autonomous driving  and  automated medical diagnoses .

Barbara Grosz , the Higgins Research Professor of Natural Sciences at the Harvard John A. Paulson School of Engineering and Applied Sciences (SEAS) is a member of the standing committee overseeing the AI100 project and Finale Doshi-Velez , Gordon McKay Professor of Computer Science, is part of the panel of interdisciplinary researchers who wrote this year’s report. 

We spoke with Doshi-Velez about the report, what it says about the role AI is currently playing in our lives, and how it will change in the future.  

Q: Let's start with a snapshot: What is the current state of AI and its potential?

Doshi-Velez: Some of the biggest changes in the last five years have been how well AIs now perform in large data regimes on specific types of tasks.  We've seen [DeepMind’s] AlphaZero become the best Go player entirely through self-play, and everyday uses of AI such as grammar checks and autocomplete, automatic personal photo organization and search, and speech recognition become commonplace for large numbers of people.  

In terms of potential, I'm most excited about AIs that might augment and assist people.  They can be used to drive insights in drug discovery, help with decision making such as identifying a menu of likely treatment options for patients, and provide basic assistance, such as lane keeping while driving or text-to-speech based on images from a phone for the visually impaired.  In many situations, people and AIs have complementary strengths. I think we're getting closer to unlocking the potential of people and AI teams.

There's a much greater recognition that we should not be waiting for AI tools to become mainstream before making sure they are ethical.

Q: Over the course of 100 years, these reports will tell the story of AI and its evolving role in society. Even though there have only been two reports, what's the story so far?

There's actually a lot of change even in five years.  The first report is fairly rosy.  For example, it mentions how algorithmic risk assessments may mitigate the human biases of judges.  The second has a much more mixed view.  I think this comes from the fact that as AI tools have come into the mainstream — both in higher stakes and everyday settings — we are appropriately much less willing to tolerate flaws, especially discriminatory ones. There's also been questions of information and disinformation control as people get their news, social media, and entertainment via searches and rankings personalized to them. So, there's a much greater recognition that we should not be waiting for AI tools to become mainstream before making sure they are ethical.

Q: What is the responsibility of institutes of higher education in preparing students and the next generation of computer scientists for the future of AI and its impact on society?

First, I'll say that the need to understand the basics of AI and data science starts much earlier than higher education!  Children are being exposed to AIs as soon as they click on videos on YouTube or browse photo albums. They need to understand aspects of AI such as how their actions affect future recommendations.

But for computer science students in college, I think a key thing that future engineers need to realize is when to demand input and how to talk across disciplinary boundaries to get at often difficult-to-quantify notions of safety, equity, fairness, etc.  I'm really excited that Harvard has the Embedded EthiCS program to provide some of this education.  Of course, this is an addition to standard good engineering practices like building robust models, validating them, and so forth, which is all a bit harder with AI.

I think a key thing that future engineers need to realize is when to demand input and how to talk across disciplinary boundaries to get at often difficult-to-quantify notions of safety, equity, fairness, etc. 

Q: Your work focuses on machine learning with applications to healthcare, which is also an area of focus of this report. What is the state of AI in healthcare? 

A lot of AI in healthcare has been on the business end, used for optimizing billing, scheduling surgeries, that sort of thing.  When it comes to AI for better patient care, which is what we usually think about, there are few legal, regulatory, and financial incentives to do so, and many disincentives. Still, there's been slow but steady integration of AI-based tools, often in the form of risk scoring and alert systems.

In the near future, two applications that I'm really excited about are triage in low-resource settings — having AIs do initial reads of pathology slides, for example, if there are not enough pathologists, or get an initial check of whether a mole looks suspicious — and ways in which AIs can help identify promising treatment options for discussion with a clinician team and patient.

Q: Any predictions for the next report?

I'll be keen to see where currently nascent AI regulation initiatives have gotten to. Accountability is such a difficult question in AI,  it's tricky to nurture both innovation and basic protections.  Perhaps the most important innovation will be in approaches for AI accountability.

Topics: AI / Machine Learning , Computer Science

Cutting-edge science delivered direct to your inbox.

Join the Harvard SEAS mailing list.

Scientist Profiles

Finale Doshi-Velez

Herchel Smith Professor of Computer Science

Press Contact

Leah Burrows | 617-496-1351 | [email protected]

Related News

Alumni profile: Jacomo Corbo, Ph.D. '08

Racing into the future of machine learning 

AI / Machine Learning , Computer Science

Ph.D. student Monteiro Paes named Apple Scholar in AI/ML

Monteiro Paes studies fairness and arbitrariness in machine learning models

AI / Machine Learning , Applied Mathematics , Awards , Graduate Student Profile

A new phase for Harvard Quantum Computing Club

SEAS students place second at MIT quantum hackathon

Computer Science , Quantum Engineering , Undergraduate Student Profile

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

• Publications
• Account settings

Preview improvements coming to the PMC website in October 2024. Learn More or Try it out now .

• Advanced Search
• Journal List
• Tzu Chi Med J
• v.32(4); Oct-Dec 2020

The impact of artificial intelligence on human society and bioethics

Michael cheng-tek tai.

Department of Medical Sociology and Social Work, College of Medicine, Chung Shan Medical University, Taichung, Taiwan

Artificial intelligence (AI), known by some as the industrial revolution (IR) 4.0, is going to change not only the way we do things, how we relate to others, but also what we know about ourselves. This article will first examine what AI is, discuss its impact on industrial, social, and economic changes on humankind in the 21 st century, and then propose a set of principles for AI bioethics. The IR1.0, the IR of the 18 th century, impelled a huge social change without directly complicating human relationships. Modern AI, however, has a tremendous impact on how we do things and also the ways we relate to one another. Facing this challenge, new principles of AI bioethics must be considered and developed to provide guidelines for the AI technology to observe so that the world will be benefited by the progress of this new intelligence.

W HAT IS ARTIFICIAL INTELLIGENCE ?

Artificial intelligence (AI) has many different definitions; some see it as the created technology that allows computers and machines to function intelligently. Some see it as the machine that replaces human labor to work for men a more effective and speedier result. Others see it as “a system” with the ability to correctly interpret external data, to learn from such data, and to use those learnings to achieve specific goals and tasks through flexible adaptation [ 1 ].

Despite the different definitions, the common understanding of AI is that it is associated with machines and computers to help humankind solve problems and facilitate working processes. In short, it is an intelligence designed by humans and demonstrated by machines. The term AI is used to describe these functions of human-made tool that emulates the “cognitive” abilities of the natural intelligence of human minds [ 2 ].

Along with the rapid development of cybernetic technology in recent years, AI has been seen almost in all our life circles, and some of that may no longer be regarded as AI because it is so common in daily life that we are much used to it such as optical character recognition or the Siri (speech interpretation and recognition interface) of information searching equipment on computer [ 3 ].

D IFFERENT TYPES OF ARTIFICIAL INTELLIGENCE

From the functions and abilities provided by AI, we can distinguish two different types. The first is weak AI, also known as narrow AI that is designed to perform a narrow task, such as facial recognition or Internet Siri search or self-driving car. Many currently existing systems that claim to use “AI” are likely operating as a weak AI focusing on a narrowly defined specific function. Although this weak AI seems to be helpful to human living, there are still some think weak AI could be dangerous because weak AI could cause disruptions in the electric grid or may damage nuclear power plants when malfunctioned.

The new development of the long-term goal of many researchers is to create strong AI or artificial general intelligence (AGI) which is the speculative intelligence of a machine that has the capacity to understand or learn any intelligent task human being can, thus assisting human to unravel the confronted problem. While narrow AI may outperform humans such as playing chess or solving equations, but its effect is still weak. AGI, however, could outperform humans at nearly every cognitive task.

Strong AI is a different perception of AI that it can be programmed to actually be a human mind, to be intelligent in whatever it is commanded to attempt, even to have perception, beliefs and other cognitive capacities that are normally only ascribed to humans [ 4 ].

In summary, we can see these different functions of AI [ 5 , 6 ]:

• Automation: What makes a system or process to function automatically
• Machine learning and vision: The science of getting a computer to act through deep learning to predict and analyze, and to see through a camera, analog-to-digital conversion and digital signal processing
• Natural language processing: The processing of human language by a computer program, such as spam detection and converting instantly a language to another to help humans communicate
• Robotics: A field of engineering focusing on the design and manufacturing of cyborgs, the so-called machine man. They are used to perform tasks for human's convenience or something too difficult or dangerous for human to perform and can operate without stopping such as in assembly lines
• Self-driving car: Use a combination of computer vision, image recognition amid deep learning to build automated control in a vehicle.

D O HUMAN-BEINGS REALLY NEED ARTIFICIAL INTELLIGENCE ?

Is AI really needed in human society? It depends. If human opts for a faster and effective way to complete their work and to work constantly without taking a break, yes, it is. However if humankind is satisfied with a natural way of living without excessive desires to conquer the order of nature, it is not. History tells us that human is always looking for something faster, easier, more effective, and convenient to finish the task they work on; therefore, the pressure for further development motivates humankind to look for a new and better way of doing things. Humankind as the homo-sapiens discovered that tools could facilitate many hardships for daily livings and through tools they invented, human could complete the work better, faster, smarter and more effectively. The invention to create new things becomes the incentive of human progress. We enjoy a much easier and more leisurely life today all because of the contribution of technology. The human society has been using the tools since the beginning of civilization, and human progress depends on it. The human kind living in the 21 st century did not have to work as hard as their forefathers in previous times because they have new machines to work for them. It is all good and should be all right for these AI but a warning came in early 20 th century as the human-technology kept developing that Aldous Huxley warned in his book Brave New World that human might step into a world in which we are creating a monster or a super human with the development of genetic technology.

Besides, up-to-dated AI is breaking into healthcare industry too by assisting doctors to diagnose, finding the sources of diseases, suggesting various ways of treatment performing surgery and also predicting if the illness is life-threatening [ 7 ]. A recent study by surgeons at the Children's National Medical Center in Washington successfully demonstrated surgery with an autonomous robot. The team supervised the robot to perform soft-tissue surgery, stitch together a pig's bowel, and the robot finished the job better than a human surgeon, the team claimed [ 8 , 9 ]. It demonstrates robotically-assisted surgery can overcome the limitations of pre-existing minimally-invasive surgical procedures and to enhance the capacities of surgeons performing open surgery.

Above all, we see the high-profile examples of AI including autonomous vehicles (such as drones and self-driving cars), medical diagnosis, creating art, playing games (such as Chess or Go), search engines (such as Google search), online assistants (such as Siri), image recognition in photographs, spam filtering, predicting flight delays…etc. All these have made human life much easier and convenient that we are so used to them and take them for granted. AI has become indispensable, although it is not absolutely needed without it our world will be in chaos in many ways today.

T HE IMPACT OF ARTIFICIAL INTELLIGENCE ON HUMAN SOCIETY

Negative impact.

Questions have been asked: With the progressive development of AI, human labor will no longer be needed as everything can be done mechanically. Will humans become lazier and eventually degrade to the stage that we return to our primitive form of being? The process of evolution takes eons to develop, so we will not notice the backsliding of humankind. However how about if the AI becomes so powerful that it can program itself to be in charge and disobey the order given by its master, the humankind?

Let us see the negative impact the AI will have on human society [ 10 , 11 ]:

• A huge social change that disrupts the way we live in the human community will occur. Humankind has to be industrious to make their living, but with the service of AI, we can just program the machine to do a thing for us without even lifting a tool. Human closeness will be gradually diminishing as AI will replace the need for people to meet face to face for idea exchange. AI will stand in between people as the personal gathering will no longer be needed for communication
• Unemployment is the next because many works will be replaced by machinery. Today, many automobile assembly lines have been filled with machineries and robots, forcing traditional workers to lose their jobs. Even in supermarket, the store clerks will not be needed anymore as the digital device can take over human labor
• Wealth inequality will be created as the investors of AI will take up the major share of the earnings. The gap between the rich and the poor will be widened. The so-called “M” shape wealth distribution will be more obvious
• New issues surface not only in a social sense but also in AI itself as the AI being trained and learned how to operate the given task can eventually take off to the stage that human has no control, thus creating un-anticipated problems and consequences. It refers to AI's capacity after being loaded with all needed algorithm may automatically function on its own course ignoring the command given by the human controller
• The human masters who create AI may invent something that is racial bias or egocentrically oriented to harm certain people or things. For instance, the United Nations has voted to limit the spread of nucleus power in fear of its indiscriminative use to destroying humankind or targeting on certain races or region to achieve the goal of domination. AI is possible to target certain race or some programmed objects to accomplish the command of destruction by the programmers, thus creating world disaster.

P OSITIVE IMPACT

There are, however, many positive impacts on humans as well, especially in the field of healthcare. AI gives computers the capacity to learn, reason, and apply logic. Scientists, medical researchers, clinicians, mathematicians, and engineers, when working together, can design an AI that is aimed at medical diagnosis and treatments, thus offering reliable and safe systems of health-care delivery. As health professors and medical researchers endeavor to find new and efficient ways of treating diseases, not only the digital computer can assist in analyzing, robotic systems can also be created to do some delicate medical procedures with precision. Here, we see the contribution of AI to health care [ 7 , 11 ]:

Fast and accurate diagnostics

IBM's Watson computer has been used to diagnose with the fascinating result. Loading the data to the computer will instantly get AI's diagnosis. AI can also provide various ways of treatment for physicians to consider. The procedure is something like this: To load the digital results of physical examination to the computer that will consider all possibilities and automatically diagnose whether or not the patient suffers from some deficiencies and illness and even suggest various kinds of available treatment.

Socially therapeutic robots

Pets are recommended to senior citizens to ease their tension and reduce blood pressure, anxiety, loneliness, and increase social interaction. Now cyborgs have been suggested to accompany those lonely old folks, even to help do some house chores. Therapeutic robots and the socially assistive robot technology help improve the quality of life for seniors and physically challenged [ 12 ].

Reduce errors related to human fatigue

Human error at workforce is inevitable and often costly, the greater the level of fatigue, the higher the risk of errors occurring. Al technology, however, does not suffer from fatigue or emotional distraction. It saves errors and can accomplish the duty faster and more accurately.

Artificial intelligence-based surgical contribution

AI-based surgical procedures have been available for people to choose. Although this AI still needs to be operated by the health professionals, it can complete the work with less damage to the body. The da Vinci surgical system, a robotic technology allowing surgeons to perform minimally invasive procedures, is available in most of the hospitals now. These systems enable a degree of precision and accuracy far greater than the procedures done manually. The less invasive the surgery, the less trauma it will occur and less blood loss, less anxiety of the patients.

Improved radiology

The first computed tomography scanners were introduced in 1971. The first magnetic resonance imaging (MRI) scan of the human body took place in 1977. By the early 2000s, cardiac MRI, body MRI, and fetal imaging, became routine. The search continues for new algorithms to detect specific diseases as well as to analyze the results of scans [ 9 ]. All those are the contribution of the technology of AI.

Virtual presence

The virtual presence technology can enable a distant diagnosis of the diseases. The patient does not have to leave his/her bed but using a remote presence robot, doctors can check the patients without actually being there. Health professionals can move around and interact almost as effectively as if they were present. This allows specialists to assist patients who are unable to travel.

