artificial inteligence is a myth!!!
-
Any programmer who thinks that we are going to reach the human intelligence by if switchs elses for loops, is either crazy or has inhaled a pound of cocaine
Oooh! A pound of coke! Right on man!
-
jschell wrote:
All that is doing is attempting to correlate some extremely rough measure to another.
Then give me a definition which can measure the intelligence of a dog and compare it to a computer. Not to mention, an equally complex neural network in a computer could theoretically out perform a biological one, just on the fact that the biological one also has to do a lot more things, like regulate bodily functions (which obviously a computer does not have).
jschell wrote:
However the correlation is not direct. And computer science (which has more impact than electrical engineering) has advanced greatly but still no AI.
Perhaps, but I'd argue neuroscience has an even bigger impact. How can we imitate what we don't understand? With greater understanding of how a brain works we may have better insight into how to replicate its functions.
lewax00 wrote:
Then give me a definition which can measure the intelligence of a dog and compare it to a computer.
You first - give me an accepted definition that provides objective criteria for intelligence. Make it easy on your self - try to find one that works for just humans.
lewax00 wrote:
Not to mention, an equally complex neural network in a computer could theoretically out perform a biological one,
There are many things that one can hypothesize.
lewax00 wrote:
but I'd argue neuroscience has an even bigger impact. How can we imitate what we don't understand? With greater understanding of how a brain works we may have better insight into how to replicate its functions.
That is simplistic. The fact that I know a great deal amount how sight works in terms of the brain doesn't mean that I can use that to create a computer that does the same thing.
-
lewax00 wrote:
Then give me a definition which can measure the intelligence of a dog and compare it to a computer.
You first - give me an accepted definition that provides objective criteria for intelligence. Make it easy on your self - try to find one that works for just humans.
lewax00 wrote:
Not to mention, an equally complex neural network in a computer could theoretically out perform a biological one,
There are many things that one can hypothesize.
lewax00 wrote:
but I'd argue neuroscience has an even bigger impact. How can we imitate what we don't understand? With greater understanding of how a brain works we may have better insight into how to replicate its functions.
That is simplistic. The fact that I know a great deal amount how sight works in terms of the brain doesn't mean that I can use that to create a computer that does the same thing.
jschell wrote:
You first - give me an accepted definition that provides objective criteria for intelligence.
I'm not the one claiming the cited article isn't a valid example, you are. If something with a brain is intelligent, I have no problem accepting a simulated brain is also intelligent.
jschell wrote:
That is simplistic. The fact that I know a great deal amount how sight works in terms of the brain doesn't mean that I can use that to create a computer that does the same thing.
That's the complete opposite of what I said. Being true in one direction does not necessarily make it true in the other. (In this case "You can simulate it only if you have the knowledge" is not equivalent to "If you have the knowledge you can simulate it". This is basic logic.) And if you know how the whole system works there's no reason you can't simulate it in software. If you can't, then you're probably missing pieces. And considering the article is specifically talking about simulating the human visual cortex, which means they HAVE simulated the part of the brain dealing with vision, so if you can't do it it's because either your knowledge of how that system works or your knowledge of software is lacking, because it's been done already, proving it can be done.
-
jschell wrote:
You first - give me an accepted definition that provides objective criteria for intelligence.
I'm not the one claiming the cited article isn't a valid example, you are. If something with a brain is intelligent, I have no problem accepting a simulated brain is also intelligent.
jschell wrote:
That is simplistic. The fact that I know a great deal amount how sight works in terms of the brain doesn't mean that I can use that to create a computer that does the same thing.
That's the complete opposite of what I said. Being true in one direction does not necessarily make it true in the other. (In this case "You can simulate it only if you have the knowledge" is not equivalent to "If you have the knowledge you can simulate it". This is basic logic.) And if you know how the whole system works there's no reason you can't simulate it in software. If you can't, then you're probably missing pieces. And considering the article is specifically talking about simulating the human visual cortex, which means they HAVE simulated the part of the brain dealing with vision, so if you can't do it it's because either your knowledge of how that system works or your knowledge of software is lacking, because it's been done already, proving it can be done.
lewax00 wrote:
I'm not the one claiming the cited article isn't a valid example, you are.
So you are claiming that that article provides a definitive objective measure of intelligence? And one that allows comparison between humans as well?
lewax00 wrote:
That's the complete opposite of what I said. Being true in one direction does not necessarily make it true in the other.
However this thread and sub thread is exactly about that - creating artificial intelligence.
lewax00 wrote:
because it's been done already, proving it can be done
Then why does research continue on that very subject?
-
Don't you mean "unimpeachable"? Neil.
impeccable (Sorry I'm late, I missed the train.)
-
lewax00 wrote:
Because as we all know, fields never improve and never become more advanced.
Look at the advances in medicine in the past 50 years. And the past 200. Look at the advances in computers in the past 50 years. Look at the advances in bio-engineering in the past 50 years. Look at the 'advances' in parapsychology in the past 50 years. New sciences which can produce results tend to advance quickly. Those that can't - don't. AI is a new science. So why isn't there an AI now that is at least as smart as, for example, a dog?
C'mon, we do have an AI as smart as a dog! http://www.theatlantic.com/technology/archive/2012/06/the-triumph-of-artificial-intelligence-16-000-processors-can-identify-a-cat-in-a-youtube-video-sometimes/259001/