Ask Ray | How do you respond to Noam Chomsky’s claim that ‘Watson is not good AI’?
February 19, 2011
Hello,
I was wondering if I could get a comment from Mr. Kurzweil. I know he thinks Watson will win the Jeopardy! match — I agree. Professor Noam Chomsky has told me that Watson is not good AI, though, and I’m curious how Kurzweil would respond to Chomsky’s words, found here.
I read your article in PC Magazine, which is what persuaded me to contact you. With regards to Chomsky, are you basically saying he’s right? It sounds like your position is that Watson may not be AI, but remains a very important step.
Sincerely,
Gavin Schmitt
Gavin,
The interview you had with Chomsky is interesting but his answers are so brief that it is difficult to understand what he is trying to say. I would say that Watson is clearly not yet “strong AI,” but it is an important step in that direction. It is the clearest demonstration I’ve seen of computers handling the subtleties of language including metaphors, puns and jokes, something people had said would not be possible. I don’t agree with Chomsky that Watson is not impressive, in that regard.
It’s not yet strong AI. As long as AI has any flaws or limitations, people will jump on these. By the time that the set of these limitations is nill, AI will have long since surpassed unaided human intelligence.
Best,
Ray
Comments (9)
by markduwe
Still, the words that come out of our mouths are only 3-5% of all of the communications we make. Most of what we want to say comes in the form of body language and tone of voice. Once you have 2 computers who can ‘talk’ to each other AND read each other’s non-verbal’ communications you’ve really got something.
by iwanttolive
chomsky is a nutjob.
his “theories” on ai are naive and stuck in the 60′s, as are his theories on politics.
If you’re seeking a nutjob’s opinion about the latest on the Middle East, then it makes sense to ask NC. While Watson would give more realistic answers to your questions than NC.
by lmb
I’m not sure you know who chomsky is. He’s kind of a big deal. His theories on how humans acquire language is still being taught in colleges around the world. it’s easy to call someone a nut job, when you don’t have to support your argument. Who are you? What have you done? my guess is you’re just come kid on the internet.
by John Goodrich
Noam Chomsky is too brilliant a man and mind to speak of in such a way.
He is an anarcho-syndicalist which thinking pertains to a future democratic and government-less society and has nothing to do with the past except that he has been consistent in his anarcho-syndicalist beliefs for decades.
It is no accident that not many on the right as you are choose to debate the man face to face.
THAT said , I believe Chomsky is more likely to be in error regarding AI than is Kurzweil.
It is yet to be determined if enough accumulated bits of knowledge can lead to AI thinking as do humans .
There is the big Singularity question of what does happen to an AI when it attains billions of Earth civilizations worth of information ; does it gain consciousness is something that we learn when and if it happens.
If it achieves consciousness, it will be based on information that is entirely human in origin. If it is “smarter” than humans then it may well take all human intelligence and and sort it so that more intelligent decisions are arrived at.
These are just some of the big questions whose answers will determine how an advanced AI would present itself, what its capabilities are and how close to human thinking, albeit far faster, far more intellectual and far more rational, those processes will be in EFFECT.
by Xcythe
I believe AI’s problems are simplistic in nature. Intelligent life developed out of need, adversity. Our moon gave us tidal pull, surge. and our tilted axis gave us seasons. I believe an ‘itch’ could change the basic parameters. How can you scratch an itch you never get? The solutions may be too simple to realize or believe. When you learn too many complex details you forget the basics. To make a pie from scratch, first you must create a universe. Your knowledge is limited by the language you use and the definition of the words in that language. What is it that you don’t know?
by mreeson
linguistics actually play a large role in the development of “artificial intelligence”. look up computational linguistics. specifically, the “chomsky hierarchy” theory has played a significant role in computer sciences. so i think his opinion is relevant. his political philosophy has made him famous to most of the world, but it would be difficult to overstate the importance of his work in the field of linguistics and the cognitive sciences.
by eberlyblog
Im no expert in any particular field of study but I often express to the fullest.
1. The smallest steps towards AI will be that of IBM, google, or any network,computer, software companies out there. The largest stepps will be made by our cognition of consciousness which is no tangible substance to science.
2 Why AI? I believe humanity’s pursuit in the products of future AI will result in that of a collective hive mind that is centralized. Who wants to be a worker be for something else? I believe the greatest pursuit of humanity will be the further evolution and cognition of our selves and individualistic expression. A good friend of mine recently said “Technology is here to optimize choices, not replace choices” I stand by that saying.
Brain Science and Neuro Science is still lagging behind and in a crawling state at the moment. We will understand how the brain works in 20 years… thats what ray says but the brain is not the only intelligence center. There are other studies that show nerve densities in both the belly region and the heart region of the body. Have you guys seen the brain scan of a monk? How much more activity there is in their brain. Google it. Spiritualists have known for decades that its not about the brain. When we activate the other centers within the body our consciousness and our intelligence becomes far more expanded. There are other technologies being used today that are way beyond what AI will be in any year from now. And it starts with awareness and us as conscious human beings. Why? Because AI will only be able to mimic that of a creator. Anyway not to shoot down AI. I have a lot of respect for those technologies and Ray’s work but I feel like today’s or tomorrows applications and attempts are not even tinkering around the edge of something that will always only mimic its source. By the way there’s a lot of things I can appreciate about AI.
@nathaneberly on twitter.
by kmac086
Asking Chomsky about AI is like asking Kurzweil about poly sci. You can do it, but it doesn’t make much sense. Noam is a linguistics/politics expert. Ray is futurist/technological science expert. If I’m seeking an opinion about the latest on the Middle East or Israel/ Palestine, it would make sense to ask NC. On the other hand, If it’s cognitive computing or nanotechnology thats the matter of discussion, it would make sense to ask RK.
While I greatly respect Ray’s continuation of Moore’s Law… I’m going to have to side with Noam on this matter. Supercomputers have been around for 50 years now and that’s exactly what ‘Watson’ is. Self replicating artificial technology and future robotics will be comparatively far more complex. Though Watson’s systematic understand of written English is a neat innovation in modern super-computing technology.
by John Goodrich
I may be mistake but I believe that computer power is determined by the number of calculations that it can do.
We’re up to 100 petaflop equipment .
I don’t know if that number of calculations equals the number of calculations a human brain does or in some way would provide what is needed to get near human intelligence in a machine but I want to point out that absent a computer that can equal all aspects of human thinking through accumulated knowledge , we aren’t yet capable of producing a machine that can think like a human due strictly to the limitations of the science AT THE MOMENT.