S OME CAUTIONS TO BE REMINDED

Despite all the positive promises that AI provides, human experts, however, are still essential and necessary to design, program, and operate the AI from any unpredictable error from occurring. Beth Kindig, a San Francisco-based technology analyst with more than a decade of experience in analyzing private and public technology companies, published a free newsletter indicating that although AI has a potential promise for better medical diagnosis, human experts are still needed to avoid the misclassification of unknown diseases because AI is not omnipotent to solve all problems for human kinds. There are times when AI meets an impasse, and to carry on its mission, it may just proceed indiscriminately, ending in creating more problems. Thus vigilant watch of AI's function cannot be neglected. This reminder is known as physician-in-the-loop [ 13 ].

The question of an ethical AI consequently was brought up by Elizabeth Gibney in her article published in Nature to caution any bias and possible societal harm [ 14 ]. The Neural Information processing Systems (NeurIPS) conference in Vancouver Canada in 2020 brought up the ethical controversies of the application of AI technology, such as in predictive policing or facial recognition, that due to bias algorithms can result in hurting the vulnerable population [ 14 ]. For instance, the NeurIPS can be programmed to target certain race or decree as the probable suspect of crime or trouble makers.

T HE CHALLENGE OF ARTIFICIAL INTELLIGENCE TO BIOETHICS

Artificial intelligence ethics must be developed.

Bioethics is a discipline that focuses on the relationship among living beings. Bioethics accentuates the good and the right in biospheres and can be categorized into at least three areas, the bioethics in health settings that is the relationship between physicians and patients, the bioethics in social settings that is the relationship among humankind and the bioethics in environmental settings that is the relationship between man and nature including animal ethics, land ethics, ecological ethics…etc. All these are concerned about relationships within and among natural existences.

As AI arises, human has a new challenge in terms of establishing a relationship toward something that is not natural in its own right. Bioethics normally discusses the relationship within natural existences, either humankind or his environment, that are parts of natural phenomena. But now men have to deal with something that is human-made, artificial and unnatural, namely AI. Human has created many things yet never has human had to think of how to ethically relate to his own creation. AI by itself is without feeling or personality. AI engineers have realized the importance of giving the AI ability to discern so that it will avoid any deviated activities causing unintended harm. From this perspective, we understand that AI can have a negative impact on humans and society; thus, a bioethics of AI becomes important to make sure that AI will not take off on its own by deviating from its originally designated purpose.

Stephen Hawking warned early in 2014 that the development of full AI could spell the end of the human race. He said that once humans develop AI, it may take off on its own and redesign itself at an ever-increasing rate [ 15 ]. Humans, who are limited by slow biological evolution, could not compete and would be superseded. In his book Superintelligence, Nick Bostrom gives an argument that AI will pose a threat to humankind. He argues that sufficiently intelligent AI can exhibit convergent behavior such as acquiring resources or protecting itself from being shut down, and it might harm humanity [ 16 ].

The question is–do we have to think of bioethics for the human's own created product that bears no bio-vitality? Can a machine have a mind, consciousness, and mental state in exactly the same sense that human beings do? Can a machine be sentient and thus deserve certain rights? Can a machine intentionally cause harm? Regulations must be contemplated as a bioethical mandate for AI production.

Studies have shown that AI can reflect the very prejudices humans have tried to overcome. As AI becomes “truly ubiquitous,” it has a tremendous potential to positively impact all manner of life, from industry to employment to health care and even security. Addressing the risks associated with the technology, Janosch Delcker, Politico Europe's AI correspondent, said: “I don't think AI will ever be free of bias, at least not as long as we stick to machine learning as we know it today,”…. “What's crucially important, I believe, is to recognize that those biases exist and that policymakers try to mitigate them” [ 17 ]. The High-Level Expert Group on AI of the European Union presented Ethics Guidelines for Trustworthy AI in 2019 that suggested AI systems must be accountable, explainable, and unbiased. Three emphases are given:

• Lawful-respecting all applicable laws and regulations
• Ethical-respecting ethical principles and values
• Robust-being adaptive, reliable, fair, and trustworthy from a technical perspective while taking into account its social environment [ 18 ].

Seven requirements are recommended [ 18 ]:

• AI should not trample on human autonomy. People should not be manipulated or coerced by AI systems, and humans should be able to intervene or oversee every decision that the software makes
• AI should be secure and accurate. It should not be easily compromised by external attacks, and it should be reasonably reliable
• Personal data collected by AI systems should be secure and private. It should not be accessible to just anyone, and it should not be easily stolen
• Data and algorithms used to create an AI system should be accessible, and the decisions made by the software should be “understood and traced by human beings.” In other words, operators should be able to explain the decisions their AI systems make
• Services provided by AI should be available to all, regardless of age, gender, race, or other characteristics. Similarly, systems should not be biased along these lines
• AI systems should be sustainable (i.e., they should be ecologically responsible) and “enhance positive social change”
• AI systems should be auditable and covered by existing protections for corporate whistleblowers. The negative impacts of systems should be acknowledged and reported in advance.

From these guidelines, we can suggest that future AI must be equipped with human sensibility or “AI humanities.” To accomplish this, AI researchers, manufacturers, and all industries must bear in mind that technology is to serve not to manipulate humans and his society. Bostrom and Judkowsky listed responsibility, transparency, auditability, incorruptibility, and predictability [ 19 ] as criteria for the computerized society to think about.

S UGGESTED PRINCIPLES FOR ARTIFICIAL INTELLIGENCE BIOETHICS

Nathan Strout, a reporter at Space and Intelligence System at Easter University, USA, reported just recently that the intelligence community is developing its own AI ethics. The Pentagon made announced in February 2020 that it is in the process of adopting principles for using AI as the guidelines for the department to follow while developing new AI tools and AI-enabled technologies. Ben Huebner, chief of the Office of Director of National Intelligence's Civil Liberties, Privacy, and Transparency Office, said that “We're going to need to ensure that we have transparency and accountability in these structures as we use them. They have to be secure and resilient” [ 20 ]. Two themes have been suggested for the AI community to think more about: Explainability and interpretability. Explainability is the concept of understanding how the analytic works, while interpretability is being able to understand a particular result produced by an analytic [ 20 ].

All the principles suggested by scholars for AI bioethics are well-brought-up. I gather from different bioethical principles in all the related fields of bioethics to suggest four principles here for consideration to guide the future development of the AI technology. We however must bear in mind that the main attention should still be placed on human because AI after all has been designed and manufactured by human. AI proceeds to its work according to its algorithm. AI itself cannot empathize nor have the ability to discern good from evil and may commit mistakes in processes. All the ethical quality of AI depends on the human designers; therefore, it is an AI bioethics and at the same time, a trans-bioethics that abridge human and material worlds. Here are the principles:

• Beneficence: Beneficence means doing good, and here it refers to the purpose and functions of AI should benefit the whole human life, society and universe. Any AI that will perform any destructive work on bio-universe, including all life forms, must be avoided and forbidden. The AI scientists must understand that reason of developing this technology has no other purpose but to benefit human society as a whole not for any individual personal gain. It should be altruistic, not egocentric in nature
• Value-upholding: This refers to AI's congruence to social values, in other words, universal values that govern the order of the natural world must be observed. AI cannot elevate to the height above social and moral norms and must be bias-free. The scientific and technological developments must be for the enhancement of human well-being that is the chief value AI must hold dearly as it progresses further
• Lucidity: AI must be transparent without hiding any secret agenda. It has to be easily comprehensible, detectable, incorruptible, and perceivable. AI technology should be made available for public auditing, testing and review, and subject to accountability standards … In high-stakes settings like diagnosing cancer from radiologic images, an algorithm that can't “explain its work” may pose an unacceptable risk. Thus, explainability and interpretability are absolutely required
• Accountability: AI designers and developers must bear in mind they carry a heavy responsibility on their shoulders of the outcome and impact of AI on whole human society and the universe. They must be accountable for whatever they manufacture and create.

C ONCLUSION

AI is here to stay in our world and we must try to enforce the AI bioethics of beneficence, value upholding, lucidity and accountability. Since AI is without a soul as it is, its bioethics must be transcendental to bridge the shortcoming of AI's inability to empathize. AI is a reality of the world. We must take note of what Joseph Weizenbaum, a pioneer of AI, said that we must not let computers make important decisions for us because AI as a machine will never possess human qualities such as compassion and wisdom to morally discern and judge [ 10 ]. Bioethics is not a matter of calculation but a process of conscientization. Although AI designers can up-load all information, data, and programmed to AI to function as a human being, it is still a machine and a tool. AI will always remain as AI without having authentic human feelings and the capacity to commiserate. Therefore, AI technology must be progressed with extreme caution. As Von der Leyen said in White Paper on AI – A European approach to excellence and trust : “AI must serve people, and therefore, AI must always comply with people's rights…. High-risk AI. That potentially interferes with people's rights has to be tested and certified before it reaches our single market” [ 21 ].

Financial support and sponsorship

Conflicts of interest.

There are no conflicts of interest.

R EFERENCES

• Resources Home 🏠
• Try SciSpace Copilot
• Search research papers
• Add Copilot Extension
• Try AI Detector
• Try Paraphraser
• Try Citation Generator
• April Papers
• June Papers
• July Papers

AI for thesis writing — Unveiling 7 best AI tools

Table of Contents

Writing a thesis is akin to piecing together a complex puzzle. Each research paper, every data point, and all the hours spent reading and analyzing contribute to this monumental task.

For many students, this journey is a relentless pursuit of knowledge, often marked by sleepless nights and tight deadlines.

Here, the potential of AI for writing a thesis or research papers becomes clear: artificial intelligence can step in, not to take over but to assist and guide.

Far from being just a trendy term, AI is revolutionizing academic research, offering tools that can make the task of thesis writing more manageable, more precise, and a little less overwhelming.

In this article, we’ll discuss the impact of AI on academic writing process, and articulate the best AI tools for thesis writing to enhance your thesis writing process.

The Impact of AI on Thesis Writing

Artificial Intelligence offers a supportive hand in thesis writing, adeptly navigating vast datasets, suggesting enhancements in writing, and refining the narrative.

With the integration of AI writing assistant, instead of requiring you to manually sift through endless articles, AI tools can spotlight the most pertinent pieces in mere moments. Need clarity or the right phrasing? AI-driven writing assistants are there, offering real-time feedback, ensuring your work is both articulative  and academically sound.

AI tools for thesis writing harness Natural Language Processing (NLP) to generate content, check grammar, and assist in literature reviews. Simultaneously, Machine Learning (ML) techniques enable data analysis, provide personalized research recommendations, and aid in proper citation.

And for the detailed tasks of academic formatting and referencing? AI streamlines it all, ensuring your thesis meets the highest academic standards.

However, understanding AI's role is pivotal. It's a supportive tool, not the primary author. Your thesis remains a testament to your unique perspective and voice.

AI for writing thesis is there to amplify that voice, ensuring it's heard clearly and effectively.

How AI tools supplement your thesis writing

AI tools have emerged as invaluable allies for scholars. With just a few clicks, these advanced platforms can streamline various aspects of thesis writing, from data analysis to literature review.

Let's explore how an AI tool can supplement and transform your thesis writing style and process.

Efficient literature review : AI tools can quickly scan and summarize vast amounts of literature, making the process of literature review more efficient. Instead of spending countless hours reading through papers, researchers can get concise summaries and insights, allowing them  to focus on relevant content.

Enhanced data analysis : AI algorithms can process and analyze large datasets with ease, identifying patterns, trends, and correlations that might be difficult or time-consuming for humans to detect. This capability is especially valuable in fields with massive datasets, like genomics or social sciences.

Improved writing quality : AI-powered writing assistants can provide real-time feedback on grammar, style, and coherence. They can suggest improvements, ensuring that the final draft of a research paper or thesis is of high quality.

Plagiarism detection : AI tools can scan vast databases of academic content to ensure that a researcher's work is original and free from unintentional plagiarism .

Automated citations : Managing and formatting citations is a tedious aspect of academic writing. AI citation generators  can automatically format citations according to specific journal or conference standards, reducing the chances of errors.

Personalized research recommendations : AI tools can analyze a researcher's past work and reading habits to recommend relevant papers and articles, ensuring that they stay updated with the latest in their field.

Interactive data visualization : AI can assist in creating dynamic and interactive visualizations, making it easier for researchers to present their findings in a more engaging manner.

Top 7 AI Tools for Thesis Writing

The academic field is brimming with AI tools tailored for academic paper writing. Here's a glimpse into some of the most popular and effective ones.

Here we'll talk about some of the best ai writing tools, expanding on their major uses, benefits, and reasons to consider them.

If you've ever been bogged down by the minutiae of formatting or are unsure about specific academic standards, Typeset is a lifesaver.

Typeset specializes in formatting, ensuring academic papers align with various journal and conference standards.

It automates the intricate process of academic formatting, saving you from the manual hassle and potential errors, inflating your writing experience.

An AI-driven writing assistant, Wisio elevates the quality of your thesis content. It goes beyond grammar checks, offering style suggestions tailored to academic writing.

This ensures your thesis is both grammatically correct and maintains a scholarly tone. For moments of doubt or when maintaining a consistent style becomes challenging, Wisio acts as your personal editor, providing real-time feedback.

Known for its ability to generate and refine thesis content using AI algorithms, Texti ensures logical and coherent content flow according to the academic guidelines.

When faced with writer's block or a blank page, Texti can jumpstart your thesis writing process, aiding in drafting or refining content.

JustDone is an AI for thesis writing and content creation. It offers a straightforward three-step process for generating content, from choosing a template to customizing details and enjoying the final output.

JustDone AI can generate thesis drafts based on the input provided by you. This can be particularly useful for getting started or overcoming writer's block.

This platform can refine and enhance the editing process, ensuring it aligns with academic standards and is free from common errors. Moreover, it can process and analyze data, helping researchers identify patterns, trends, and insights that might be crucial for their thesis.

Tailored for academic writing, Writefull offers style suggestions to ensure your content maintains a scholarly tone.

This AI for thesis writing provides feedback on your language use, suggesting improvements in grammar, vocabulary, and structure . Moreover, it compares your written content against a vast database of academic texts. This helps in ensuring that your writing is in line with academic standards.

Isaac Editor

For those seeking an all-in-one solution for writing, editing, and refining, Isaac Editor offers a comprehensive platform.

Combining traditional text editor features with AI, Isaac Editor streamlines the writing process. It's an all-in-one solution for writing, editing, and refining, ensuring your content is of the highest quality.

PaperPal , an AI-powered personal writing assistant, enhances academic writing skills, particularly for PhD thesis writing and English editing.

This AI for thesis writing offers comprehensive grammar, spelling, punctuation, and readability suggestions, along with detailed English writing tips.

It offers grammar checks, providing insights on rephrasing sentences, improving article structure, and other edits to refine academic writing.

The platform also offers tools like "Paperpal for Word" and "Paperpal for Web" to provide real-time editing suggestions, and "Paperpal for Manuscript" for a thorough check of completed articles or theses.

Is it ethical to use AI for thesis writing?

The AI for writing thesis has ignited discussions on authenticity. While AI tools offer unparalleled assistance, it's vital to maintain originality and not become overly reliant. Research thrives on unique contributions, and AI should be a supportive tool, not a replacement.

