Re: Old news but still worth noting: Computers rule chess
- From: denis feldmann <denis.feldmann.sansspam@xxxxxxxxxxxxxxxx>
- Date: Thu, 05 Jan 2006 13:30:36 +0100
Frank de Groot a écrit :
<royls@xxxxxxxxx> wrote in message news:43bc7825.43139474@xxxxxxxxxxxxxxxxxxxxxxxx
So it is irrelevant to know how long it took for the human brain to evolve.
the only relevant issue is how long it takes to evolve a Go eval function.
OK, how long will that take?
Depends how much money you will throw at it, LOL.
Like I said, a genetic solution scales (linearly?) with money (= CPU's + RAM)
But it also requires that the problem be amenable to evolutionary solutions. I'm not sure go is in that category.
It is 100% proven that a computer program can evolve to human level in Go, because the human brain has likewise /evolved/ from (protons, electrons and neutrons and various types of radiation and forces ("weak" & "strong").
Mmmm... Besides the fact you confuse two (or even three) different meanings of "evolve" (general meaning of change, from primeval post Big-Bang state towards, say, solar system and solid Earth / biological meaning of (darwinian or such) evolution (from, say, early cells to modern organisms) / computer science meaning of "evolution of genetic algorithmes") , the key word is *can* : from early cells to human being able to play professional go, human beings who couldn't care less (like yourself) and, for instance, snakes, trees and flowers, what gives you any mean of *choosing* the "right" path ?
Of course if you are religious and do not believe in evolution, all bets are off :)
The _most_ complex problems, or just more complex problems than have been solved to date?
From the set of all problems people are trying to solve and have been trying to solve/have solved in computer science.
.
- Follow-Ups:
- Re: Old news but still worth noting: Computers rule chess
- From: Frank de Groot
- Re: Old news but still worth noting: Computers rule chess
- References:
- Re: Old news but still worth noting: Computers rule chess
- From: Xiaoding
- Re: Old news but still worth noting: Computers rule chess
- From: royls
- Re: Old news but still worth noting: Computers rule chess
- From: Frank de Groot
- Re: Old news but still worth noting: Computers rule chess
- From: royls
- Re: Old news but still worth noting: Computers rule chess
- From: Frank de Groot
- Re: Old news but still worth noting: Computers rule chess
- Prev by Date: Re: Old news but still worth noting: Computers rule chess
- Next by Date: Re: Old news but still worth noting: Computers rule chess
- Previous by thread: Re: Old news but still worth noting: Computers rule chess
- Next by thread: Re: Old news but still worth noting: Computers rule chess
- Index(es):
Relevant Pages
|