Center for Analysis and Design of Intelligent Agents

Readings README (Do not skip!)

What is intelligence?
How is intelligence realized in the physical world?
What uniquely separates the phenomenon of intelligence from other similar phenomena?

• Intelligent Agents on Wikipedia
• Cognitive Architecture on Wikipedia
• A Collection of Definitions of Intelligence by Legg & Hutter
• No, The Universe Isn't Made of Pure Mathematics by E. Siegel
• Wild Elephants Give Names to Each Other on NPR

What is AI?
How is AI used in the physical world?
What is required to make the use of AI ethical?

• AI Could Help Us Talk to Animals—But Should It? by B. Warner
• On Defining Artificial Intelligence by P. Wang
• AI risk repository at MIT
• Computing Machinery and Intelligence by A. M. Turing
• Artificial Intelligence: examples of ethical dilemmas by UNESCO
• Limits of Transformers on Compositionality by Dziri et al.

No accepted theory of meaning and understanding exists
Systematic action requires knowledge of cause-effect relations
Reasoning in the physical world can only be non-axiomatic

• Reasoning on Wikipedia
• The limitation of Bayesiansm by P. Wang
• A Theory of Foundational Meaning Generation, Natural & Artificial by K. R. Thórisson & G. Talevi
• ChatGPT Can't Reason by K. Arkoudas
• Adequate Determinism on the Information Philosopher
• Bayesianism and Causality, or, Why I am Only a Half-Beyesian by J. Pearl

Transparency means explanation
Explanation means cause-effect relations
AI systems that can represent cause-effect relations are few

• Explicit Goal-Driven Autonomous Self-Explanation Generation by K.R.Thorisson & G. Talevi
• The 'Explanation Hypothesis' in General Self-Supervised Learning by K.R.Thórisson
• About Understanding by K. R. Thórisson et al.
• Big tech fails transparency test by Gary Marcus on The Big Think
• No One Should Trust AI by J. Bryson

• Responsible Artificial Intelligence – from Principles to Practice by V. Dignum
• The Uselessness of AI Ethics by L. Munn
• Responsibility on The Information Philosopher
• The global landscape of AI ethics guidelines by Jobin, et al.
• Philosophy of Technology on Stanford Encyclopedia of Philosophy
• Ethics of Artificial Intelligence and Robotics on Stanford Encyclopedia of Philosophy

• The Future of AI Research by K. R. Thórisson and H. Minsky
• Do Machines Make History? by R. L. Heilbroner
• Printing and the Growth of a Protestant Movement in Germany from 1517 to 1524 by L. W. Holborn
• https://en.wikipedia.org/wiki/Futures_studiesFutures Studies on Wikipedia.

• Democracy & Artificial General Intelligence by E. Contio & J. Salmi
• Iceland & The Fourth Industrial Revolution by H. F. Thorsteinsson et al.
• Ethical Issues of AI by Bernd Carsten Stahl: Chapters 3, 4 (except 4.5 Metaphysical Issues) and 5
• Example of ethical review panel principles at EUPATI Open Classroom
• Why Are There Still So Many Jobs? The History and Future of Workplace Automation by D. H. Autor
• unified framework of five principles for AI in society by L. Floridi and J. Cowls.
• AI in EU-funded projects on Privanova
• Understanding Ethical and Legal Obligations in a Pandemic: A Taxonomy of “Duty” for Health Practitioners by Sheahan & Lamont on the NIH Website.

• Self-Consuming Generative Models Go MAD by S. Alemohammad et al.
• Gary Marcus' commentary blog on 60 Minutes interview with Geoffrey Hinton by G. Marcus
• Beyesian Networks by J. Pearl
• The Human Use of Human Beings by N. Wiener (1950) – possibly the world's first treatise on AI & ethics. “The 'mechanical brain' and similar machines can destroy human values or enable us to realize them as never before.”
• Claude Shannon and Norbert Wiener (cybernetics) by B. Collins.
• Claude Shannon demonstrates machine learning in 1952 on YouTube

• The Singularity Is Coming Soon. Here’s What It May Mean. in Forbes magazine

How to Use This Page

Note: DO NOT SKIP READING THE BELOW TEXT

Papers under each section are ordered from most to least important, so start counting from the top.

[ x,y ]
x: necessary mandatory number of papers to be read – absolute minimum number.
y: the recommended number.
No number: Read all the papers listed.

It is your responsibility to ensure that you grasp the concepts covered in this class; the readings are my top choices (suggestions) for getting this done. However, if you are aware of alternative sources of treatment of the concepts covered in these you may prefer to read about them from your preferred source. If in doubt, ask me.

Assigned readings should be read before class.
If you do so you will already have some familiarity with the subject matter, which not only means you will remember it better but also that you can ask questions for clarification during the lecture and partially steer its direction.

Reading the papers after class is less effective.

You are expected to read all of the papers assigned in this course, at least 2-3 papers per week (4 recommended). Keep at it and you'll be fine!

Warning: Do not attempt to read papers during the group sessions as this is the absolute worst way to cover this material if you truly are interested in learning (you may of course have it open for reference).

Reading the assigned readings not at all should generally be avoided.

As you read papers from each of the following categories I want you ask yourself a few questions:

• For each paper in each category X, ask yourself:
  • What is X?
  • How does the human mind do X?
  • Do current computers do X?
  • …and …
  • Do we need (to replicate or capture) what the human mind does to achieve X to create a machine that rivals the ability of humans to do X?

If you can answer them satisfactorily when you're done reading you're good! Even if you can't you'll be fine if you: Write down the discrepancies and bring them to class in the form of questions. There is no such thing as a 'stupid question' when you're learning something new.

EOF