The Incredible Thinking Machines

I finished my final multimedia class unit of the semester, on games and game design, with a brief overview of AI.  After discussing Turing's Polite Convention, I asked my students if they believed that computers could think.  The confident response, with which the whole class claimed to agree was, "yes, computers can think because they have brains."

Since in the secondary schools, students in ICT classes are forced to memorize that the CPU is the brain of the computer, this isn't too surprising, though like many things secondary school students are forced to memorize in ICT classes, it's not really accurate.  It might be accurate in a broad metaphorical sense, but broad metaphors are not helpful in this case.  Computers do not have lumps of grayish organic matter inside them.  That would be gross.  Nor do computers have a simulation of a human brain.  That is currently impossible.

So I spent a while asking my students questions about what it means to think, which confused them, but they did bring up remembering and making decisions. So I talked about the differences between human memory and computer memory, and then we talked about how we  humans make simple decisions, e.g., what clothes to wear in the morning, and what facts (cold, dirtiness) we might use to make decisions, and the cultural factors that they don't necessary consider when making decisions (I told them that American university students might show up to class in their sleeping clothes.  My students were shocked.).

Since Tanzanian children play a game completely equivalent to Tic-Tac-Toe (the proof is left to the reader), I drew a board and had the students tell me which opening move they would make if they wanted to win, and which move the next player makes to try to win and so on.  Then I asked how they decided a move would help them win and how a computer could decided.   I actually talked them through the minimax algorithm for solving it, and told them that I could assure them they were not doing that in their brains when I asked them to try to win.

A Tanzanian Tic-Tac-Toe board.  Stones may be placed at
line intersections.  The object is still 3 in a row.   

Yes, I know I'm being ethnocentric calling it Tic-Tac-Toe, and I'm sorry, but the students all said they didn't know of a name for this, but most of them had played it as children, so I'm calling it TZ Tic-Tac-Toe because it is grammatically convenient for stuff to have names.

I always get this feeling that students never believe me when I contradict something they've been memorizing and receiving rewards for regurgitating, but I think the majority of the class actually thought a little about what it means to make decisions, and how complicated it can get, and that's all I was really trying to get across.  Well, that and don't just repeat stuff to me without thinking about it.  Arguably a lot of computer science is describing things in a very precise and unambiguous way.