ITV9070 2023

Allikas: Lambda
märgid

Code: ITV9070
Link: http://lambda.ee/wiki/Semantika_ja_anal%C3%BC%C3%BCtiline_filosoofia
Title: Semantics and Analytical Philosophy Archive: year 2022 archive
Lecturer: Tanel Tammet
Contact: tanel.tammet@ttu.ee, 6203457, ICT 426

"A sign is anything that can be used to tell a lie." --Umberto Eco

“I’m writing a book on magic,” I explain, and I’m asked, “Real magic?” By real magic people mean miracles, thaumaturgical acts, and supernatural powers. “No,” I answer: “Conjuring tricks, not real magic.” Real magic, in other words, refers to the magic that is not real, while the magic that is real, that can actually be done, is not real magic. --Lee Siegel


NB! This is an archive of the year 2023, not the actual materials!

Time, place, credits

Time: Every Friday 16:00 - 17:30

  • Normally we will have student presentations during the lecture time, along with planning following presentations and brief intros by the lecturer. We will follow the schedule
  • There will be additional video lectures by the lecturer during some of the weeks to study, in addition to the Friday time above.

Location: IT building room ICT-507AB

Credits: 6

The lectures will be in English, unless everybody understands Estonian, in which case they will be in Estonian.

The presentations in practice sessions can be given in English or Estonian, as you prefer.

The contents of practice sessions are presentations by students and their discussion. For all students, presence in the practice sessions is important and is accounted for in the final grade.

We will probably succeed with saving the videos of lectures: will check out during the first weeks.

The course will finish with a written exam. The materials below marked with a red E will be critical for studying for the exam, while the materials marked with blue L are also used but less important. Materials without such marks will not be used for the exam questions.

In order to be allowed to take the exam, each student should give at least four presentations. Presentations are not graded, but when they are determined not to be satfisfactory, they should be re-done.

Some of the later lecture times will be used as additional practice sessions.

Initial course plan, materials and topics

We consider several topics, starting from general ones and moving to more specific.

The main goal of the lectures is to make it easier to understand the concepts and ideas in the topics, in order to better understand the reading materials.

Intro, semiotics, philosophy

  • Source for the first round of presentations: Tähendus, tõde, meetod.. Separate original articles can be found by googling.

Main materials:

Useful to read/listen:

About the alphabet:


Notes on structuralism, deconstruction and friends:

Notes on Propp:

Notes on the Vienna circle and logical positivism:

Evolution of brain, intelligence and society

Sociobiology, evolutionary psychology and memetics.

Recording of the lecture: 15 Sept

Main material:

Recommended overviews:

Recommended books:

Important papers:

Additional notes and recommendations, not obligatory:

Brain science

Intro material

Intro talk video: 29 sept.

For exam:

Good general overviews:

Important podcasts:


General notes:

Read also:

Infinity, logic and philosophy of mathematics

The first topic is intro to the philosophy of mathematics, mainly the questions of infinity and unsolvability.

Materials to read:

Videos have heads down:

More about infinity and computability:

Helpful read: cardinality wikipedias. Read more about Gödel's incompleteness theorem from SEP: a part of a larger incompleteness-article.

A longer story about the philosophy of mathematics from SEP.

Nonclassical logics, related to the philosophy of mathematics:


More on nonclassical logics:

Thinking and AI

Detailed arguments and counterarguments for the

Some good core stuff


Living classics from the lex fridman podcast:


Semantics of natural language

We will look into both

  • Vector semantics
  • Semantics based on world model

Details will be added when the block starts

Dialog, pragmatics, uncertain knowledge

The first subject is dialogue pragmatics. Several related branches (have a brief look, but no need to read deeply unless you become interested):

To start, certainly look/listen these short lectures:

In case you just want to read original Grice:

Have a look at

Additionally, nice to listen and :


The next theme is uncertain information and probabilities: how to understand, write down and make conjectures.

General intro here as a part of this course.


Read this intro: Numeric uncertainty (L as explained next):

  • Certainly up the the section "Different ways to encode confidences in logic", also looking at the wikipages pointed to
  • Briefly look into the following sections, but do not focus or spend much time on these

Then read the first parts and have brief looks at the SEP articles:

Here are some of our papers on our system, about confidence and reasoning with exceptions.


Extra stuff

Presentations

Schedule

Tarmo, Anton, Rahman, Fuad: scheduled for 13 and 20 oct, see below.

15. Sept

22. Sept

29. Sept

6. Oct

  • Villem - Frame problem in AI. Start with wiki and continue to Dennett for philosophy and this for blocks world example. For deeper stuff read SAP article and another classic about frame problem.
  • Sadiq - modal logics S4 ja S5 and possible worlds (Kripke semantics). wiki is immediately technical and a bit hard to read. The SAP article with the Kripke addendum is better. A good detailed, yet approachable intro + further details are here. Read also this about epistemic modal logic.
  • Zhe - frequentist vs bayesian vs other interpretations of probability. First have a look at the SAP page. Notice the Bayesian approach is a part of the wider "Subjective Interpretation" chapter. This is a good comparison of Bayesian/frequentist approaches.

