My 2/20/11 Missoulian column
Did you watch the “Jeopardy!” match last week between the supercomputer named Watson and two previous human “Jeopardy!” champions?
The development of Watson at IBM and the concepts behind the very complex software algorithms and “learning” that went into the supercomputer were amazing and were outlined on PBS’ “Nova” science show last week, too.
The technology developed will only get more complex and powerful, and the practical applications have “Star Trek”- and “2001: A Space Odyssey”-like implications.
Watching the actual “Jeopardy!” match was interesting. One could see a small part of Watson’s “thinking,” where “he” came up with multiple questions for the answers and narrowed them down. And one could wonder about “his” strategy when it came to selecting a category and placing a bet on the daily double.
But I got bored half an hour or so into the match. Why? The match wasn’t fair.
I don’t have so much of an argument with the amount of data that was fed to Watson; storing and using vast amounts of data is one of the strengths of information technology.
It was an advantage for Watson, but in this case I don’t think it was a crucial one. Ken Jennings and Brad Rutter have proven themselves to be capable of absorbing and recalling the millions of facts needed to be become the past champions they are.
But one advantage Watson had over the humans was that “he” could push the answer button within 10 milliseconds.
That’s a much faster reaction time than the poor human past champions. I could see their frustration as they missed by a fraction of a second again and again.
If you’re going to treat a machine like Watson as something of a human, make it fair. Give the poor humans a few more milliseconds. If IBM wanted to make it fair, a human reaction time would have made the match more realistic.
I think the whole match turned on that single advantage, and to me, that’s not the soul of the game.
This might seem like a small thing to quibble over. But an idea of fairness isn’t. It’s typical of the speed of technological change; something always gets left behind, and in this instance, it’s a moral idea of what is fair.
And the brings forth ethical debates regarding technology. If the creators of Watson can’t play fair, what does that mean for the rest of the machine? And how such technology we make will be applied in the future?
This week in Mac Q & A: Windows, Taxes and a Mac