The key question: Can a thesis, significantly aided by AI, still be viewed as an original piece of work?

AI tools can simplify research, offer grammar corrections, and even produce content. However, there's a fine line between using AI as a helpful tool and becoming overly dependent on it.

In essence, while AI offers numerous advantages for thesis writing, it's crucial to use it judiciously. AI should complement human effort, not replace it. The challenge is to strike the right balance, ensuring genuine research contributions while leveraging AI's capabilities.

Wrapping Up

Nowadays, it's evident that AI tools are not just fleeting trends but pivotal game-changers.

They're reshaping how we approach, structure, and refine our theses, making the process more efficient and the output more impactful. But amidst this technological revolution, it's essential to remember the heart of any thesis: the researcher's unique voice and perspective .

AI tools are here to amplify that voice, not overshadow it. They're guiding you through the vast sea of information, ensuring our research stands out and resonates.

Try these tools out and let us know what worked for you the best.

Love using SciSpace tools? Enjoy discounts! Use SR40 (40% off yearly) and SR20 (20% off monthly). Claim yours here 👉 SciSpace Premium

Frequently Asked Questions

Yes, you can use AI to assist in writing your thesis. AI tools can help streamline various aspects of the writing process, such as data analysis, literature review, grammar checks, and content refinement.

However, it's essential to use AI as a supportive tool and not a replacement for original research and critical thinking. Your thesis should reflect your unique perspective and voice.

Yes, there are AI tools designed to assist in writing research papers. These tools can generate content, suggest improvements, help with formatting, and even provide real-time feedback on grammar and coherence.

Examples include Typeset, JustDone, Writefull, and Texti. However, while they can aid the process, the primary research, analysis, and conclusions should come from the researcher.

The "best" AI for writing papers depends on your specific needs. For content generation and refinement, Texti is a strong contender.

For grammar checks and style suggestions tailored to academic writing, Writefull is highly recommended. JustDone offers a user-friendly interface for content creation. It's advisable to explore different tools and choose one that aligns with your requirements.

To use AI for writing your thesis:

1. Identify the areas where you need assistance, such as literature review, data analysis, content generation, or grammar checks.

2. Choose an AI tool tailored for academic writing, like Typeset, JustDone, Texti, or Writefull.

3. Integrate the tool into your writing process. This could mean using it as a browser extension, a standalone application, or a plugin for your word processor.

4. As you write or review content, use the AI tool for real-time feedback, suggestions, or content generation.

5. Always review and critically assess the suggestions or content provided by the AI to ensure it aligns with your research goals and maintains academic integrity.

You might also like

What is a thesis | A Complete Guide with Examples

MIT Libraries home DSpace@MIT

• DSpace@MIT Home
• MIT Libraries
• Undergraduate Theses

The physics of artificial intelligence

Other Contributors

Terms of use, description, date issued, collections.

• Graduate School of Design
• GSD Theses and Dissertations
• Communities & Collections
• By Issue Date
• FAS Department
• Quick submit
• Waiver Generator
• DASH Stories
• Accessibility
• COVID-related Research

Terms of Use

• Privacy Policy
• By Collections
• By Departments

Interlacing Latent Features: Synthesis of Past and Present in Architectural Design through Artificial Intelligence in a Case Study of Japanese Houses

Citable link to this page

Collections.

• GSD Theses and Dissertations [355]

Contact administrator regarding this item (to report mistakes or request changes)

Thesis: A strategic perspective on the commercialization of artificial intelligence

Submitted by Siddhartha Ray Barua.

Abstract: Many companies are increasing their focus on Artificial Intelligence as they incorporate Machine Learning and Cognitive technologies into their current offerings. Industries ranging from healthcare, pharmaceuticals, finance, automotive, retail, manufacturing and so many others are all trying to deploy and scale enterprise Al systems while reducing their risk. Companies regularly struggle with finding appropriate and applicable use cases around Artificial Intelligence and Machine Learning projects. The field of Artificial Intelligence has a rich set of literature for modeling of technical systems that implement Machine Learning and Deep Learning methods. This thesis attempts to connect the literature for business and technology and for evolution and adoption of technology to the emergent properties of Artificial Intelligence systems. The aim of this research is to identify high and low value market segments and use cases within the industries, prognosticate the evolution of different Al technologies and begin to outline the implications of commercialization of such technologies for various stakeholders. This thesis also provides a framework to better prepare business owners to commercialize Artificial Intelligence technologies to satisfy their strategic goals.

To read the complete article, visit DSpace at the MIT Libraries .

Artificial Intelligence Thesis

03 Jan 2024 - 29 Dec 2024

Permission of the Head of Department or delegated authority.

An externally examined piece of written work that reports on the findings of supervised research.

Teaching Periods and Locations

If your paper outline is not linked below, try the previous year's version of this paper .

Indicative Fees

You will be sent an enrolment agreement which will confirm your fees. Tuition fees shown are indicative only and may change. There are additional fees and charges related to enrolment - please see the  Table of Fees and Charges for more information.

Available subjects

Artificial intelligence, additional information.

Subject regulations

• Paper details current as of 28 Jan 2024 00:05am
• Indicative fees current as of 9 Apr 2024 01:30am

You’re viewing this website as a domestic student

You’re currently viewing the website as a domestic student, you might want to change to international.

You're a domestic student if you are:

• A citizen of New Zealand or Australia
• A New Zealand permanent resident

You're an International student if you are:

• Intending to study on a student visa
• Not a citizen of New Zealand or Australia

Cookie Acknowledgement

This website uses cookies to collect information to improve your browsing experience. Please review our Privacy Statement for more information.

• College of Engineering
• News Center

Auburn Engineering to offer new artificial intelligence programs beginning this fall

Published: Apr 10, 2024 10:05 AM

By Joe McAdory

Auburn Engineering’s Department of Computer Science and Software Engineering (CSSE) will offer three artificial intelligence (AI) engineering degree and certificate programs beginning in Fall 2024.

The programs — master’s degree in AI engineering, graduate certificate in AI engineering and undergraduate certificate in AI engineering — will provide various levels of technical depth to broaden students’ skillsets as they enter the workforce, CSSE Chair Hari Narayanan, said.

“Having a solid artificial intelligence education is a growing demand for the engineering workforce and we look forward to supplying organizations with highly skilled graduates,” Narayanan said. “Employers are looking for people with AI-related skills… even in the most unlikely places. Artificial intelligence promises to revolutionize sectors such as business, defense, education, government and health care. To succeed in these roles, people need to understand how to utilize AI systems productively and efficiently. We aim to meet that need.”

The 30-hour master’s degree curriculum in AI engineering, which can be taken online or on-campus, is open to all students who have already earned a baccalaureate degree from an institution of recognized standing, either in computer science, software engineering, or another STEM discipline, and to those who have professional experience related to computing or IT and familiarity with mathematics. This program requires students to take three core courses — artificial intelligence, machine learning and data mining — and seven technical electives related to AI.

“The graduate program provides in-depth training, including the possibility for students to specialize,” Narayanan said. “For example, a student could say, ‘I want in-depth training in AI, but I also want to know how to use it within cyber security, computer networks, or something else.’”

In the graduate program, students will learn how to:

• Develop algorithms and methodologies for AI and machine learning systems and technologies.
• Incorporate software engineering principles to analyze, design and implement AI and machine learning software.
• Apply AI and machine learning techniques to solve complex engineering problems and problems of societal importance.
• Deliver written and oral presentations to non-technical and technical audiences.

Like the standard graduate degree program, admission into the graduate certificate program in AI engineering is open to all students with a baccalaureate degree from an institution of recognized standing, either in computer science, software engineering, another STEM discipline or have relevant professional experience and background knowledge.

“The graduate certificate might be appealing to students who have been working within industry for a while and want to upskill,” Narayanan said. “Maybe not enough to design and build a new AI system, but enough to make decisions about what AI technologies and tools to use and how best to use them.”

This 12-hour program, available on-campus or online, includes courses in AI, machine learning and data mining, along with an approved AI-related elective.

Narayaran said the undergraduate certificate in AI engineering is currently open only to CSSE majors.

“Students in this program will develop the deep technical knowledge and skills to solve real-world problems and issues faced by government, industry and society,” he said. “Blending this AI certificate with their chosen CSSE major in computer science or software engineering can be very advantageous for these students.”

The undergraduate certificate program’s 12-hour curriculum models the graduate certificate and includes courses in AI, machine learning and data mining with an approved AI-related elective.

“Developing cutting-edge programs like these — and meeting the growing needs of industry — is expected of leading engineering colleges in the nation,” Narayanan said. “In fact, these programs on engineering AI systems are the first of their kind in Alabama.”

Among the largest departments at Auburn University, CSSE currently offers a variety of degree and certificate programs, including B.S. in Computer Science, Bachelor of Software Engineering, Bachelor of Computer Science (online), Computer Science Minor, Undergraduate Cyber Defense Certificate, M.S. in Computer Science and Software Engineering (thesis or non-thesis options), M.S. in Cybersecurity Engineering, Graduate Certificate in Cybersecurity Engineering, M.S. in Data Science and Engineering (data science or data engineering options), Graduate Certificate in Data Engineering, and Doctor of Philosophy.

Visit here for more information about computer science and software engineering at Auburn University.

New programs include a master’s degree in AI engineering, graduate certificate in AI engineering and undergraduate certificate in AI engineering.

Featured Faculty

Computer Science and Software Engineering

Recent Headlines

The obscure politics of artificial intelligence: a Marxian socio-technical critique of the AI alignment problem thesis

• Original Research
• Open access
• Published: 08 April 2024

Cite this article

You have full access to this open access article

• Federico Cugurullo   ORCID: orcid.org/0000-0002-0625-8868 1  

179 Accesses

Explore all metrics

There is a growing feeling that artificial intelligence (AI) is getting out of control. Many AI experts worldwide stress that great care must be taken on the so-called alignment problem , broadly understood as the challenge of developing AIs whose actions are in line with human values and goals. The story goes that ever more powerful AI systems are escaping human control and might soon operate in a manner that is no longer guided by human purposes. This is what we call the AI-out-of-control discourse which, in this paper, we critically examine and debunk. Drawing on complementary insights from political theory, socio-technical studies and Marxian political economy, we critique the supposed animistic and autonomous nature of AI, and the myth of the uncontrollability of AI. The problem is not that humanity has lost control over AI, but that only a minority of powerful stakeholders are controlling its creation and diffusion, through politically undemocratic processes of decision-making. In these terms, we reframe the alignment problem thesis with an emphasis on citizen engagement and public political participation. We shed light on the existing politics of AI and contemplate alternative political expressions whereby citizens steer AI development or stop it in the first place.

Similar content being viewed by others

The Ethics of AI Ethics: An Evaluation of Guidelines

Thilo Hagendorff

Education for AI, not AI for Education: The Role of Education and Ethics in National AI Policy Strategies

Daniel Schiff

In AI We Trust: Ethics, Artificial Intelligence, and Reliability

Avoid common mistakes on your manuscript.

1 Introduction: AI out of control?

Recently, there seems to be a growing feeling that the development of artificial intelligence (AI) is getting out of control. Such concern has escalated with the release of ChatGPT-4 and it can be observed by looking at three dimensions. First, the public. As of this writing, ChatGPT-4 has over 150 million users and many of them, according to recent studies, seem to believe that this AI has unprecedented human-like properties [ 1 ]. This is one of the most powerful AIs ever built, which has entered the mundane spaces of everyday life, with people now able to have conversations with a digital intelligence from the comfort of their homes [ 2 ]. Second, the computer scientists and engineers who are building AIs. Emblematic is the paper published by a Microsoft team in which the authors argue that GPT-4 exhibits remarkable capabilities that are strikingly close to human-level performance, thereby showing early signs of general intelligence [ 3 ]. They also warn us about the risks of this AI technology, stressing that ‘great care would have to be taken on alignment and safety’ [3 p.2]. Third, there are AI experts worldwide who are monitoring the development of AI and expressing concern over the rapid pace of its innovation. Here the most prominent example, which we use in this paper to critically discuss the seemingly out-of-control nature of AI development, is the Future of Life Institute’s (FLI) open letter published in March 2023.

The letter in question was written and then endorsed by prominent AI scientists and public intellectuals such as Max Tegmark, Stuart Russell and Yuval Noah Harari, asking all AI companies to pause the development of AI systems with a computational power akin or superior to that of GPT-4 for six months [ 4 ]. Moreover, the letter calls on governments from all around the world to impose a moratorium on AI development should AI companies like OpenAI and Google keep operating business as usual. Together with its compendium, a longer policy brief [ 5 ], the FLI letter offers very useful materials to understand the seemingly general feeling that the development of AI is getting out of control, why this is happening and, above all, what the biggest risks are according to what we refer to as the AI-out-of-control discourse .

The FLI letter notes that ‘recent months have seen AI labs locked in an out-of-control race to develop and deploy ever more powerful digital minds that no one – not even their creators – can understand, predict or reliably control’ [4 p.1]. They describe a type of technological development that has gotten out of control in terms of both pace and outcomes. Among the key risks that the authors of the letter highlight are the creation of ‘non-human minds that might eventually outnumber, outsmart, obsolete and replace us’, and ‘the loss of control of our civilization’ [4 p.1]. Here the meaning is not simply that the development of AI technology is getting out of control, but that AI itself as a non-human mind might soon get out of control and operate in manners that go against human values, interests and goals. In relation to emerging AI systems akin to GPT-4, the letter’s compendium states that ‘the systems could themselves pursue goals, either human or self-assigned, in ways that place negligible value on human rights, human safety or, in the most harrowing scenarios, human existence’ [4 p.4]. In this regard, one of their policy recommendations is ‘a significant increase in public funding for technical AI safety research in […] alignment: development of technical mechanisms for ensuring AI systems learn and perform in accordance with intended expectations, intentions and values’ so that they are ‘aligned with human values and intentions’ [4 p.11].

Essentially, what the FLI is talking about is the so-called alignment problem . This has been confirmed by Max Tegmark (founder of the FLI) in an interview that he gave to announce and discuss the FLI open letter and the proposed pause and moratorium on AI development. He said: ‘The AI alignment problem is the most important problem for humanity to ever solve’ [ 6 ]. In this paper we take a different stance. Using the FLI open letter as an entry point, our aim is to critically examine the alignment problem thesis, expose some of the main conceptual flaws that undermine its understanding, and propose an alternative thesis on the way contemporary AI technologies are aligned. We argue that, as it is currently formulated in mainstream public discourses, the alignment problem presents several conceptual issues that are problematic not simply from a theoretical perspective. The alignment problem has also policy implications that are influencing the development of AI technologies in a way that fails to recognize important political aspects.

