October 20, 2007
Artifical Intelligence Gets Sidetracked
Over at MIT, a video of Marvin Minsky's discussion of AI and commonsense:
Marvin Minsky is worried that after making great strides in its infancy, AI has lost its way, getting bogged down in different theories of machine learning. Researchers “have tried to invent single techniques that could deal with all problems, but each method works only in certain domains.” Minsky believes we’re facing an AI emergency, since soon there won’t be enough human workers to perform the necessary tasks for our rapidly aging population.
So while we have a computer program that can beat a world chess champion, we don’t have one that can reach for an umbrella on a rainy day, or put a pillow in a pillow case. For “a machine to have common sense, it must know 50 million such things,” and like a human, activate different kinds of expertise in different realms of thought, says Minsky.
Minsky suggests that such a machine should, like humans, have a very high-level, rule-based system for recognizing certain kinds of problems.
Posted by Robin Varghese at 04:17 PM | Permalink
Comments
What work has been done toward the development of a contextual analysis algorithm, such that any AI would in fact be, as humans are, several different intelligence systems with the ability to assess the situation and make a determination toward the best system to use?
Posted by: Damien | Oct 20, 2007 10:30:32 PM
Marvin's being pretty cheeky here (to say the least). After a failed career to create anything intelligent, or even to understand the nature of intelligence -- natural or artificial -- he's witnessed a new generation of researchers who found success and solved some interesting (if elementary) problems of AI by completely ignoring him, his work and most of all his book "Perceptrons". I understand -- Marvin's upset by the fact that he has been proven to be wrong on almost everything he's ever said about AI, and he's just trying to get back at the people who humiliated him by their success. There's one big problem with his comments, though -- if it were possible for a "high-level, rule-based system" to have "common sense", which is more or less what Minsky's been working on for 50 years, surely he'd have it, or at least a demo, by now. If he wants a robot to make his bed, he's looking under the wrong streetlight.
Posted by: JO'N | Oct 22, 2007 12:57:54 PM
"Minsky believes we’re facing an AI emergency, since soon there won’t be enough human workers to perform the necessary tasks for our rapidly aging population."
He can believe that, but where is his evidence? All advocates and researchers of life extension are adamant that it’s progress is in prolonging the period of relative youth, not the period of decrepitude. Social commentators, by and large, also do not see any forthcoming crisis due to rapidly ageing population in the few most developed countries - with the exception of Japan. And it is, of course, a fallacy to see a social problem as solvable exclusively by technology of a very specific - and very ambitious - kind.
Plus, as Peter Drucker stated, the innovator should address the problems of today, not those of the future.
All things taken into account, this looks very much like a substitute of the real explanation for his dream to live like a brain in a vat, untroubled by any manual (i.e. physiological) tasks. Perhaps if he could face this fact, he might doubt the soundness of his aspiration.
Posted by: Ivan | Feb 14, 2008 11:46:54 AM
Post a comment