Many people think of strong artificial intelligence as an artificial device being able to perceive, and reason, in ways similar to a human being. I agree that there are other, lesser bars for artificial intelligence, but the difference is why I use the term "strong AI".
2007-07-05 03:01:07 · 6 answers · asked by Chuck T 2 in Science & Mathematics ➔ Engineering
The perception of AI is as you describe. Few realize that it is only a captured set of evaluation steps to solve specific problems, for example AI in fly-by-wire jet fighters - all it knows is how to fly the plane and translate pilot input into control impulses that make the plane fly. There is no "reasoning" there at all, just a list of recipes that people have entered to solve very spedific problems.
Another good example of AI in practice is medical diagnostic computers, but all they do is go through a list of questions that men have programmed into the computer - there is no "thinking" going on there...
A VERY interesting application is the AI shrink program that acts like a therapist , where people cannot tell it is not a person - but that probably says more about therapists than it does about AI.
A favorite idea from someone I cannot remember was that if a computer could amplify intelligence, it would be equally able to amplify stupidity. Thr problem for the computer is how to tell the difference.
Strong AI? An interesting comparison might be between computers and the human brain. You can start by counting the connections in the brain, but you quickly get lost because the brain is a parallel processing analog device, and for "reasoning" skills, analog circuits require many orders of magnitude less connectiions than digital. How many connectioins in an ant's brain?
It is interesting that there are no digital analogs between an ant's brain and any computer program, but there is a guy who builds robots that behave just like bugs with a handful of parts. But they are analog, and not digital.
Strong AI? Nah, thats is for SciFi and not the real world.
It's really too bad they put the word "intelligence" in there, because that is not what it ever was intended to be - except by the Japanese, perhaps...
Ron.
2007-07-05 03:26:05 · answer #1 · answered by Anonymous · 0⤊ 0⤋
Garry Kasparov defeated Deep Blue in a rematch. Unless there was another match, Deep Blue stands defeated such as, computers are beatable in chess. Ultimately computers are predictable. Even the random number generators being used to encrypt government level encryption are predictable to certain degree. As of now, computers only know what we program them with. True, one could install the data of every book ever written, but unless software was designed to utilize the knowledge without human interaction, the computer would still only know what its been programmed to do. As such current computers can not be "smarter" then humans. I believe it will be possible to write software that is cable of sentient thought, but I think we're still a good 20 years or more away from the hardware ability to handle said software.
2016-05-18 22:18:11 · answer #2 · answered by ? 3 · 0⤊ 0⤋
We first need to get a grip on regular intelligence.
2007-07-05 03:04:48 · answer #3 · answered by fangtaiyang 7 · 0⤊ 0⤋
the problems could be solved in our lifetime. It all depends upon how old you are, and how much longer you plan on living.
2007-07-05 03:09:07 · answer #4 · answered by vtothef 5 · 0⤊ 0⤋
I doubt it. We are less than 1% of the way there.
2007-07-05 03:04:52 · answer #5 · answered by Anonymous · 0⤊ 0⤋
?
2007-07-05 03:04:31 · answer #6 · answered by Anonymous · 0⤊ 0⤋