Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Will It Take A New Kind of Chess Engine to Whip Kramnik?

Author: Robert Hyatt

Date: 10:52:57 12/20/02

Go up one level in this thread


On December 20, 2002 at 11:33:57, Bob Durrett wrote:

>On December 20, 2002 at 10:51:54, Robert Hyatt wrote:
>
>>On December 20, 2002 at 08:52:44, Bob Durrett wrote:
>>
>>>On December 19, 2002 at 23:04:17, Robert Hyatt wrote:
>>>
>>><snip>
>>>
>>>>I think the basic problem is in understanding how _we_ do something.  For
>>>>specific applications, no matter how complex they are, we can develop software
>>>>to accomplish that particular task.  But only _if_ we know how we solve the
>>>>problem.  That is the missing link so far.  We simply don't know how we do
>>>>what we do, we just know we do it.
>>>
>>><snip>
>>>
>>>Hmmm.  That sounds VERY philosophical.  Almost religious.
>>>
>>>So you really do not believe the science fiction writers when they portray
>>>computers of the future creating new computers?  Programs writing programs?
>>
>>Certainly I believe that will happen.  I believe that before long, gene therapy
>>will eliminate most disease.  But it will _only_ happen after the complete
>>genome is mapped along with all deviations that cause problems.
>
>Although perhaps TOO off-topic, I should "warn" everybody that our descendents
>may run out of oil, coal, and nuclear fuel before all these good things happen.
>[Solar Power Satellites may be their only salvation.]  But I share your
>enthusiasm regarding that genome business.  It is almost as interesting as
>computer chess!  : )
>
>>The point is,
>>once you understand something, you can fiddle with it to make it better.  But if
>>you don't understand it, fiddling depends on serendipity to make it better, and
>>there is just as good a chance it will end up worse.
>
>Seriously, that is my experience too.  But I rarely "understand" well enough.
>But don't knock the "trial and error" method.  That's what Thomas Edison is
>supposed to have done to produce a practical light bulb.  You do like light
>bulbs, don't you? : )

Yes, but that is a bit different.  He _knew_ what he was trying to do, find
a material for the filament that would produce significant light and last a
reasonable length of time.  His "trial and error" was within a tight
experimental setup, with the only variable being "material".

We aren't "there" yet...  Everything is a variable and how each variable
interacts with others is totally unknown, which is a problem...


>
>>
>>I can write a program for any process I understand fully, no matter how
>>complex it is.  But for the human mind, we are simply not "there" yet...  I
>>fully believe we _will_ get there at some point in time however....
>>
>>>
>>>It is not clear to me that the computers [or robots] of the future will NEVER be
>>>able to cut the umbilical cord to their human "parents."
>>
>>Nor to me...  Just not yet...  :)
>>
>>
>>>
>>>At the risk of offending human chess programmers, let me suggest that smart
>>>computers or robots will create the chess engines of the future.  The humans
>>>will be told to stay home and keep out of trouble.  : )
>>>
>>>Bob D.
>>
>>
>>That may well happen at some point in time...



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.