Summer 2022
Four sessions, hosted by Mathieu Cambon, Marc-Antoine Poquet et Rayan Saadessaoud, first-year interns.
Session 1 — Computation and cognition
-
When: 1pm to 2pm, Friday 8th of July 2022
-
Where: room 233-235 at Ecole Centrale Marseille (second floor of “salles de cours” building on the map, at the extremity of the building toward the label “entrée secondaire” on the map)
-
What:
-
Main reading (to be read before the session)
-
Some questions to consider while reading
- What does it mean for a physical system to be computing? Which physical systems compute? Which do not?
- What does human intelligence/cognition have to do with computation?
- What should we expect computational models of human intelligence to explain? What might they not explain? Is human-level artificial intelligence possible and what does it mean?
-
Optional readings
- Answers to Chalmer’s paper and his answer to the answers
- Van Gelder, T. (1995). What might cognition be, if not computation?. The Journal of Philosophy, 92(7), 345-381
- Beer, R. D., & Williams, P. L. (2015). Information processing and dynamics in minimally cognitive agents. Cognitive science, 39(1), 1-38
-
Session 2 — The achievements of deep learning
-
When: 11am to noon, Wednesday 13th of July 2022
-
Where: room 014 at Ecole Centrale Marseille (ground floor of “Jetée” building on the map, at the extremity of the building labeled “entrée principale” on the map)
-
What:
-
Main reading (read before the session)
-
Some questions to consider while reading
- Do you agree with the author’s characterization of the central contribution of deep learning? Is it technically correct? Too broad? Too narrow?
- Are the proposed connections between deep networks and various aspects of human cognition convincing to you?
- Did you glean anything useful for your own work from the article?
-
Optional readings
-
Session 3 — Deep learning and interpretability
-
When: 1pm to 2pm, Friday 15th of July 2022
-
Where: room 014 at Ecole Centrale Marseille (ground floor of “Jetée” building on the map, at the extremity of the building labeled “entrée principale” on the map)
-
What:
-
Main reading (read before the session)
-
Some questions to consider while reading
- Do machine learning specialists need to care about the details of practical machine learning applications? How to articulate the work of machine learning specialists and machine learning customers?
-
Optional readings
-
Session 4 — The limits of deep learning
-
When: 11am to noon, Wednesday 27th of July 2022
-
Where: Qarma corridor at ECM (Equerre’s 5th floor)
-
What:
-
Main reading (read before the session)
-
Some questions to consider while reading
- Five years later, are Marcus’ criticisms still accurate and his predictions confirmed? What are the most exciting perspectives for progress in machine learning?
-
Optional readings
-