This is of course a straw-man claim, computers are merely transistors arranged in patterns that execute algorithms that are pre-defined. The brain also performs basic operations at a neuronal level as well it has been shown that neurons can do a wide range of operations on a given input. The real question is what is the algorithm that is being executed and how does the brain know which algorithm to use or which operation to execute? And of course the brain can change its wiring patterns making things even more interesting.

"Deep Blue played chess barely better than the world's top humans, but a heck of a lot better than its own programmers. Deep Blue's programmers had to know the rules of chess - since Deep Blue wasn't enough of a general AI to learn the rules by observation - but the programmers didn't play chess anywhere near as well as Kasparov, let alone Deep Blue."

A feeble example since chess is a mathematical exercise that readily lends itself to mathematical analysis and algorithm development. Human intelligence while more predictable then some like to believe, still is not nearly as predictable as chess. Lets be real honest about Deep Blue the real advantage over a person is that Deep Blue can analyze far more moves then a person can and can plan further ahead. If we change the game to something like facial recognition the computer is far worse for the most part though facial recognition has been improving.

"If you can understand steering-the-future, hitting-a-narrow-target as the work performed by intelligence - then, even without knowing exactly how the work gets done, it should become more imaginable that you could build something smarter than yourself."

Key word imaginable, not necessarily possible since the problem of AGI is not nearly as simple as Chess. Nor does a good chess computer make even the slightest case for AGI. AGI is only proven through construction the rest is babble and hollow claims. It is akin to claiming to be able to build a house all by yourself and then never really doing it, to make that claim one must prove it by the doing.

"There are those who will see it as almost a religious principle that no one can possibly know that a design will work, no matter how good the argument, until it is actually tested."

What works in theory and what works in praxis are two different things sometimes, since your not an engineer you wouldn't know. Try building a design and you will understand until then you really are an outsider.

"Just like the belief that no one can possibly accept a scientific theory, until it is tested. But this is ultimately more of an injunction against human stupidity and overlooked flaws and optimism and self-deception and the like - so far as theoretical possibility goes, it is clearly possible to get a pretty damn good idea of which designs will work in advance of testing them."

Never the less proof through the testing of the idea is needed no matter how good the theory, without testing you have nothing. The point of testing is to ensure that the theory works in practice the same way as it does on paper.

Sorry for cutting out swaths of your post, they were swaths of verbiage not worth the answering. Please if your going to speak about these thing try engineering a real AI and then come back to these philosophical meanderings. Until you have real experience in the programming and the math of AI, beyond Bayesian Statistics you cannot hope to make a case for AGI or build one yourself.

Posts

Wikitag Contributions

Comments