Author: Christophe Theron
Date: 05:11:32 03/15/01
Go up one level in this thread
On March 15, 2001 at 02:13:13, Bruce Moreland wrote:
>On March 15, 2001 at 00:42:03, Christophe Theron wrote:
>
>>On March 14, 2001 at 16:14:02, Bruce Moreland wrote:
>>
>>>On March 14, 2001 at 12:11:41, Fernando Villegas wrote:
>>>
>>>>Well, Bruce, your answer widens the scope of the debate. To begin with, the
>>>>concept of intelligence as something proper of an entity capable of intelligent
>>>>behaviour does not solve the problem as much intelligence as such is not clearly
>>>>stated first. Sounds a little bit like a circular reasonning. Nevertheless I
>>>
>>>I consider that for these purposes, intelligence is the ability to provide good
>>>solutions to the majority of problems inside a difficult and wide problem
>>>domain.
>>
>>
>>I can still see some circularity here in the definition of "good" in the context
>>of your sentence.
>>
>>Am I just splitting hairs, or isn't it something that has to do with the
>>basement of the problem?
>
>I don't think it is circular. There isn't a clear dividing line between what
>problem is complex enough, and what competence is high enough, but that's not
>necessarily fatal. Perhaps this can be placed on a continuum and the problem
>disappears.
>
>So a system is more intelligent if it provides better solutions in wider
>domains.
>
>bruce
This is why I pointed out the fact that "intelligence" cannot be defined so
easily, because it's like the elo level (but maybe on several dimensions).
And the Turing test in this case looks like a game (or a match) which is used to
estimate the elo level of another player.
It is true that the elo level of the player which is used to calibrate the other
one can be totally inadequate, as you pointed out, which is probably the main
weakness of the Turing test.
Christophe
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.