6.868 / MAS.731

Society of Mind

Spring 2013

« Return to the main page

 

The subject of our first student-driven discussion will be emotions—what are they for, how do they work, are they beneficial, can you write algorithms for them? The discussion will take place in the latter half of class next Wednesday (13 Feb 20 Feb).

Some of you are on the discussion panel for next week. You are expected to write out your answers to the discussion questions below. You may bring them with you during the discussion, and may write notes on them as well. We'll collect them at the end of class— our grading criterion will be evidence of thoughtfulness and preparation.

Others of you will be in the audience. You are also expected to prepare your answers to these questions so that you can participate actively, but you are not required to write them out. (We do encourage you to do so, however, because it promotes coherent and concise discussion.)

You are all enthusiastically encouraged to work together on these assignments; of course, the final write-up must be your own work.

Assignment 1: Emotions and Rationality

In The Emotion Machine, Prof. Minsky argues that “each of our major ‘emotional states’ results from turning some set of resources on and turning another set of them off. Each such selection will change how we think by changing our brain's activities.”

Your task is to critically analyze this thesis, by determining your answers to the following discussion topics.

  1. Is there an evolutionary advantage to having (at least some of our) emotions? If so, what is the advantage, and how does it outweigh apparent disadvantages? If not, why haven't we evolved out of having emotions yet?
  2. Prof. Minsky denies that there is a meaningful difference between emotional and rational thinking. Find as many counterexamples as you can to this hypothesis. For example: After you collect a list of such differences between emotional and rational thinking, try to reconcile them with Prof. Minsky's hypothesis. Is he right or wrong, after all?
  3. Is it possible to program a computer with emotions? Would it be desireable?
  4. How would you construct a program which would solve problems ‘emotionally’ rather than ‘rationally’? Is a different design strategy required?
  5. Some critics of AI (e.g. Searle, Penrose) have argued that even if you could program a computer to behave as if it had emotions, it nonetheless would not actually feel anything the way that we do— it would only be able to execute subroutines, which only cause actions, not feelings. What do you think?

    (See also: Prof. Minsky's reply to Penrose)