Our critique proceeds in five steps. First, we present the AI alignment problem thesis as it is generally understood particularly in the fields of computer science and philosophy. Second, we review the academic literature within which this contribution is situated, with a focus on socio-technical approaches to the study of AI. Third, we introduce three fundamental misconceptions that undermine the understanding of the alignment problem as it is currently formulated in line with the FLI open letter. These are the idea that AI has the capacity to act on its own volition, the belief that AI development is getting out of control, and the assumption that these are technical and philosophical issues that can be fixed by improving the technology and philosophy underpinning AI innovation. Fourth, we unpack each misconception in an attempt to explain what is causing it. More specifically, we analyze the animistic tendency that people have had particularly since the Industrial Revolution toward new technologies, intended as the belief that lifelike properties such as will and consciousness can be found in machines. We explain this attitude in the age of AI as a “shortcut” that people resort to, in order to make sense of an otherwise unintelligible technology; with the average user more willing to project consciousness on a complex technology such as ChatGPT-4 than to study its algorithms. Then, we draw on the work of Langton Winner and Karl Marx to debunk the myth of the uncontrollability of AI technology; we do so conceptually by building upon complementary insights from political theory and political economy, and empirically through the support of real-life examples showing how across different scales (countries, regions and cities) there are specific actors who are steering the development of AI. Subsequently, we expose the limits of computer science and philosophy when it comes to understanding and solving the political side of the alignment problem; that is the hidden network of stakeholders behind the production of AI. Finally, we reframe the alignment problem thesis with an emphasis on questions of citizen participation and public political engagement, and define key areas for future research.

Overall, as in this paper we are adopting a theoretical approach, our contribution is theoretical in nature and made of two parts. It expands and updates critical political theory on the alleged uncontrollability and autonomy of technology, in light of the recent concerns raised by AI development. It also fleshes out theoretically the AI alignment problem thesis and contributes to its understanding by adding a hitherto missing Marxian socio-technical perspective.

2 Socio-technical perspectives on AI: a review of the literature

This study is informed by and situated within an emerging socio-technical approach to the analysis of the development of AI. At the core of this approach is the awareness of a profound ‘interaction of social and technical processes’ in the production of AI [ 7 p. 180]. Drawing on Latour, Venturini reminds us how ‘the evolution of humans and technologies is a chronicle of mutual entanglement and escalating interdependence’ which we can now observe in the way AI technologies are being generated as part of broader socio-political systems [ 8 p. 107]. There are many examples of interactions between social forces and technological components in the genesis and operation of AI. Pasquinelli, for instance, points out that AI technologies learn by absorbing data that are labelled by humans [ 9 ]. Much of this learning process, so-called machine learning , takes place in the human (and hence social) environment par excellence: the city [ 10 ]. This is where multiple AIs assimilate vast volumes of data by observing human behavior in practice [ 2 ]. In so doing, AIs populate urban spaces and influence the composition of the city itself which is evolving into a complex socio-technical system made of both human and artificial intelligences [ 11 , 12 ].

Recent socio-technical approaches to the study of AI also remind us that this is a technology that is ultimately experienced by humans whose feelings toward its potential benefits and harms influence how it spreads across society [ 13 ]. Moreover, ‘people’s attitudes and perceptions are crucial in the formation and reproduction of the socio-technical imaginaries that sustain technological development’ in the field of AI [ 14 p. 455]. In these terms, in line with Jasanoff’s theory, human feelings contribute to the development of visions of desirable (or undesirable) futures animated by social expectations of technology, which in turn steer the course of technological development [ 15 ].

This approach then recognizes human responsibility in the trajectory and speed of AI development, which is particularly relevant in the context of this critical study on the alignment problem and what we referred to in the introduction as the AI-out-of-control discourse . Korinek and Balwit, for example, frame the alignment problem as a social problem, by taking into account the welfare of the plethora of individuals affected by AI systems, as well as the many stakeholders who actively govern their technological development [ 16 ]. Similarly, from a politico-economic perspective, Dafoe stresses how the production of AI is controlled by countries and multinationals seeking to seize economic benefits [ 17 ]. We adopt a similar approach and draw on Marxian political economy to contribute to the emerging socio-technical literature on AI and its development. As Pasquinelli remarks, back in the 19th century Marx had already understood that ‘the machine is a social relation, not a thing’ and, in this sense, he can be considered a precursor of the socio-technical approach reviewed in this section. [ 18 p. 119]. Marx’s theory of political economy has not been mobilized yet, together with more recent socio-technical perspectives, to critique the mainstream understanding of the alignment problem and identify its adverse policy implications, and this is precisely the gap in knowledge where the following contribution is situated.

3 The alignment problem

The so-called alignment problem is part of a complex and multifaceted debate that extends well beyond academia and encompasses industry, policy and news media too [ 19 ]. This is an ongoing and rapidly evolving debate to which many heterogeneous voices, ranging from AI experts to public intellectuals, are contributing and, as such, it cannot be confined to one single expression. As the purpose of this paper is to critically examine the AI-out-of-control discourse expressed prominently by the FLI open letter discussed in the introduction, in this section we begin by explaining the nature of the alignment problem as it is understood by the FLI and those who adhere to its stance. Step by step, we will then expose the limitations of the FLI’s view on the alignment problem and integrate alternative perspectives, in order to present a more balanced view on the alignment of AI development.

Through the perspective of the FLI open letter, the alignment problem refers mainly to the challenge of developing AIs whose actions are in line with human values and goals [ 20 , 21 , 22 ]. This is described and presented as a problem because, for scholars such as Yudkowsky [ 23 ] and Bostrom [ 24 ], there is no guarantee that powerful AI systems (especially hypothetical superintelligent and conscious ones) will act in a way that is compatible with human values and with the preservation of humanity itself. Bostrom [ 24 ] explains this philosophically through the Orthogonality thesis pointing out that intelligence and goals are two orthogonal axes moving along different directions. He concludes that any type of intelligence can in principle follow any type of goal, and that it would be thus safe to assume that a non-human superintelligence might not necessarily follow a humanistic goal [ 24 ].

In turn, Bostrom’s reflections have led computer scientists like Russell [ 25 ] and Hendrycks et al. [ 26 ] to conclude that it is vital to build human-compatible AIs. In practice, as Gabriel [ 27 ] explains, solving the alignment problem has two components, each one encapsulating a set of different but interrelated tasks. The first one is technical in nature and is about formally encoding human values in AI [ 27 ]. This is essentially a computer science challenge consisting in building an ethical understanding of what is good or bad, right or wrong, directly into the algorithms that steer the actions of a given AI [ 25 ]. The second one is philosophical in nature and is about determining what is good and what is bad in the first place, and choosing what specific values will be integrated into AI [ 27 , 28 ].

This is not the place to discuss the practicalities of the alignment problem in-depth, but it is worth mentioning that both components are intrinsically connected to each other, and that over the years have led to interdisciplinary collaborations between computer scientists and ethicists. In this respect, Gabriel’s example is emblematic given that he is a philosopher working for DeepMind i.e. Google. In addition, each individual component is per se hyper complex. The first one is a matter of how AI learns, which makes it a machine learning issue. In his overview of the technical challenge underpinning the alignment problem, Gabriel [ 27 ] notes how there are different machine learning approaches ranging from supervised learning to unsupervised learning, and more experimental techniques such as inverse reinforcement learning whereby AI is not implanted with objectives, but instead ‘can learn more about human preferences from the observation of human behavior’ [ 29 p.8].

There is still no consensus in the computer science community over what the best approach is. As for the philosophical problem, it is important to remember that ethics has been changing and varying for centuries across spaces and times, and moral dilemmas over what is good or bad in the age of AI abound. For instance, what the right conduct should be for AIs that are already present in our society, such as autonomous cars, is a controversial research topic that shows a considerable degree of dissensus within both academia and the public [ 30 , 31 , 32 ]. Above all, it is important to remember that there is a lack of consensus regarding the nature of the alignment problem itself and its very premises. For example, the theories of Yudkowsky [ 23 ] and Bostrom [ 24 ] concerning the hypothetical emergence of superintelligences have been critiqued and dismissed by other voices in the debate, as sci-fi distractions that risk hindering our focus on already existing problems [ 33 , 34 ].

Similarly, critical social scientists have stressed that AI does not need to be superintelligent to cause harm and that contemporary narrow AIs are already responsible for causing social injustice and environmental degradation [ 35 , 36 ]. These are AIs whose detrimental actions are not based on malign intents since, as it has been repeatedly stressed in critical AI studies, artificial intelligences are amoral entities that are indifferent to questions of right or wrong [ 35 ]. In these terms, the critical side of the debate on the alignment problem ascribes responsibility to the many human stakeholders who shape the development of AI and, therefore, its outcomes. This is why, according to Korinek and Balwit, governance is a key component to the resolution of the alignment problem which then emergences as a social problem, rather than a solely technical and philosophical challenge [ 16 ]. It is in line with these critical perspectives, that we now turn to the FLI’s stance and expose its limitations in an attempt to contribute to a more comprehensive and balanced understanding of the alignment problem.

4 Conceptual misunderstandings of the alignment problem

The alignment problem, as it is understood by the FLI and recognized by scholars in favor of its open letter, presents some critical and interrelated issues in the way the problem is conceptualized, which then risk generating practical repercussions in terms of policy. Conceptual issues can undermine how we come to understand the alignment problem as such, and what in theory is causing it. This is crucial and worth examining because conceptual misunderstandings can severely undermine all the actions and policies that are supposed to mitigate or fix the alignment problem. In essence, if we fail to properly understand a problem, any solution that we conceive becomes useless.

We argue that there are currently three main conceptual issues affecting the mainstream understanding of the AI alignment problem, promoted by the FLI:

The idea of AI as agentic ; meaning a non-human intelligence that has the capacity to act by drawing upon a force of its own, in a way that risks being not aligned with the goals and interests of humanity. Such misunderstanding erroneously pictures humans out-of-the-loop and autonomous AIs in pursuit of non-human objectives.

The perception that the development of powerful AIs and the outcomes of their actions are getting out of control. This is the essence of the AI-out-of-control discourse that we mentioned in introduction, according to which AI development has rapidly become unmanageable in the sense that humans are neither capable to stop it, nor able to direct it towards a scenario where AI does exactly what we want it to do. This misunderstanding erroneously depicts AIs that are no longer guided by human purposes.

The belief that this is a technical and philosophical problem. This misunderstanding erroneously presents a problem that can be solved by means of better algorithms that make AI act exactly the way we want it to act, according to ethical ideals of what is good or bad defined by philosophers.

Such ideas, beliefs and perceptions present some fundamental misconceptions that it is important and urgent to shed light on. There is indeed an alignment problem but, as we will see in the remainder of the paper, it is much more complex than how the FLI and its open letter are depicting it. The FLI’s stance on the alignment problem is not complete and misses crucial social and political aspects that we aim to identify and discuss, in order to add to the debate. Next, we are going to critically examine each of the conceptual issues introduced above, by drawing upon a combination of insights from political theory, socio-technical literature, and Marxian political economy.

5 First conceptual issue: the supposed animus of AI

The idea of technology as an agentic entity capable of drawing upon a force of its own to act and influence the surrounding physical and social environment, is not new. This is a recurring idea that tends to appear again and again at the dawn of major technological revolutions that spawn machines presenting unprecedented capabilities and functions. One of the main examples of this phenomenon in modern history is Marx’s account in the 1800s of the new technologies of the Industrial Revolution, in which we can find clear signs of animism . This is the belief that lifelike properties can be found beyond the human realm and that objects, for instance, including human-made technologies can be animated and alive, thus exhibiting consciousness and possessing the capacity to act according to intentions of their own.

An animistic line of thought posits that animated object can be benign or malign in the way they engage with the human population. In Das Kapital (1867), Marx takes a negative stance toward the animus driving the actions and outcomes of the technologies of the Industrial Revolution. He depicts the 1800s factory and its technological apparatus as a gigantic entity animated by ‘demonic power’, pulsing with engines described as ‘organs’ and acting by means of ‘limbs’ that extend endlessly like mechanical tentacles [ 37 p.416]. His account draws upon a complex techno-gothic imaginary merging then popular gothic novels, including the work of John William Polidori and Mary Shelley, with the technological innovations of the 19th century, through which technology is at times portrayed like a vampire that drains the life and soul of the workers who are creating it [ 38 ].

Similar traces of animism can be found during another major technological revolution: the development of home computers in the 1970s and their diffusion in the 1980s. This becomes particularly evident in the mid 1980s when computer technology begun to be widely popularized through personal computers (PC). It is through the PC that a number of people started to interact with then new computer technologies, finding themselves in the position of having to make sense of this type of technology. As Stahl [ 39 ] observes, in the 1980s discourses surrounding personal computers do not simply refer to technology as an object. For him, the discourse in question has a quasi-magical tone and describes computers as machines possessing intelligence and capable of talking and acting [ 39 ]. He notes that back then ‘the machine was frequently portrayed as the active partner’ in the emerging relationships between humans and computers whom ‘were spoken of in the active voice, as if they had volition’ [ 39 p.246].

Volition is the act of making a conscious choice. It implies having intentions, will and other life-like properties that we commonly associate with humanity rather than machinery. But in the 1980s this conceptual association was being reshaped by new cultural perceptions of technology. When Marx was writing about the technologies of the 19th century, the zeitgeist was characterized (particularly in the West) by a techno-gothic imaginary whereby the properties of the new machines of the industrial age were being interpreted through the lens of monstrous myths. Throughout the 1970s and 1980s instead, we find an emerging sci-fi imaginary promoted by authors such as Isaac Asimov and William Gibson, that introduced to the public the idea of intelligent machines and cultivated the early fantasies about AI [ 40 ]. It is not a coincidence that, around the same time, we have important sociological studies, notably the work of Sherry Turkle [ 41 ], showing that some people were under the impression that computers were intelligent and alive.

The same animistic thread remerges in our contemporary society, starting from the 2010s when relatively powerful AIs begin to enter domestic spaces and everyday life, thereby interacting with a growing number of people. A prominent example of this phenomenon is Siri launched in 2011 by Apple. This is a powerful AI in the sense that it can engage in basic forms of conversation and mediate several activities. It is a step forward compared to the PCs of the 1980s not simply in terms of computational capacity, but also because of how pervasive the technology is. This is a technology that like Amazon’s Alexa is meant to penetrate into people’s private spaces and life; its designed personality seemingly human-like expressed through names, gender and a feminine tone of voice [ 42 ].

For scholars such as Marenko [ 43 p.221], AIs like Siri are triggering a new wave of animism, a digital neo-animism , as ‘we often end up treating our smartphone as if it is alive.’ In these terms, neo-animism is understood as the contemporary belief that novel AI technologies possess lifelike properties such as will, intentionality and consciousness [ 43 ]. Within this strand of literature, the perception of emerging AIs is that of actants which animate our everyday spaces and objects, and ultimately influence our life [ 44 ]. In reality, as we have seen in the first part of this section, there is nothing new about such animistic response to new technologies. Today the response is essentially the same that Marx [ 37 ] gave in the 1800s and that Stahl [ 39 ] reports in the 1980s. The nature of technology has considerably changed, but people’s reaction has not: we find an analogous animistic tendency that portrays a given new technology as an agentic entity capable of acting upon a force and intention of its own. There is thus a long-standing connection between animism and the idea of agentic technologies, which cuts across the last two centuries of the history of technology and our social attitudes toward it.

