So I've now attended at least one session of each of my classes. At least so far, I'm happy with the decision I've made to go back to school. Then again...I haven't had any tests yet.

This semester I'm taking:

Mind and Language, which deals with the underlying physiology of speech production and language comprehension. We discuss rival theories which basically boil down in many ways to the old nature vs. nurture debate. Some see language ability as mostly innate, with hard-wired, specialized bits of the brain attuned specificially for language. Others see it more as a general skill that emerges from interaction.

History and Foundations of Cognitive Science is a history of how the discipline came into being, and what it is. The field really began to gel in the 1970's, and because it is relatively new, and extremely interdisciplinary, there seems to be a lot of introspection about the field itself. This class is gonna cover all that.

Cognitive Neuroscience is basically looking at the physiology of the brain at its various hierarchies, from the system level, the network level, and the cellular level. Because I'm mostly interested in modeling functionality, I don't care too much about the lowest level (this would be like an engineering building a bridge really focusing on the electrons and protons in iron), but it's all interesting anyway.

I'm also co-assisting two professors this semester. One a linguist, the other a computer scientist...though nothing's gotten started with either yet.

Anyway, I'll post here about how things are going. The classes are smaller and will tend more toward discussion rather than lecture. And I can already tell that I may get into healthy disagreements with some of my professors.

For example, one of the assigned readings for the History and Foundations class is Alan Turing's 1950 paper Computing Machinery and Intelligence. Here's one of my blogs on this paper, which is pretty silly really. It may have been useful in the sense that it got people seriously considering the idea of intelligent machines (although this was also the Golden Age of science fiction, which played a pretty large role in getting people to imagine thinking machines). And Turing was a genious, no doubt. His theoretical work laid the basis for modern computing.

But his test is silly. And this paper is an embarrassing artifact. In that blog entry I talked about how his test is theoretically poor because it is looking for certain attributes that are obviously not necessary requisites for intelligence. I didn't even get into some of the kookier stuff in the essay.

Turing was brilliant, but like a lot of brilliant people, he was a bit off-balance. The last half of his essay is meant to address objections to its viability. One objection is that people with ESP would screw up the test. No, he wasn't kidding.

I assume that the reader is familiar with the idea of extra-sensory perception, and the meaning of the four items of it, viz. telepathy, clairvoyance, precognition and psycho-kinesis. These disturbing phenomena seem to deny all our usual scientific ideas. How we should like to discredit them! Unfortunately the statistical evidence, at least for telepathy, is overwhelming.

Well no, it wasn't. Isn't now and never was.

And how would somebody with ESP screw up the test?

Let us play the imitation game, using as witnesses a man who is good as a telepathic receiver, and a digital computer. The interrogator can ask such questions as 'What suit does the card in my right hand belong to?' The man by telepathy or clairvoyance gives the right answer 130 times out of 400 cards. The machine can only guess at random, and perhaps gets 104 right, so the interrogator makes the right identification.

Um...okay. I remember first reading this essay, not in the context of a class. Nobody else was around. I was reading one of the greatest mathematical minds of the century talk about how ESP would muck up his test. Wha? I reread it for irony and didn't find any. This stuff got published.

Anyway, that doesn't necessarily discredit the idea of the Turing Test in and of itself...but it does show some pretty flaky thinking going on.

So I think this essay actually ended up hurting the field of AI more than helping it. Look in pretty much any AI text and they'll cite it as a seminal paper in the field. I'm just gonna have to disagree.

So we'll see how that, and the other classes, go.