Author: Amir Ban
Date: 06:48:03 01/28/98
Go up one level in this thread
On January 28, 1998 at 00:41:26, Don Dailey wrote: >This is degrading into a philisophical discussion but here goes: > Yes, no hash techniques discussed in this thread :) >As computers do more and more amazing things we will >tend to keep changing our definition of what intelligence is (so as >to leave them out!) This is what keeps happening with the whole >field of AI. They keep doing interesting things and then it gets >redefined as not really being AI. Maybe at some point the ultimate >Robot will take a good look at us humans and say nah... they do >some clever tricks but I wouldn't call them intelligent! > This has been said before but I don't think it's true. I think a necessary condition for intelligence (possibly also sufficient) are the basic cognitive tasks that not only humans but all the higher animals do at the pre-conscious, inate level (I wrote this in more detail in another thread). I didn't invent this, and I probably don't have anything original to say on this. I just read it in various sources and when I think about it it appears self-evident. These are the basic tasks that every AI researcher would dearly love to achieve but has no clue as to how. I would define AI as the discipline that, given that artificial intelligence has not been achieved yet, aims at achieving it and researches into methods to do that. AI people should be forgiven, actually even commended, for going in many confused directions in search of their target, since they don't yet know where it is and must experiment: Expert systems, computer chess, neural networks, whatever. However, since the target is so hard and elusive, AI people have the tendency to fall in love with their research topic and stick to it, losing sight of the fact that it is no longer a useful direction in search of the real target, and some of them insult our (natural) intelligence by claiming that what they are doing is already artificial intelligence achieved. I think the AI community is quite right to say that computer chess is not an AI subject. It used to be a hot and legitimate AI topic, but once it stopped contributing to the goal, they righty dropped it. We are not doing AI, whatever the origin of the methods we use, because we are not aiming to achieve intelligence but only in making our programs stronger to win games. Amir
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.