From the industrial machines of the 1800s to the PCs of the 1980s and, more recently, in relation to the many AIs that permeate our daily life we find the same animistic common denominator. To dig deeper into the subject matter, we need to ask the following questions: Why is this happening? Why does this keep happening through the ages? Why do we develop these animistic tendencies toward new technologies? The answers lie in the longstanding habit of human users to ‘project agency, feelings and creativity onto machines’ [ 45 p.2]. This is animism in action, whereby an inanimate technology, nowadays AI, is perceived as an agentic entity. According to Marenko and Van Allen [ 44 p.54], ‘users tend to attribute personality, agency and intentionality to devices because it is the easiest route to explain behavior.’ In these terms, it is easier to believe that a machine has an animus, than to comprehend the complex mechanisms that make it work in a certain way. Following this line of thought and applying it to the recent wave of AI technologies that is investing our society, we can posit that it is easier to believe that Alexa has a personality or that ChatGPT exhibits some degree of consciousness, compared to how incredibly difficult it would be for most people to study the algorithms whereby these AIs perceive reality and act on it. Animism becomes then a “shortcut” to make sense of AI’s behavior, by assuming that this is an agentic technology possessing lifelike properties, such as consciousness and volition, instead of making an effort to understand the algorithms that make AI behaves in a certain way.

In addition, there is another common denominator underpinning the various waves of animism discussed in this section, and that is the occult of new technologies. A new technology tends to be occult in the sense that its functioning is often beyond the realm of human comprehension, apart from a small group of initiates: those who are building and studying the technology in question. This was true in the 1800s when anyone without a working knowledge of engineering could barely comprehend how then new engines were functioning and why mechanical devices seemed to move on their own. The same was true in the 1980s when most people without a background in computer science could not understand the software that was making PCs work. This is even truer today when the very computer scientists and engineers who are designing and building Large Language Models (LLMs) such as GPT-4 do not fully understand their own creations. As computer scientist Sam Bowman remarks, ‘any attempt at a precise explanation of an LLM’s behavior is doomed to be too complex for any human to understand’ given the countless connections among artificial neurons at play in the production of just one piece of text [ 46 ].

As Greenfield [ 47 ] puts it, AI is an arcane technology that tends to escape human comprehension. Recent empirical research suggests that, around the world, levels of so-called AI literacy , which is the ability to understand and use AI technologies, are low and many users exhibit ‘the tendency to attribute human-like characteristics or attributes to AI systems’ [ 48 p.5]. This problematic epistemological aspect has been repeatedly stressed in Explainable Artificial Intelligence (XAI) literature where AI is often portrayed as a black box intended as a device whose inner workings are extremely difficult to understand [ 49 , 50 , 51 ]. However, while the black box narrative is getting very popular nowadays to highlight the scarce intelligibility of AI systems, it is far from being new. Black box is the same exact term that Stahl [ 39 ] employed to unpack the animistic and quasi-magical attitudes that people had in the 1980s toward PCs. As he observed back then: ‘Computers were powerful, but also mysterious. Their power was ours to use, but not to understand. When technology is a black box, it becomes magical’ [ 39 p.252].

6 Second conceptual issue: the myth of the uncontrollability of AI

The first conceptual issue is connected to the second one: the belief that the development of AI is getting out of control. As Nyholm [ 52 ] notes, ‘whenever there is talk about any form of AI or new technologies more generally, worries about control tend to come up.’ In essence, this is the worry that we humans are unable to safely manage (let alone stop) the creation and diffusion of ever more powerful AIs that appear to be acting upon their own volition. In the previous section, we have critically discussed the animistic notion of technology as a human tendency to project personality and consciousness on new technologies, and to believe that such technologies have volition and can make conscious choices. As we have seen, this is a tendency that becomes evident from the Industrial Revolution onwards and that is triggered by humans’ incapacity to understand how a device showing unprecedented capabilities (being it a 1800s engine or a 21st century Large Language Model such as GPT-4) actually works.

In this section, we tackle the conceptual issue of AI as a technology that seems to be getting out of control. That of technology out of control is a recurring theme in the work of political theorist Langton Winner who has contributed to the development of a critical political theory of technology with his notion of autonomous technology . His is one the initial attempts to develop a socio-technical perspective on the study of technology, in a way that recognizes human responsibility in the trajectory of technological development. According to Winner [ 53 pp.13–15], ‘the idea of autonomous technology’ is ‘the belief that somehow technology has gotten out of control and follows its own course.’ For him [ibid], a technology out of control is one that is running amok ‘and is no longer guided by human purposes’ or controlled by human agency. As we can see from this passage, Winner himself talks about the phenomenon of technology out of control not as a fact, but as a belief . This is a belief that he questions, trying to understand why for a very long time (but particularly in modern and contemporary history) many people seem to believe that the development of technology and its outcomes have gotten out of control.

In his explanation, Winner stresses one issue in particular: speed. He reflects on the velocity of technological innovation, remarking how quickly ‘technology-associated alterations take place’ [ 53 p.89]. In addition, he notes that many of the changes triggered by technological development are usually unintended, concluding that ‘technology always does more than we intend’ [ 53 p.98]. We can build on Winner’s reflections and add that technology-associated alterations that are unintended tend to be also unexpected . While a major technological development triggers changes and leads to outcomes that were unintended, it naturally finds many people unprepared since much of those changes and outcomes were not expected. Referring back to the examples and historical periods discussed in the previous section, we know for instance that some of the major changes and outcomes produced by the Industrial Revolution were both unintended and unexpected, at least for most of the population. The increase in productivity enabled by the new machines of the 19th century were indeed intended and expected [ 54 ]. However, the negative environmental changes (heavy pollution and destruction of natural habitat, in particular) that were triggered by the extraction and consumption of the resources necessary to build and power 1800s machines were not. Nor were the radical socio-economic and geographical transformations associated with the technological innovation of that period. This includes the rapid growth of large and polluted industrial cities and the process of suburbanization whereby the rich were trying to escape from the smokes of industry [ 55 , 56 ]. Not to mention the historical records indicating a significant increase in infectious diseases, alcoholism, domestic violence and, thus, death rates in the large and overcrowded cities of the Industrial Revolution [ 54 , 57 ].

If you were experiencing similar changes (substantial in nature and taking place at a fast pace) it would be easy to feel in a position of no control over technological development. When the production of new technology leads to outcomes that were unintended, unexpected and alter the surrounding social and physical environment, in the words of Winner [ 53 pp.89, 97] ‘we find ourselves both surprised and impotent – victims of technological drift’ with ‘societies going adrift in a vast sea of unintended consequences’ triggered by technological innovation. However, this is partly an illusion. As Winner [ 53 p.53] notes, ‘behind modernization are always the modernizers; behind industrialization, the industrialists.’ The reality then is not that we have lost control of technological development: some of us are in control, and this is usually a minority of powerful individuals who make conscious and deliberate decisions that shape the direction and outcomes of technological innovation. There are thus potent socio-political forces that steer the trajectory of the development of technology, but these are associated with a type of power (the power to shape technological innovation) that is unevenly distributed across society.

Winner [ 53 p.53] is adamant in affirming that ‘the notion that people have lost any of their ability to make choices or exercise control over the course of technological change is unthinkable; for behind the massive process of transformation one always finds a realm of human motives and conscious decisions in which actors at various levels determine which kinds of apparatus, technique, and organizations are going to be developed and applied.’ The problem in contemporary discourses about technological development is the use of general terms such as people and society , implying (like in the case of the FLI open letter) that the whole humanity has lost control over the development of AI technology. Winner’s studies remind us that there are always specific actors making choices that determine the nature, scope and place of technology.

This is the same conclusion that Marx had reached upon reflecting on the origin of 19th century technologies. Despite the animistic passages in Das Kapital quoted in the previous section, Marx was well aware of the fact that the new technologies that he was observing were neither animated nor spawn by demonic forces. There was human agency behind them or, in the words of Marx [ 37 p.462], a ‘master’ who was consciously deploying technology to fulfil specific agendas driven by the will to accumulate capital to the detriment of both laborers and the environment. In essence, the work of Winner and Marx is useful to remember that we can always find someone who is to some extent in control of technological development. Technological innovation is not a process bereft of human intervention. Quite the opposite: it is a human strategy whereby a powerful minority of individuals attempt to get hold of the production of new technology to achieve their own goals. There are plenty of historical examples of such power dynamics as they intersect with and alter technological development, ranging from the development of railways championed by George Stephenson in Victorian England to the mechanization of Soviet agriculture led by Joseph Stalin in the early 1930s [ 54 ]. The problem is that, in the present, we do not see these individuals. We do not see them acting and making choices that shape technological development. What we do see are the technologies that are being produced and the changes that they cause, altering society and the environment at a fast pace.

This problem is connected to the black box problem discussed above. In a way it is an extension of the black box, which is worth unpacking in an attempt to get a glimpse of the big picture. New technologies can be understood as a black box, whether it is a 1980s PC [ 39 ] or a 2020s AI [ 51 ], because their mechanics and functioning remain obscure to their users who ignore their impenetrable operations. However, users also do not see the political economy underpinning the production of new technology, which remains more obscure and inscrutable than the inner mechanics of the technology itself. In other words, most people are unaware of the many political agendas and decisions that set the direction of economic development at different scales (companies, cities, regions and states, for example), which in turn dictate what new technologies will be produced, where and how. Of course, some people do manage to see these intricate politico-economic dynamics, but achieving such awareness requires a considerable effort in terms of research and critical thinking, since this aspect of technological innovation cannot be found at the surface level. A case in point is Marx who was capable of identifying the hand and the mind of the capitalist behind the rapid diffusion of seemingly out-of-control technologies in 19th century England, because he had extensively studied 19th century English political economy. He was fully aware of the big picture. He had penetrated the black box.

More recently, critical social scientists are beginning to shed light on the actors that today are steering the development of AI, and in so doing, triggering significant social and environmental transformations [ 58 , 59 ]. Penetrating the black box means shedding light not simply on the technical aspects of new technologies to understand what makes them function the way they do (whether it is a steam engine in 19th century machines, or software in 1980s PCs, or algorithms in contemporary AIs). It also means exposing the complex political economy that drives the production of new technologies and render visible the network of stakeholders who, guided by human purposes, make choices in an attempt to control the course of technological development and its outcomes. Some of these choices include speed . It would be erroneous to think that technological development proceeds at a faster and faster pace, gaining momentum autonomously like a rock that rolls down a hill and continues moving because of its inertia. The pace of technological development is based on conscious decisions that specific actors make to accelerate its course, in line with pre-determined politico-economic rationales according to which a rapid roll-out of new technologies is expected to accrue certain benefits. The development of new technology might have some degree of inertia but, for the most part, it is tenaciously pushed forward by human hands under the guidance of human logics. Historically, a case in point is that of the Ford Motor Company and its iconic car, the model T, developed in 1908, just five years after the establishment of Ford’s company [ 54 ]. Back then, as Henry Ford himself acknowledged, ‘speed’ was one of the key principles that his company was actively adhering to, implementing ‘high-speed tools’ and perfecting the factory’s assembly line to purposely speed up the production of cars as much as possible [ 60 p.143, 61 p.2170].

In relation to contemporary AI technologies, the phenomenon that has been described so far in conceptual terms, can be also observed in practice across three different scales: country , region and city . The aim here is not to provide an in-depth empirical analysis of how specific actors attempt to take control over AI innovation, but rather to offer an overview of such dynamics and of the logics underpinning them. At the first scale, within many countries, we find national AI strategies. Bareis and Katzenbach [ 62 ], for example, have analyzed the national AI strategies of four key players in the field of AI development, namely China, the United States, Germany and France. In their analysis, they note how ‘the role of the state remains crucial’ as it is in national AI strategies that ‘ideas, announcements and visions start to materialize in projects, infrastructures and organizations’ [ 62 pp.859, 875]. As part of a national AI strategy there are also national research grants through which states finance AI research, thereby boosting AI development, as well as trade sanctions meant to hinder AI development in other states, such as the recent American restrictions on the export of chips to China [ 58 ]. These are important dynamics to highlight because a strategy is a plan of action intended to accomplish specific goals. National AI strategies, therefore, show that AI development is not following its own course independent of human direction. These are evident state-led attempts to capture and steer AI development. In this context, as Bareis and Katzenbach [ 62 p.875] point out, governments explicitly ‘claim agency’ in the production of AI technologies.

At the scale of the region, the dynamics illustrated above become even more evident as the more we zoom in on specific places, the more the agency of specific actors emerges. An emblematic case of regional AI development is Neom in Saudi Arabia. Neom [ 63 ] is a megaproject consisting in the creation of new cities and infrastructures in the north-west of the Arabian Peninsula. The Neom development which includes a new linear city called The Line [ 64 ] has AI as its common denominator, the plan being that all services and infrastructures will be automated by means of algorithms and robotics, and that robots will hold citizenship and coexist with humans in the same urban spaces [ 65 ]. The plan in question is the product of one actor in particular: Mohamed bin Salman (MBS) Crown Prince of the Kingdom of Saudi Arabia and Chairman of Neom. There is a specific rationale behind the production of Neom and its technological apparatus. The power of MBS in Saudi Arabia is growing, but the ambitious Crown Prince is relatively young and has a lot of opponents who resent his autocracy [ 66 ]. In this politically unstable context, MBS is seeking to crystallize his position as the sole leader of the Kingdom by investing in megaprojects like Neom, in a bet to boost both his prestige and economic assets. As Hope and Schek [ 67 ] observe, MBS is purposely accelerating the speed of technological development in order to consolidate his power as soon as possible and take his adversaries by surprise. This is a clear example of how the rapid pace of AI innovation does not depend on the inertia of out-of-control technologies, but on the agency of specific individuals.

Finally, when we look at cities, we can identify a fine-grained network of actors who join forces to steer the development of AI in urban spaces. The work of Zhang, Bates and Abbott [ 68 ], for instance, reveals the variegated groups of stakeholders behind smart-city initiatives in China, whereby multiple AI technologies are developed and integrated into the built environment. Recent smart-city studies focused on AI show how the genesis and diffusion of AI technology are often the product of the cooperation between public stakeholders (city councils, planning departments and city managers) and private stakeholders (tech companies, in particular) that have different stakes and pursue different but compatible goals [ 2 ]. As Lee [ 69 ] explains, Chinese tech companies need to push AI innovation forward as for them innovation is the only way to survive in a very competitive and ruthless market. In the same context, city councils tend to follow regional and national strategies of AI development set by China’s State Council, in a way that problematically excludes citizen engagement and bottom-up inputs from the local population [ 70 ]. Referring back to one of the key points made above, the problem is not that humanity has lost control over AI development, but that only a small percentage of it is controlling the creation and diffusion of AI.

7 Third conceptual issue: the overbelief in computer science and philosophy

In addition, the misconceptions that we have discussed so far create a third misunderstanding: the belief that we can regain control over AI by means of better algorithms that compel AI to act exactly the way we want it to act, according to well-defined ethical principles. However, as we have argued in the previous section, we already have control over AI. We humans do not have to gain (or regain) control over artificial intelligences, because we already have it. The problem is that most of us do not have control over AI, and only a minority of powerful stakeholders are controlling and steering the development of AI, often through procedures and decision-making processes that are undemocratic. Therefore, while there is indeed an alignment problem, it is neither a matter of computer science (designing better algorithms), nor simply a matter of philosophy (formulating better ethical principles). The question is political .

