Author: Sam Lloyd
Date: 10:28:32 03/29/04
Go up one level in this thread
On March 29, 2004 at 11:57:09, Steven Edwards wrote: >On March 29, 2004 at 10:56:46, Robert Hyatt wrote: >>On March 29, 2004 at 05:48:37, Steven Edwards wrote: >> >>>See: http://mitpress.mit.edu/e-books/Hal/chap5/five1.html >>> >>>Any comments on the second paragraph? > >>if we don't do it like humans do it, it isn't AI. Notwithstanding the small >>problem that we don't even know how humans do it in the first place. :) > >Submarines use propellers instead of fins, automobiles use wheels instead of >feet, and (most) chess programs use big tree iterative A/B search instead of >cognition. > >Yet it might be possible to build a submarine with fins; it would be more >complex but could also be quieter and more energy efficient. An automobile with >articulated appendages for locomotion would be slower but could also travel in >extremely rough terrain. A chess program that worked using recognition, >deduction, and introspection might be more prone to blunders, but its techniques >could be ported to other areas of application. > >I agree in general with Campbell's speculations about Hal. It appears that a >machine like that could be given the rules of the game and eventually play well >without explicit domain specific programming. And maybe that is one of the >identifying characteristics of AI: applicability across domains that are loosely >related, if at all. If a machine is created with the albilities of HAL, which may happen within the next 50 years, maybe sooner, than this machine could do both cognition and an iterative A/B search. It would be for all intents and purposes, "incapable of error";-) Lloyd
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.