13. Oct

20. Oct

  • Fuad: pragmatist philosophy
  • Eva-Lotta: Gödels incompleteness theorems and their influence on philosophy. There is an OK overview in Estonian by the lecturer. You can get the main details from SEP.
  • Tiina: Overview of the main points, arguments, thought examples and counter-arguments of/for the functionalist school. As always, SEP is a good detailed material. The computational mind thought line is very closely related to functionalism.

27. Oct


3. Nov

All of the following are really popular, hence it is easy to google for different papers, presentations and other materials. The symbolic vs non-symbolic is harder, due to an extreme amount of heated discussion.

  • Deniss: Donald D. Hoffman and the user interface (MUI) theory of mind. Please prepare already for 27. Oct, as a spare.
  • Ian: Kahneman & Tversky: briefly about Prospect theory and more on System 1 and System 2 thinking
  • Jaanus: symbolic vs non-symbolic debate. If possible, please connect it also the classic nature-vs-nurture and individual-learning-vs-evolution and plato-vs-aristotle thought lines. It may be also interesting to have a look at this paper.

10. Nov

NB! The following schedule is preliminary: some people are still waiting for the slot, so please email me (again, in case you are not scheduled despite previous emails etc) to schedule you for the 10. Nov, pushing some of the presentations below to the 17. Nov:

  • Veiko: Main points of Chomsky's philosophy of language. Main criticisms and support. Notice that it is a prime example of "evolution-first" as opposed to "learning-first" position.
  • Villem: Paul Grice: overview of the most important of Grice's claims and theory, including dialog (start with https://en.wikipedia.org/wiki/Implicature and https://en.wikipedia.org/wiki/Speech_act and have a look at https://www.youtube.com/watch?v=CGJGIWqnSFY ).
  • Sadiq: Intuitionistic logic + examples of proofs of theorems which are (a) possible classically, but not intuitionistically, (b) easy classically, but hard intuitionistically. Please prepare already for 13. Oct, as a spare. Start with wiki and continue to SEP. Have a look at the curry-howard stuff. Most importantly, google for materials talking about the irrationality(?) of the square root of two and around this.

17 Nov

Normal 30-min talks:


24 Nov

Normal 30-min talks:

  • Rahman: Yann LeCun on the problems of LLMs, like the need to find new architectures for machine learning and the need to connect LLMs to tools. lecun about his views, recent, another video with slides. An important detailed paper and a larger tool use paper
  • Fuad: arguments about the danger / high risk of AI (e.g. Stuart Russell, Geoffrey Hinton, Yoshua Bengio), vs arguments about the usefulness / small risk of AI (e.g. Yann LeCun, Marc Andreessen, Andrew Ng). Best to google these guys presenting their views, since the topic is hugely popular.
  • Tiina: vector semantics. Use this chapter from this book as the main material (surely you can reuse slides from here as well. Obviously, searching for additional examples etc would be also nice.

1 Dec

Quick 15-minute talks:

  • Moved to 15 dec: Eva-Lotta: panpsychism and Philip Goff. A useful podcast episode. Also, have a quick intro look into the “hard problem of consciousness” of Chalmers.
  • Moved to 15 dec: Tiina: Max Tegmark and the mathematical multiverse. A useful podcast episode.
  • Deniel: David Deutsch and the probabilistic (or rather, not) nature of scientific theories. A useful podcast. Also the homepage. Here is one complicated paper on constructor theory, as a potential additional deep source.
  • Eva: Jeff Hawkins and the main Numenta theories of intelligence. See numenta, some of the talks by Hawkins and, for example, a useful paper from there.
  • Siim: Yejin Choi on AI and Common Sense. A suitable podcast episode and the homepage.
  • Muaan: Anil Seth on consciousness. A useful podcast episode and a homepage.

8 Dec

Quick 15-minute talks:


15 Dec

Quick 15-minute talks:

  • Tarmo: Antonio Damasio and the origin of important stuff in homeostasis. Interview and a ted talk. See also https://en.wikipedia.org/wiki/Somatic_marker_hypothesis
  • Anton: Donald D. Hoffman and the user interface (MUI) theory of mind. Was planned for Deniss on 3. Nov, not presented? Easy to google, highly popular.
  • Not present: Rahman: Randy Gallistel on memory interview 1 and maybe also interview 2. Some additional background: here
  • Fuad: Confirmation bias
  • Eva-Lotta: panpsychism and Philip Goff. A useful podcast episode. Also, have a quick intro look into the “hard problem of consciousness” of Chalmers.
  • Tiina: Max Tegmark and the mathematical multiverse. A useful podcast episode.

plus 1 reserve slot to be filled as needed

22 Dec

  • Consultation for exam + some spare slots, including:
  • Tiina: functionalism

Some themes for selection

Ideas for the first rounds from this book (search originals from the web):

  • Putnam: tähendus ja osutus. (Meaning and reference)
  • Quine: empirismi kaks dogmat. (Two dogmas of empiricism)
  • Dummett: Truth.
  • Tarski: semantiline tõekontseptsioon.
  • Strawson: tähendus ja tõde
  • Ramsey subjektiivne tõenäosusteooria.

Further:

Pärit leheküljelt "http://lambda.ee/w/index.php?title=ITV9070_2023&oldid=21437"