It is no mystery that there is a politics to AI [ 71 , 72 , 73 ]. Contrary to what animistic interpretations of AI might suggest, there is no magic in AI. What we are witnessing is not a magic show, but rather a game of politics that a Marxian politico-economic perspective can help us comprehend. In these terms, the AI ethics industry which has been producing a voluminous corpus of ethical guidelines regarding the creation and deployment of AI technology, is not helping us to overcome the impasse discussed in this section. This is a diverse sector comprising a variety of voices, ranging from international organizations and corporations to business consultancies and independent ethicists [ 74 ]. Yet, apart from few critical voices, there is a problematic common denominator [ 75 ]. By and large, this sector is not opposing the diffusion of AI, proposing instead top-down technical and philosophical solutions which, by targeting the development of ethically sound algorithms and refined ethical codes, fail to engage with citizens who are ultimately marginalized in the politics of AI. A significant portion of contemporary AI ethics is thus part of the same political economy that we have critiqued so far and risks causing ‘ethics shopping’ whereby some ‘stakeholders may be tempted to “shop” for the most appealing’ ethical principles and ignore the issue of citizen engagement to which we now turn [ 75 p.390, 76 p.2].

8 Reframing the alignment problem thesis from a Marxian perspective

In relation to the alignment problem discussed so far, a Marxian perspective is useful to note that much of the labor involved in the production of AI is problematically not involved in steering the course of its development. In this context, labor includes for example the myriad ghost workers who are underpaid to train AIs, as well as the countless citizens who become data points and get their personal information extracted, mostly via social media, through processes of surveillance capitalism that, as Zuboff remarks, ultimately feed AI systems [ 77 , 78 ]. There is thus an evident unethical situation of exploitation at play, since many of the people whose labor and data are used to develop AI, are not contributing to the political agendas and decisions that actually shape AI developments.

In Marxian terms then, reframing the alignment problem means acknowledging this problematic political issue, beyond the already recognized technical and philosophical problems that dominate much of current public discourses. In these terms, AI ethics needs to recognize the fundamental presence of humans in the making of AI systems by integrating a socio-technical perspective, and take into serious consideration the uneven power relations that control their development. In theory, this calls for more participation and extended stakeholder engagement in AI ethics. In practice, we envision public engagement, particularly at the smallest scale examined in this paper i.e. the city, in line with two examples. First is the case of Barcelona where citizens’ opinions are increasingly being included in the local AI-driven platform for urban governance, de facto influencing its application and purpose [ 79 ]. Second is the case of San Francisco where in 2022 citizens protested against weaponized police robots, and managed to stop their deployment in the city [ 80 ]. The first example is a story of democratic political engagement, while the second story is about an agonistic political act carried out by expressing dissensus [ 81 , 82 ].

9 Conclusions: illuminating the obscure politics of AI

As this paper has shown, there are three fundamental misconceptions about the alignment problem as it is often formulated and understood in mainstream public discourses, such as in the case of the FLI open letter. The first and the second one depict a situation in which AI is acting on its own volition and getting out of control. These two misunderstandings are two sides of the same black box issue discussed throughout the paper. We do not understand the complex machine learning techniques and algorithms whereby AI learns about the surrounding environment and act on it and, therefore, many of us resort to animism as a “shortcut” to explain the behavior of hyper complex and unintelligible AI technologies such as ChatGPT. In addition, we do not understand the complex political economy driving the production of AI technology and we do not see the fine-grained network of actors who, across different scales (countries, regions and cities), make decisions that steer the development of AI. In this regard, we have drawn on the examples of national AI strategies by means of which state-actors explicitly attempt to control the production of AI technologies, of the high-tech Neom project of regional development tightly controlled by the Saudi Crown Prince, and of Chinese smart-city initiatives steered by partnerships between public stakeholders and AI companies.

Failing to comprehend these politico-economic dynamics gives us the illusion that AI innovation occurs at a fast pace propelled by an incontrollable momentum, while speed itself is a conscious strategy implemented by human stakeholders in line with human-made agendas. These two misunderstandings give rise to a third misunderstanding, the belief that computer science and philosophy alone can help us solve the alignment problem and, in turn, invalidate much of the current strategies and policies that are being developed worldwide to realign AI development, such as the moratorium proposed by the FLI in its open letter. AI is neither acting on its own volition nor is it getting out of control. In this paper we have debunked the AI-out-of-control discourse and stressed that AI is controlled by a minority of powerful human stakeholders. This makes the alignment problem not a computer science problem or a philosophical issue, but rather an acute political problem.

The politics of AI is still, by and large, an uncharted and obscure territory that needs to be empirically understood in detail across different scales. As states forge national AI strategies, regions develop AI infrastructures, and cities integrate AI technologies into the built environment, it is key that future research seeks to identify who exactly is controlling the production of AI and how. This is the politico-economic side of the black box that we know is there, but that we have not penetrated yet, by digging into the thick layers of decision-making processes whereby AI innovation takes place. In addition, areas of future research should include public participation in the politics of AI in both theory and practice. Empirically, this means examining in more detail cases, such as those of Barcelona and San Francisco, in which citizens are manifesting different forms of political engagement [ 79 , 80 ]. Theoretically, this is about drawing on political theory to theorize alternative participatory politics of AI.

In this regard, Mouffe’s theory of agonistic politics and Rancière’s notion of dissensus can be useful to imagine a politics of AI whereby citizens can opt against some types of AI [ 81 , 82 ]. This would rectify the current imbalance in contemporary AI ethics in which most efforts go into improving the use of AI rather than objecting it in the first place. As this paper has shown from a socio-technical perspective, AI development is not a magic show: it is a game of politics. If we pay attention to the hidden puppeteers, not on the puppet, then we can start realigning the development of AI to common goals, and this could also mean say no to AI and end the game.

Ma, X., Huo, Y.: Are users willing to embrace ChatGPT? Exploring the factors on the acceptance of chatbots from the perspective of AIDUA framework. Tech. Soc. (2023). https://doi.org/10.1016/j.techsoc.2023.102362

Article   Google Scholar  

Cugurullo, F., Caprotti, F., Cook, M., Karvonen, A., McGuirk, P., Marvin, S.: Artificial Intelligence and the City: Urbanistic Perspectives on AI. Routledge, London and New York (2023)

Book   Google Scholar  

Bubeck, S., Chandrasekaran, V., Eldan, R., Gehrke, J., Horvitz, E., Kamar, E., Lee, P.: Sparks of artificial general intelligence: Early experiments with gpt-4. arXiv preprint arXiv:2303.12712 (2023)

Future of Life Institute: Pause Giant AI Experiments: An Open Letter. (2023). https://futureoflife.org/open-letter/pause-giant-ai-experiments/ Accessed 23 March 2024

Future of Life Institute: Policymaking in the Pause. (2023). https://futureoflife.org/wp-content/uploads/2023/04/FLI_Policymaking_In_The_Pause.pdf Accessed 23 March 2024

Fridman, L.: Max Tegmark: The Case for Halting AI Development | Lex Fridman Podcast #371 [2:11:00]. (2023). https://www.youtube.com/watch?v=VcVfceTsD0A&t=7948s Accessed 23 March 2024

Holton, R., Boyd, R.: Where are the people? What are they doing? Why are they doing it?’(Mindell) situating artificial intelligence within a socio-technical framework. J. Sociol. 57 (2), 179–195 (2021)

Venturini, T.: Bruno Latour and Artificial Intelligence. Tecnoscienza–Italian J. Sci. Technol. Stud. 14 (2), 101–114 (2023)

Google Scholar  

Pasquinelli, M.: How a Machine Learns and Fails: A Grammar of Error for Artificial Intelligence. Spheres: Journal for Digital Cultures. 5 : 1–17 (2019)

Cugurullo, F., Caprotti, F., Cook, M., Karvonen, A., MGuirk, P., Marvin, S.: The rise of AI Urbanism in post-smart Cities: A Critical Commentary on Urban Artificial Intelligence, vol. 00420980231203386. Urban Studies (2023)

Palmini, O., Cugurullo, F.: Charting AI urbanism: Conceptual sources and spatial implications of urban artificial intelligence. Discover Artif. Intell. 3 (1), 15 (2023)

Palmini, O., Cugurullo, F.: Design culture for sustainable urban artificial intelligence: Bruno Latour and the search for a different AI urbanism. Ethics Inf. Technol. 26 (1), 11 (2024)

Cugurullo, F., Acheampong, R.A.: Fear of AI: An Inquiry into the Adoption of Autonomous cars in Spite of fear, and a Theoretical Framework for the Study of Artificial Intelligence Technology Acceptance, pp. 1–16. AI & SOCIETY (2023). https://doi.org/10.1007/s00146-022-01598-6

Sartori, L., Bocca, G.: Minding the gap (s): Public perceptions of AI and socio-technical imaginaries. AI Soc. 38 (2), 443–458 (2023)

Jasanoff, S., Kim, S.H.: Dreamscapes of Modernity: Sociotechnical Imaginaries and the Fabrication of Power. University of Chicago Press, Chicago (2015)

Korinek, A., Balwit, A.: Aligned with whom? Direct and social goals for AI systems. National Bureau of Economic Research. (2022). https://www.nber.org/papers/w30017 Accessed 23 March 2024

Dafoe, A.: AI governance: a research agenda. Future of Humanity Institute. (2018). https://www.fhi.ox.ac.uk/wp-content/uploads/GovAI-Agenda.pdf Accessed 23 March 2024

Pasquinelli, M.: The eye of the Master: A Social History of Artificial Intelligence. Verso Books, London (2023)

Ji, J., Qiu, T., Chen, B., Zhang, B., Lou, H., Wang, K., Gao, W.: AI alignment: A comprehensive survey. arXiv Preprint (2023). arXiv:2310.19852

Christian, B.: The Alignment Problem: Machine Learning and Human Values. WW Norton & Company, New York (2020)

Erez, F.: Ought we align the values of artificial moral agents? AI Ethics. (2023). https://doi.org/10.1007/s43681-023-00264-x

Friederich, S.: Symbiosis, not alignment, as the goal for liberal democracies in the transition to artificial general intelligence. AI Ethics. (2023). https://doi.org/10.1007/s43681-023-00268-7

Yudkowsky, E.: Artificial intelligence as a positive and negative factor in global risk. In: Bostrom, N., Cirkovic, M.M. (eds.) Global Catastrophic Risks, pp. 308–345. Oxford University Press, Oxford (2008)

Bostrom, N.: Superintelligence. Paths, Dangers, Strategies. Oxford University Press, Oxford (2017)

Russell, S.: Human Compatible: Artificial Intelligence and the Problem of Control. Penguin, London (2019)

Hendrycks, D., Burns, C., Basart, S., Critch, A., Li, J., Song, D., Steinhardt, J.: Aligning AI with shared human values. arXiv preprint arXiv:02275 (2020). (2008)

Gabriel, I.: Artificial intelligence, values, and alignment. Minds Mach. 30 (3), 411–437 (2020)

McDonald, F.J.: AI, alignment, and the categorical imperative. AI Ethics. 3 (1), 337–344 (2023)

Russell, S.: Human-compatible artificial intelligence. In: Muggleton, S., Chater, N. (eds.) Human-Like Machine Intelligence pp, pp. 3–23. Oxford University Press, Oxford (2021)

Chapter   Google Scholar  

Cugurullo, F.: Good and evil in the Autonomous City. In: Mackinnon, D., Burns, R., Fast, V. (eds.) Digital (in) Justice in the Smart City pp, pp. 183–194. University of Toronto, Toronto (2022)

Tolmeijer, S., Arpatzoglou, V., Rossetto, L., et al.: Trolleys, crashes, and perception—a survey on how current autonomous vehicles debates invoke problematic expectations. AI Ethics. (2023). https://doi.org/10.1007/s43681-023-00284-7

Othman, K.: Understanding how moral decisions are affected by accidents of autonomous vehicles, prior knowledge, and perspective-taking: A continental analysis of a global survey. AI Ethics. (2023). https://doi.org/10.1007/s43681-023-00310-8

Floridi, L.: Artificial intelligence as a public service: Learning from Amsterdam and Helsinki. Philos. Technol. 33 (4), 541–546 (2020)

Cugurullo, F., Caprotti, F., Cook, M., Karvonen, A., McGuirk, P., Marvin, S.: Conclusions: The present of urban AI and the future of cities. In: Cugurullo, F., Caprotti, F., Cook, M., Karvonen, A., McGuirk, P., Marvin, S. (eds.) Artificial Intelligence and the City pp, pp. 361–389. Routledge, London and New York (2023)

Cugurullo, F.: Frankenstein Urbanism: Eco, Smart and Autonomous Cities, Artificial Intelligence and the end of the city. Routledge, London and New York (2021)

Lindgren, S.: Handbook of Critical Studies of Artificial Intelligence. Edward Elgar Publishing, Cheltenham (2023)

Marx, K.: Capital: Volume I. Penguin UK, London (2004)

Fromm, E., Marx, K.: Marx’s Concept of Man: Including ‘Economic and Philosophical Manuscripts’. Bloomsbury Publishing, London (2013)

Stahl, W.A.: Venerating the black box: Magic in media discourse on technology. Science, Technology, & Human Values 20, no. 2: 234–258 (1995)

Hermann, I.: Artificial intelligence in fiction: Between narratives and metaphors. AI Soc. 38 (1), 319–329 (2023)

Turkle, S.: The Second Self: Computers and the Human Spirit. MIT Press, Cambridge (1984)

Strengers, Y., Kennedy, J.: The Smart wife: Why Siri, Alexa, and Other Smart home Devices need a Feminist Reboot. MIT Press, Cambridge (2021)

Marenko, B.: Neo-animism and design: A new paradigm in object theory. Des. Cult. 6 (2), 219–241 (2014)

Marenko, B., Van Allen, P.: Animistic design: How to reimagine digital interaction between the human and the nonhuman. Digit. Creativity. 27 (1), 52–70 (2016)

Natale, S., Henrickson, L.: The Lovelace effect: Perceptions of creativity in machines. New. Media Soc. 14614448221077278. (2022)

Bowman, S.R.: Eight Things to Know about Large Language Models. (2023). https://doi.org/10.48550/arXiv.2304.00612

Greenfield, A.: Radical Technologies: The Design of Everyday life. Verso Books, London (2017)

Bewersdorff, A., Zhai, X., Roberts, J., Nerdel, C.: Myths, mis-and preconceptions of artificial intelligence: A review of the literature. Computers Education: Artif. Intell. 100143 (2023). https://doi.org/10.1016/j.caeai.2023.100143

de Fine Licht, K., de Fine Licht, J.: Artificial intelligence, transparency, and public decision-making: Why explanations are key when trying to produce perceived legitimacy. AI Soc. 35 , 917–926 (2020)

Rai, A.: Explainable AI: From black box to glass box. J. Acad. Mark. Sci. 48 , 137–141 (2020)

Zednik, C.: Solving the black box problem: A normative framework for explainable artificial intelligence. Philos. Technol. 34 (2), 265–288 (2021)

Nyholm, S.: A new control problem? Humanoid robots, artificial intelligence, and the value of control. AI Ethics. (2022). https://doi.org/10.1007/s43681-022-00231-y

Winner, L.: Autonomous Technology: Technics-out-of-control as a Theme in Political Thought. MIT Press, Cambridge (1977)

Johnson, S., Acemoglu, D.: Power and Progress: Our Thousand-Year Struggle Over Technology and Prosperity. Basic Books, London (2023)

