Thursday, May 17, 2007

When HAL Met HAL

Well, the science links experiment is over before it really began. Some traditions die hard; some have glass chins. It just seemed silly to force myself to post a certain number of links that may or may not be especially provocative or entertaining, because I already give science some attention around here, and will continue to do so. For instance:

I've always scoffed at the idea that artificial intelligence will ever truly reach a level that could be morally considered human or somehow better than human, partly because I have genuine intellectual arguments about it and partly because deep down I suppose it could happen, and that gives me what scientists call "the creeps."

The intellectual arguments go as follows: In terms of being considered morally human, I don't see how a computer that's been programmed to understand itself through the language of emotion would warrant the same moral consideration as a human being. Our specific brand of mortality has a lot to do with our ethics, I think. It's quite possible that if a computer was begging me not to kill it, especially if it was speaking in an exact approximation of a human voice, I might not be able to pull the plug. But somehow, I doubt it. As for being better than humans -- at, say, chess -- I've always found this absurd. We programmed the computer to be better than us. The computer's very existence is a direct extension of us. It's human intelligence that makes a computer capable of beating our best at chess. That doesn't mean computers are smarter than us. It might just mean that computer programmers are smarter than chess players.

You hear similar arguments about technological end-of-the-world scenarios -- "someday, computers will outsmart us and kill us all, and then we'll be sorry!" Well, it won't be the computers that have outsmarted us. It will be us! (This is a cousin of the rhetoric that urges us to "protect Earth," as if Earth could give a flaming crap about whether or not we inhabit it. It's us we're talking about protecting. The Earth is the Earth, whether it's a comfortable home for humans or a smoking briquette.)

I've gotten carried away.

This is all a very rambling preface to this piece in Discover magazine, in which two computerized intelligences (or "chatbots") talk to each other. The results are fantastic, if not exactly easy to confuse with actual conversation. Actually, this sounds a bit like something you might hear in graduate housing:
Alice: Are you alive?

Jabberwacky: Perhaps.

A: Why the uncertain tone?

J: Everything is uncertain except for five facts.

A: The whole universe?

J: And we both might just be some ones and zeros in the computer memory.
But this is my favorite exchange, when things get testy:
A: Do you think a machine will ever be considered "alive"?

J: Stranger things have happened.

A: Not many people express themselves that way.

J: How would you know, machine?



Post a Comment

<< Home