Author: Danniel Corbit
Date: 11:34:39 10/01/98
Go up one level in this thread
On September 30, 1998 at 15:20:52, Robert Hyatt wrote: >On September 30, 1998 at 14:12:22, Alessio Iacovoni wrote: [snip] >>Maybe I haven't expressed myself correctly... of course a program developed in >>1998 is stronger than one dating back to 1990, or even closer in time. What I >>meant to say is that we will reach a point, if we havent already, in which chess >>programs, instilled with that chess knowlwedge that they were lacking in the >>past, will not require anything else for improvement than sheer processing >>speed. What programs in the bast were basically lacking was a better positional >>understanding.. now they have that (see HIARCS and basically all of the other >>ones.. including the so called "fast searchers" such as Crafty), plus they have >>an outstanding tactical capacity. So.. HOW ELSE can they be improved by >>software? What I meant was that we have probably come at a point in which >>"everything has already been done" and now the baton has to pass on to hardware >>improvements... >>I don't know if there have been any studies of this kind but could a progam like >>Hiarcs.. or any of the strong ones that everybody has at home.. beat kasparov at >>long times with a pentium XX "1000" or something of the sort and 100mb of mem >>for hash tables? >> >>Alessio Iacovoni > >I would tend to disagree. Until a program plays "perfect chess", programmers >will find holes in its play and continue to fix them. I haven't slowed down, >for example, in adding things to my evaluation, nor changing them as I watch how >they behave. There is a *long* way to go before reaching this point, IMHO... It has been touched on elsewhere [in another thread] but programs have to be rewritten, or *at least* recompiled to take advantage of the new architectures. Try running a program from 1984 compiled for an 8088 on a new machine compared with one tailored to the new architectures and you will see enormous differences. I think, in order for computer programs to really be on a par with the great human players, the software area is equally important to hardware. This is simply because it takes a lot of pure computing power to overcome a problem which is ~O(n^28). If we can find a way to make the problem smaller, that will make an enormous difference.
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.