Benevolo, L.: The European City. Blackwell, Oxford (1993)

Hall, P.: Cities of Tomorrow: An Intellectual History of Urban Planning and Design since 1880. Blackwell, Oxford (2002)

Finer, S.E.: The life and Times of Sir Edwin Chadwick. Routledge, London and New York (2016)

Rella, L.: Close to the metal: Towards a material political economy of the epistemology of computation. Soc. Stud. Sci. 54 (1), 3–29 (2024)

van der Vlist, F., Helmond, A., Ferrari, F.: Big AI: Cloud infrastructure dependence and the industrialisation of artificial intelligence. Big Data Soc. 11 (1) (2024). https://doi.org/10.1177/20539517241232630

Nye, D.E.: Consuming Power: A Social History of American Energies. MIT Press, Cambridge (1999)

Hounshell, D.: From the American System to mass Production, 1800–1932: The Development of Manufacturing Technology in the United States (No. 4). Johns Hopkins University, Baltimore (1984)

Bareis, J., Katzenbach, C.: Talking AI into being: The narratives and imaginaries of national AI strategies and their performative politics. Sci. Technol. Hum. Values. 47 (5), 855–881 (2022)

Neom: What is Neom? (2023). https://www.neom.com/en-us Accessed 23 March 2024

Batty, M.: The Linear City: Illustrating the logic of spatial equilibrium. Comput. Urban Sci. 2 (1), 8 (2022)

Article   MathSciNet   Google Scholar  

Parviainen, J., Coeckelbergh, C.: The political choreography of the Sophia robot: Beyond robot rights and citizenship to political performances for the social robotics market. AI Soc. 36 , 3: 715–724 (2021)

Davidson, C.M.: Mohammed Bin Salman Al Saud: King in all but name. In: Larres, K. (ed.) Dictators and Autocrats: Securing Power Across Global Politics pp, pp. 320–345. Routledge, London PP (2022)

Hope, B., Scheck, J.: Blood and Oil: Mohammed Bin Salman’s Ruthless Quest for Global Power. Hachette UK, London (2020)

Zhang, J., Bates, J., Abbott, P.: State-steered smartmentality in Chinese smart urbanism. Urban Studies 59 , 14: 2933–2950. (2022)

Lee, K.: AI Superpowers: China, Silicon Valley, and the new World Order. Houghton Mifflin, Boston and New York (2018)

Xu, Y., Cugurullo, F., Zhang, H., Gaio, A., Zhang, W.: The Emergence of Artificial Intelligence in Anticipatory Urban Governance: Multi-scalar evidence of China’s transition to City brains. J. Urban Technol. 1–25 (2024). https://doi.org/10.1080/10630732.2023.2292823

Coeckelbergh, M.: The Political Philosophy of AI: An Introduction. Polity, Cambridge (2022)

Crawford, K.: Atlas of AI: Power, Politics, and the Planetary Costs of Artificial Intelligence. Yale University Press, New Haven (2021)

Kalluri, P.: Don’t ask if artificial intelligence is good or fair, ask how it shifts power. Nature. 583 , 7815: 169–169 (2020)

Jobin, A., Ienca, M., Vayena, E.: The global landscape of AI ethics guidelines. Nat. Mach. Intell. 1 (9), 389–399 (2019)

Attard-Frost, B., De los Ríos, A., Walters, D.R.: The ethics of AI business practices: A review of 47 AI ethics guidelines. AI Ethics. 3 (2), 389–406 (2023)

Floridi, L., Cowls, J.: A unified framework of five principles for AI in society. Harv. Data Sci. Rev. (2019). https://hdsr.mitpress.mit.edu/pub/l0jsh9d1/release/8

Gabrys, J.: Programming environments: Environmentality and citizen sensing in the smart city. Environ. Plann. D: Soc. Space. 32 (1), 30–48 (2014)

Zuboff, S.: The age of Surveillance Capitalism: The Fight for a Human Future at the new Frontier of Power. Profile books, London (2019)

Cardullo, P., Ribera-Fumaz, R., González Gil, P.: The Decidim ‘soft infrastructure’: democratic platforms and technological autonomy in Barcelona. Computational Culture, (9). (2023). http://computationalculture.net/the-decidim-soft-infrastructure/ Accessed 23 March 2024

Blanchard, A.: Autonomous force beyond armed conflict. Mind. Mach. 33 (1), 251–260 (2023)

Mouffe, C.: Agonistics: Thinking the World Politically. Verso Books, London (2013)

Rancière, J.: Disagreement: Politics and Philosophy. University of Minnesota, Minneapolis (1999)

Download references

Acknowledgements

Intellectually this paper benefitted enormously from the comments of four anonymous reviewers, and financially from the support of the Irish Research Council.

Open Access funding provided by the IReL Consortium. https://doi.org/10.13039/501100002081 Irish Research Council, IRCLA/2022/3832. Prof. Federico Cugurullo https://orcid.org/0000-0002-0625-8868 .

Open Access funding provided by the IReL Consortium

Author information

Authors and affiliations.

Department of Geography, Museum Building, Trinity College Dublin, Dublin 2, Ireland

Federico Cugurullo

You can also search for this author in PubMed   Google Scholar

Corresponding author

Correspondence to Federico Cugurullo .

Ethics declarations

Conflict of interest.

The corresponding author states that there is no conflict of interest.

Additional information

Publisher’s note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ .

Reprints and permissions

About this article

Cugurullo, F. The obscure politics of artificial intelligence: a Marxian socio-technical critique of the AI alignment problem thesis. AI Ethics (2024). https://doi.org/10.1007/s43681-024-00476-9

Download citation

Received : 29 October 2023

Accepted : 29 March 2024

Published : 08 April 2024

DOI : https://doi.org/10.1007/s43681-024-00476-9

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Artificial intelligence
• Alignment problem
• Autonomous technology
• Political theory
• Marxian political economy
• Socio-technical studies
• Find a journal
• Publish with us
• Track your research

How Tech Giants Cut Corners to Harvest Data for A.I.

OpenAI, Google and Meta ignored corporate policies, altered their own rules and discussed skirting copyright law as they sought online information to train their newest artificial intelligence systems.

Researchers at OpenAI’s office in San Francisco developed a tool to transcribe YouTube videos to amass conversational text for A.I. development. Credit... Jason Henry for The New York Times

Supported by

• Share full article

By Cade Metz ,  Cecilia Kang ,  Sheera Frenkel ,  Stuart A. Thompson and Nico Grant

Reporting from San Francisco, Washington and New York

• Published April 6, 2024 Updated April 8, 2024

In late 2021, OpenAI faced a supply problem.

The artificial intelligence lab had exhausted every reservoir of reputable English-language text on the internet as it developed its latest A.I. system. It needed more data to train the next version of its technology — lots more.

Listen to this article with reporter commentary

Open this article in the New York Times Audio app on iOS.

So OpenAI researchers created a speech recognition tool called Whisper. It could transcribe the audio from YouTube videos, yielding new conversational text that would make an A.I. system smarter.

Some OpenAI employees discussed how such a move might go against YouTube’s rules, three people with knowledge of the conversations said. YouTube, which is owned by Google, prohibits use of its videos for applications that are “independent” of the video platform.

Ultimately, an OpenAI team transcribed more than one million hours of YouTube videos, the people said. The team included Greg Brockman, OpenAI’s president, who personally helped collect the videos, two of the people said. The texts were then fed into a system called GPT-4 , which was widely considered one of the world’s most powerful A.I. models and was the basis of the latest version of the ChatGPT chatbot.

The race to lead A.I. has become a desperate hunt for the digital data needed to advance the technology. To obtain that data, tech companies including OpenAI, Google and Meta have cut corners, ignored corporate policies and debated bending the law, according to an examination by The New York Times.

At Meta, which owns Facebook and Instagram, managers, lawyers and engineers last year discussed buying the publishing house Simon & Schuster to procure long works, according to recordings of internal meetings obtained by The Times. They also conferred on gathering copyrighted data from across the internet, even if that meant facing lawsuits. Negotiating licenses with publishers, artists, musicians and the news industry would take too long, they said.

Like OpenAI, Google transcribed YouTube videos to harvest text for its A.I. models, five people with knowledge of the company’s practices said. That potentially violated the copyrights to the videos, which belong to their creators.

Last year, Google also broadened its terms of service. One motivation for the change, according to members of the company’s privacy team and an internal message viewed by The Times, was to allow Google to be able to tap publicly available Google Docs, restaurant reviews on Google Maps and other online material for more of its A.I. products.

The companies’ actions illustrate how online information — news stories, fictional works, message board posts, Wikipedia articles, computer programs, photos, podcasts and movie clips — has increasingly become the lifeblood of the booming A.I. industry. Creating innovative systems depends on having enough data to teach the technologies to instantly produce text, images, sounds and videos that resemble what a human creates.

The volume of data is crucial. Leading chatbot systems have learned from pools of digital text spanning as many as three trillion words, or roughly twice the number of words stored in Oxford University’s Bodleian Library, which has collected manuscripts since 1602. The most prized data, A.I. researchers said, is high-quality information, such as published books and articles, which have been carefully written and edited by professionals.

For years, the internet — with sites like Wikipedia and Reddit — was a seemingly endless source of data. But as A.I. advanced, tech companies sought more repositories. Google and Meta, which have billions of users who produce search queries and social media posts every day, were largely limited by privacy laws and their own policies from drawing on much of that content for A.I.

Their situation is urgent. Tech companies could run through the high-quality data on the internet as soon as 2026, according to Epoch, a research institute. The companies are using the data faster than it is being produced.

“The only practical way for these tools to exist is if they can be trained on massive amounts of data without having to license that data,” Sy Damle, a lawyer who represents Andreessen Horowitz, a Silicon Valley venture capital firm, said of A.I. models last year in a public discussion about copyright law. “The data needed is so massive that even collective licensing really can’t work.”

Tech companies are so hungry for new data that some are developing “synthetic” information. This is not organic data created by humans, but text, images and code that A.I. models produce — in other words, the systems learn from what they themselves generate.

OpenAI said each of its A.I. models “has a unique data set that we curate to help their understanding of the world and remain globally competitive in research.” Google said that its A.I. models “are trained on some YouTube content,” which was allowed under agreements with YouTube creators, and that the company did not use data from office apps outside of an experimental program. Meta said it had “made aggressive investments” to integrate A.I. into its services and had billions of publicly shared images and videos from Instagram and Facebook for training its models.

For creators, the growing use of their works by A.I. companies has prompted lawsuits over copyright and licensing. The Times sued OpenAI and Microsoft last year for using copyrighted news articles without permission to train A.I. chatbots. OpenAI and Microsoft have said using the articles was “fair use,” or allowed under copyright law, because they transformed the works for a different purpose.

More than 10,000 trade groups, authors, companies and others submitted comments last year about the use of creative works by A.I. models to the Copyright Office , a federal agency that is preparing guidance on how copyright law applies in the A.I. era.

Justine Bateman, a filmmaker, former actress and author of two books, told the Copyright Office that A.I. models were taking content — including her writing and films — without permission or payment.

“This is the largest theft in the United States, period,” she said in an interview.

‘Scale Is All You Need’

In January 2020, Jared Kaplan, a theoretical physicist at Johns Hopkins University, published a groundbreaking paper on A.I. that stoked the appetite for online data.

His conclusion was unequivocal: The more data there was to train a large language model — the technology that drives online chatbots — the better it would perform. Just as a student learns more by reading more books, large language models can better pinpoint patterns in text and be more accurate with more information.

“Everyone was very surprised that these trends — these scaling laws as we call them — were basically as precise as what you see in astronomy or physics,” said Dr. Kaplan, who published the paper with nine OpenAI researchers. (He now works at the A.I. start-up Anthropic.)

“Scale is all you need” soon became a rallying cry for A.I.

Researchers have long used large public databases of digital information to develop A.I., including Wikipedia and Common Crawl, a database of more than 250 billion web pages collected since 2007. Researchers often “cleaned” the data by removing hate speech and other unwanted text before using it to train A.I. models.

In 2020, data sets were tiny by today’s standards. One database containing 30,000 photographs from the photo website Flickr was considered a vital resource at the time.

After Dr. Kaplan’s paper, that amount of data was no longer enough. It became all about “just making things really big,” said Brandon Duderstadt, the chief executive of Nomic, an A.I. company in New York.

Before 2020, most A.I. models used relatively little training data.

Mr. Kaplan’s paper, released in 2020, led to a new era defined by GPT-3, a large language model, where researchers began including more data in their models …

… much, much more data.

When OpenAI unveiled GPT-3 in November 2020, it was trained on the largest amount of data to date — about 300 billion “tokens,” which are essentially words or pieces of words. After learning from that data, the system generated text with astounding accuracy, writing blog posts, poetry and its own computer programs.

In 2022, DeepMind, an A.I. lab owned by Google, went further. It tested 400 A.I. models and varied the amount of training data and other factors. The top-performing models used even more data than Dr. Kaplan had predicted in his paper. One model, Chinchilla, was trained on 1.4 trillion tokens.

It was soon overtaken. Last year, researchers from China released an A.I. model, Skywork , which was trained on 3.2 trillion tokens from English and Chinese texts. Google also unveiled an A.I. system, PaLM 2 , which topped 3.6 trillion tokens .

Transcribing YouTube

In May, Sam Altman , the chief executive of OpenAI, acknowledged that A.I. companies would use up all viable data on the internet.

“That will run out,” he said in a speech at a tech conference.

Mr. Altman had seen the phenomenon up close. At OpenAI, researchers had gathered data for years, cleaned it and fed it into a vast pool of text to train the company’s language models. They had mined the computer code repository GitHub, vacuumed up databases of chess moves and drawn on data describing high school tests and homework assignments from the website Quizlet.

By late 2021, those supplies were depleted, said eight people with knowledge of the company, who were not authorized to speak publicly.

OpenAI was desperate for more data to develop its next-generation A.I. model, GPT-4. So employees discussed transcribing podcasts, audiobooks and YouTube videos, the people said. They talked about creating data from scratch with A.I. systems. They also considered buying start-ups that had collected large amounts of digital data.

OpenAI eventually made Whisper, the speech recognition tool, to transcribe YouTube videos and podcasts, six people said. But YouTube prohibits people from not only using its videos for “independent” applications, but also accessing its videos by “any automated means (such as robots, botnets or scrapers).”

OpenAI employees knew they were wading into a legal gray area, the people said, but believed that training A.I. with the videos was fair use. Mr. Brockman, OpenAI’s president, was listed in a research paper as a creator of Whisper. He personally helped gather YouTube videos and fed them into the technology, two people said.

Mr. Brockman referred requests for comment to OpenAI, which said it uses “numerous sources” of data.

Last year, OpenAI released GPT-4, which drew on the more than one million hours of YouTube videos that Whisper had transcribed. Mr. Brockman led the team that developed GPT-4.

Some Google employees were aware that OpenAI had harvested YouTube videos for data, two people with knowledge of the companies said. But they didn’t stop OpenAI because Google had also used transcripts of YouTube videos to train its A.I. models, the people said. That practice may have violated the copyrights of YouTube creators. So if Google made a fuss about OpenAI, there might be a public outcry against its own methods, the people said.

