November 28, 2006
the ghost in the machine
ONE OF THE great broken promises of the 20th-century view of the future, right up there with personal jet-packs, was the promise of artificial intelligence. AI was supposed to lead to computers that wouldn't just calculate and organize, but reason and analyze; computers that could really think, like HAL in "2001" or KITT on the 1980s TV show "Knight Rider." (Of course, HAL turned out to be a homicidal psychopath and KITT was a smug know-it-all, but still, it seemed like a good idea.)Recent efforts to realize the promise of AI have centered on teaching computers to better deduce meaning from the vast content of the Web, but there's still a long way to go. In the meantime, however, there's an alternative type of computerized system that is actually making big strides toward getting computers to think like humans. Publisher Tim O'Reilly calls it intelligence augmentation (IA for short), and it uses a very clever technique. It cheats.
more from Boston Globe Ideas here.
Posted by Morgan Meis at 12:13 PM | Permalink
Comments
Is this in any way related to "fuzzy logic"?
Posted by: Bill Ectric | Nov 28, 2006 1:25:51 PM
"Deduce meaning from the vast content of the Web?" Hey, even I can hardly do that, and I'm a Grade A Homo sap.!
And very clever of Mozes Mob and Google to get slaves (oops, I mean "volunteers") to do their work for zilch. Eventually, I predict, a large proportion of the population will be slaves to Google.
Posted by: JonJ | Nov 28, 2006 2:19:43 PM
You can come to one of two conclusions here. Either:
1) Humans are too stupid to figure out how intelligence works;
or
2) Humans are too smart for machines to emulate.
Take your pick ;-)
Posted by: Bruno Mota | Nov 28, 2006 9:20:07 PM
A bit more seriously this time, I think the main lesson we can take from the 'failure' of AI research is that the human brain is even more wondrous and intricate than previously thought.
At every stage in the history of AI the assumptions about how intelligence worked were proven naive, and the human brain was shown to be even be more capable than was then assumed. This happened not so much because we found that our brains could perform new complex tasks, but because we realized that what we used to consider simple tasks were actually quite involved.
When one thinks of supercomputers beating human grand masters at chess, it is easy to lose perspective. To be competitive, Deep Blue had to project many orders of magnitude more moves than Kasparov ever could. It could recall every move of thousands of past games, including all in which Kasparov played. And yet the human was able to hold his own. Clearly the brain works in ways that for some tasks are far more efficient than any computer ever built. Why this is so I have no idea. It doesn't have to be something exotic like quantum computing or some sort of supra-Turing machine, but I expect it to be something highly non-trivial we haven't even touched upon yet.
Posted by: Bruno Mota | Nov 28, 2006 9:41:05 PM
Deep Blue was a hoax. And playing chess is not exactly a way to prove your intelligence - chess is merely go for beginners.
Posted by: Scott Holon | Nov 29, 2006 12:08:45 PM
Well, how big is a neuron compared to a computer chip?
Posted by: Bill Ectric | Nov 29, 2006 12:31:55 PM
Post a comment