hougen@umn-cs.CS.UMN.EDU (Dean Hougen) (01/04/90)
In article <35@tygra.UUCP> jpp@tygra.UUCP (John Palmer) writes: > >Artificial neural nets will still be unable to solve hard problems >(patttern recognition, REAL language processing, etc) because they >are implemented in silicon (usually as a virtual machine on top of >a standard digital computer) and are therefore inherently inefficient. >In theory (Church-Turing Thesis) it is possible for such problems to >be solved by digital computers, but most of the hard problems are >intractable. We are very quickly reaching the limits of >speed of silicon devices. > >My point: We are not going to solve the hard problems of AI by >simply developing programs for our digital computers. We have to Although you may be correct that we may never "solve the hard problems of AI by simply developing programs for our digital computers," this says almost nothing about whether or not machines can be constructed (in theory) which think. A slow machine may not be practical, but would its speed prevent it from thinking? It might not pass the Turing Test, but remember, the Turing Test is not the last word in everything that we might choose to call intelligence. Would we wish to say a system that came up with Relativity in 30,000 years was not thinking because it simply took too long? (What if we put a man on a space ship travelling near the speed of light, who during his 30,000 year (our time) round trip came up with this brilliant theory - would we say that he was not thinking because it took him too long? The speed of his brain, relative to ours, is of no consequence, right?) Dean Hougen -- "I'm 20,000 light years from home." - the Stones.