Matt Bryant, a Google spokesman, said the company had no knowledge of OpenAI’s practices and prohibited “unauthorized scraping or downloading of YouTube content.” Google takes action when it has a clear legal or technical basis to do so, he said.

Google’s rules allowed it to tap YouTube user data to develop new features for the video platform. But it was unclear whether Google could use YouTube data to build a commercial service beyond the video platform, such as a chatbot.

Geoffrey Lottenberg, an intellectual property lawyer with the law firm Berger Singerman, said Google’s language about what it could and could not do with YouTube video transcripts was vague.

“Whether the data could be used for a new commercial service is open to interpretation and could be litigated,” he said.

In late 2022, after OpenAI released ChatGPT and set off an industrywide race to catch up , Google researchers and engineers discussed tapping other user data. Billions of words sat in people’s Google Docs and other free Google apps. But the company’s privacy restrictions limited how they could use the data, three people with knowledge of Google’s practices said.

In June, Google’s legal department asked the privacy team to draft language to broaden what the company could use consumer data for, according to two members of the privacy team and an internal message viewed by The Times.

The employees were told Google wanted to use people’s publicly available content in Google Docs, Google Sheets and related apps for an array of A.I. products. The employees said they didn’t know if the company had previously trained A.I. on such data.

At the time, Google’s privacy policy said the company could use publicly available information only to “help train Google’s language models and build features like Google Translate.”

The privacy team wrote new terms so Google could tap the data for its “A.I. models and build products and features like Google Translate, Bard and Cloud AI capabilities,” which was a wider collection of A.I. technologies.

“What is the end goal here?” one member of the privacy team asked in an internal message. “How broad are we going?”

The team was told specifically to release the new terms on the Fourth of July weekend, when people were typically focused on the holiday, the employees said. The revised policy debuted on July 1, at the start of the long weekend.

How Google Can Use Your Data

Here are the changes Google made to its privacy policy last year for its free consumer apps.

Google uses information to improve our services and to develop new products, features and technologies that benefit our users and the public. For example, we use publicly available information to help train Google’s language AI models and build products and features like Google Translate , Bard, and Cloud AI capabilities .

In August, two privacy team members said, they pressed managers on whether Google could start using data from free consumer versions of Google Docs, Google Sheets and Google Slides. They were not given clear answers, they said.

Mr. Bryant said that the privacy policy changes had been made for clarity and that Google did not use information from Google Docs or related apps to train language models “without explicit permission” from users, referring to a voluntary program that allows users to test experimental features.

“We did not start training on additional types of data based on this language change,” he said.

The Debate at Meta

Mark Zuckerberg, Meta’s chief executive, had invested in A.I. for years — but suddenly found himself behind when OpenAI released ChatGPT in 2022. He immediately pushed to match and exceed ChatGPT , calling executives and engineers at all hours of the night to push them to develop a rival chatbot, said three current and former employees, who were not authorized to discuss confidential conversations.

But by early last year, Meta had hit the same hurdle as its rivals: not enough data.

Ahmad Al-Dahle, Meta’s vice president of generative A.I., told executives that his team had used almost every available English-language book, essay, poem and news article on the internet to develop a model, according to recordings of internal meetings, which were shared by an employee.

Meta could not match ChatGPT unless it got more data, Mr. Al-Dahle told colleagues. In March and April 2023, some of the company’s business development leaders, engineers and lawyers met nearly daily to tackle the problem.

Some debated paying $10 a book for the full licensing rights to new titles. They discussed buying Simon & Schuster, which publishes authors like Stephen King, according to the recordings.

They also talked about how they had summarized books, essays and other works from the internet without permission and discussed sucking up more, even if that meant facing lawsuits. One lawyer warned of “ethical” concerns around taking intellectual property from artists but was met with silence, according to the recordings.

Mr. Zuckerberg demanded a solution, employees said.

“The capability that Mark is looking for in the product is just something that we currently aren’t able to deliver,” one engineer said.

While Meta operates giant social networks, it didn’t have troves of user posts at its disposal, two employees said. Many Facebook users had deleted their earlier posts, and the platform wasn’t where people wrote essay-type content, they said.

Meta was also limited by privacy changes it introduced after a 2018 scandal over sharing its users’ data with Cambridge Analytica, a voter-profiling company.

Mr. Zuckerberg said in a recent investor call that the billions of publicly shared videos and photos on Facebook and Instagram are “greater than the Common Crawl data set.”

During their recorded discussions, Meta executives talked about how they had hired contractors in Africa to aggregate summaries of fiction and nonfiction. The summaries included copyrighted content “because we have no way of not collecting that,” a manager said in one meeting.

Meta’s executives said OpenAI seemed to have used copyrighted material without permission. It would take Meta too long to negotiate licenses with publishers, artists, musicians and the news industry, they said, according to the recordings.

“The only thing that’s holding us back from being as good as ChatGPT is literally just data volume,” Nick Grudin, a vice president of global partnership and content, said in one meeting.

OpenAI appeared to be taking copyrighted material and Meta could follow this “market precedent,” he added.

Meta’s executives agreed to lean on a 2015 court decision involving the Authors Guild versus Google , according to the recordings. In that case, Google was permitted to scan, digitize and catalog books in an online database after arguing that it had reproduced only snippets of the works online and had transformed the originals, which made it fair use.

Using data to train A.I. systems, Meta’s lawyers said in their meetings, should similarly be fair use.

At least two employees raised concerns about using intellectual property and not paying authors and other artists fairly or at all, according to the recordings. One employee recounted a separate discussion about copyrighted data with senior executives including Chris Cox, Meta’s chief product officer, and said no one in that meeting considered the ethics of using people’s creative works.

‘Synthetic’ Data

OpenAI’s Mr. Altman had a plan to deal with the looming data shortage.

Companies like his, he said at the May conference, would eventually train their A.I. on text generated by A.I. — otherwise known as synthetic data.

Since an A.I. model can produce humanlike text, Mr. Altman and others have argued, the systems can create additional data to develop better versions of themselves. This would help developers build increasingly powerful technology and reduce their dependence on copyrighted data.

“As long as you can get over the synthetic data event horizon, where the model is smart enough to make good synthetic data, everything will be fine,” Mr. Altman said.

A.I. researchers have explored synthetic data for years. But building an A.I system that can train itself is easier said than done. A.I. models that learn from their own outputs can get caught in a loop where they reinforce their own quirks, mistakes and limitations.

“The data these systems need is like a path through the jungle,” said Jeff Clune, a former OpenAI researcher who now teaches computer science at the University of British Columbia. “If they only train on synthetic data, they can get lost in the jungle.”

To combat this, OpenAI and others are investigating how two different A.I. models might work together to generate synthetic data that is more useful and reliable. One system produces the data, while a second judges the information to separate the good from the bad. Researchers are divided on whether this method will work.

A.I. executives are barreling ahead nonetheless.

“It should be all right,” Mr. Altman said at the conference.

Read by Cade Metz

Audio produced by Patricia Sulbarán .

An earlier version of this article misstated the publisher of J.K. Rowling’s books. Her works have been published by Scholastic, Little, Brown and others. They were not published by Simon & Schuster.

How we handle corrections

Cade Metz writes about artificial intelligence, driverless cars, robotics, virtual reality and other emerging areas of technology. More about Cade Metz

Cecilia Kang reports on technology and regulatory policy and is based in Washington D.C. She has written about technology for over two decades. More about Cecilia Kang

Sheera Frenkel is a reporter based in the San Francisco Bay Area, covering the ways technology impacts everyday lives with a focus on social media companies, including Facebook, Instagram, Twitter, TikTok, YouTube, Telegram and WhatsApp. More about Sheera Frenkel

Stuart A. Thompson writes about how false and misleading information spreads online and how it affects people around the world. He focuses on misinformation, disinformation and other misleading content. More about Stuart A. Thompson

Nico Grant is a technology reporter covering Google from San Francisco. Previously, he spent five years at Bloomberg News, where he focused on Google and cloud computing. More about Nico Grant

Explore Our Coverage of Artificial Intelligence

News  and Analysis

U.S. clinics are starting to offer patients a new service: having their mammograms read not just by a radiologist, but also by an A.I. model .

OpenAI unveiled Voice Engine , an A.I. technology that can recreate a person’s voice from a 15-second recording.

Amazon said it had added $2.75 billion to its investment in Anthropic , an A.I. start-up that competes with companies like OpenAI and Google.

The Age of A.I.

A.I. tools can replace much of Wall Street’s entry-level white-collar work , raising tough questions about the future of finance.

The boom in A.I. technology has put a more sophisticated spin on a kind of gig work that doesn’t require leaving the house: training A.I, models .

Teen girls are confronting an epidemic of deepfake nudes in schools  across the United States, as middle and high school students have used A.I. to fabricate explicit images of female classmates.

A.I. is peering into restaurant garbage pails  and crunching grocery-store data to try to figure out how to send less uneaten food into dumpsters.

David Autor, an M.I.T. economist and tech skeptic, argues that A.I. is fundamentally different  from past waves of computerization.

Economists doubt that A.I. is already visible in productivity data . Big companies, however, talk often about adopting it to improve efficiency.

Advertisement

• Election 2024
• Entertainment
• Newsletters
• Photography
• Personal Finance
• AP Investigations
• AP Buyline Personal Finance
• Press Releases
• Israel-Hamas War
• Russia-Ukraine War
• Global elections
• Asia Pacific
• Latin America
• Middle East
• Election Results
• Delegate Tracker
• AP & Elections
• March Madness
• AP Top 25 Poll
• Movie reviews
• Book reviews
• Personal finance
• Financial Markets
• Business Highlights
• Financial wellness
• Artificial Intelligence
• Social Media

Brazil Supreme Court strikes down military intervention thesis in symbolic vote for democracy

FILE - A supporter of Brazilian President Jair Bolsonaro salutes while singing the nation’s anthem outside a military base during a protest against his reelection defeat in Sao Paulo, Brazil, Nov. 3, 2022. Brazil’s Supreme Court unanimously voted Monday, April 8, 2024, that the armed forces have no constitutional power to intervene in disputes between government branches, marking a largely symbolic decision aimed at bolstering democracy after years of increasing threat of military intervention.(AP Photo/Matias Delacroix, File)

FILE - Brazilian President Jair Bolsonaro, center, and his Defense Minister Walter Braga Netto, second from right, watch a military convoy pass Planalto presidential palace, alongside military officials in Brasilia, Brazil, Tuesday, Aug. 10, 2021. Brazil’s Supreme Court unanimously voted Monday, April 8, 2024, that the armed forces have no constitutional power to intervene in disputes between government branches, marking a largely symbolic decision aimed at bolstering democracy after years of increasing threat of military intervention. (AP Photo/Eraldo Peres, File)

FILE - Protesters, supporters of Brazil’s former President Jair Bolsonaro, storm the Supreme Court building in Brasilia, Brazil, Jan. 8, 2023. Brazil’s Supreme Court unanimously voted Monday, April 8, 2024, that the armed forces have no constitutional power to intervene in disputes between government branches, marking a largely symbolic decision aimed at bolstering democracy after years of increasing threat of military intervention.(AP Photo/Eraldo Peres, File)

• Copy Link copied

SAO PAULO (AP) — Brazil’s Supreme Court unanimously voted Monday that the armed forces have no constitutional power to intervene in disputes between government branches, a largely symbolic decision aimed at bolstering democracy after years of increasing threat of military intervention.

The court’s decision came in response to an argument that right-wing former President Jair Bolsonaro and his allies deployed in recent years. They have claimed that Article 142 of Brazil’s Constitution affords the military so-called “moderating power” between the executive, legislative and judicial branches.

Bolsonaro presented this interpretation in an April 2020 meeting with his ministers, telling them that any of the three powers can request the armed forces take action to restore order in Brazil. In the years since, posters invoking Article 142 became a fixture at rallies calling for military takeover – and culminated in an uprising by Bolsonaro supporters seeking to summon the military to oust his successor from power.

All of the 11 justices — including both justices appointed by Bolsonaro — rejected that thesis.

FILE - Brazilian President Jair Bolsonaro, center, and his Defense Minister Walter Braga Netto, second from right, watch a military convoy pass Planalto presidential palace, alongside military officials in Brasilia, Brazil, Tuesday, Aug. 10, 2021. (AP Photo/Eraldo Peres, File)

While the constitution empowers the military to protect the nation from threats and guarantee constitutional powers, “that does not comport with any interpretation that allows the use of the armed forces for the defense of one power against the other,” the case’s rapporteur, Justice Luiz Fux, wrote in his vote.

Article 142’s vague wording had allowed room for some interpretation — although the one espoused by Brazil’s far right was “absolutely crazy,” said João Gabriel Pontes, a constitutional lawyer at Daniel Sarmento e Ademar Borges in Rio de Janeiro.

“This is not a Supreme Court ruling that will safeguard Brazilian democracy from new attacks,” Pontes said by phone. “However, it sends an important message to society that a military intervention has no constitutional basis.”

The constitution dates from 1988, three years after the country cast off its 21-year military dictatorship.

Bolsonaro’s 2018 election in a sense marked the return of the armed forces to power. The former army captain who openly waxed nostalgic for the dictatorship era appointed high-ranking officers to his Cabinet and thousands of active-duty service members and reservists to civilian positions throughout his administration.

For his 2022 reelection bid, he tapped a general as his running mate and tasked the military with auditing electronic voting machines whose reliability he cast doubt upon, without ever providing evidence. Following his defeat to leftist rival Luiz Inácio Lula da Silva , his supporters set up camp outside military barracks for months to demand military intervention.

Bolsonaro never conceded defeat nor asked them to demobilize, and on Jan. 8, 2023 they stormed the capital , Brasilia, invading and vandalizing the Supreme Court, Congress and the presidential palace.

Federal Police later confiscated the cell phone of Bolsonaro’s aide-de-camp and found conversations between close advisers and military officials debating whether conditions and the constitution allowed for military intervention. The seizure was part of an investigation into whether the former president and top aides incited the uprising to restore him to power. He has denied any involvement.

Debate over the constitutional role of the armed forces reflects “the historic vice of an institution that never conformed to subordinating itself to civil order,” and the court’s vote reaffirms what is clear from any constitutional law textbook, said Conrado Hubner, a professor of constitutional law at the University of Sao Paulo.

“Nothing has the power to avoid a coup in the future. Nothing,” Hubner said. But the court’s position helps to combat justifications for a coup, he said.

Meantime, Lula has endeavored to stay on good terms with the military’s top brass. Last week, he forbade any official events observing the 60th anniversary of the date the military deposed the president and ushered in Brazil’s dictatorship, on March 31, 1964.

Virtually all historians characterize it as a coup. Others disagree, including Bolsonaro’s then-vice president, Gen. Hamilton Mourão, who wrote Sunday on X that the date represents the day “the nation saved itself from itself!” and that history cannot be rewritten.

In his vote, Justice Flávio Dino wrote that “echoes of that past stubbornly refuse to pass,” and that the court’s decision should be forwarded to Lula’s defense minister for dissemination to every military organization in the country.

Doing so “would aim to eradicate misinformation that has reached some members of the armed forces,” Dino wrote. “Any theories that go beyond or distort the true meaning of Article 142 of the federal constitution must be eliminated.”

Follow AP’s coverage of Latin America and the Caribbean at https://apnews.com/hub/latin-america