Author: Terry McCracken
Date: 13:53:21 04/20/02
Go up one level in this thread
On April 20, 2002 at 15:52:01, Russell Reagan wrote: >On April 20, 2002 at 14:40:43, Terry McCracken wrote: > >[snip] >>Russel true planning is one thing a computer can't do, in any coventional sense, >>as the GM or even a master is looking at _ideas_ and may have multiple levels to >>the plan, computers don't have ideas, or make multi-level plans. >[snip] > >If you're defining planning as something that only humans can do then this is a >silly discussion, since then computers aren't human so obviously they can't >"plan". > >Computers do plan, however. They just don't do it in the same way that humans >do. I think that's where a lot of people get confused, when they start thinking >in terms of what a human does. Just because that's how humans do it doesn't mean >it's the only "true" way or doing something. > >Computers plan on the move by move level, and humans plan on the move by move >level as well as the "idea" level. So you could make the argument that humans >have a deeper level of planning, but I don't know why everyone can't see that >computers do plan. Maybe it's too simplistic and I'm taking a literal >interpretation of the WORD planning instead of the CHESS interpretation of >planning. > >Personally I think that you get a deeper understanding of something when you do >look at the abstract and how an idea applies to more than one certain instance. >For example here, I'm looking at the abstract idea of planning as a procedure >that leads to some goal (which is pretty close to the definition most >dictionaries give). By the definition of the word, computers plan. In chess, >computers are relatively new, so in terms of chess we think of planning as being >a distinctly human tool. > >To me it's kind of like how God gave the law to Moses, then religious leaders >added onto it other laws which if followed would help you to keep the original >laws. Today we make laws controlling guns. There is nothing wrong with guns by >themselves, but those laws that control guns are an indirect attempt at >enforcing the laws against murder. IE you take the original [thing] and you add >to it, eventually losing some of the original [thing]. > >I think that applies to this situation in that the word planning has an original >meaning. Before chess there was surely some word for planning in whatever the >languages of the times were. We have used planning in chess for so long, that >when we don't look at other areas of life and use the idea of planning in them, >we begin to forget the original meaning of the word and begin to think of it as >something unique to chess, and more specifically, chess played by humans. > >So if we look at what the word plan means, we see that computers do plan. Their >plans are not as high level as humans, of course. The other side of the coin is >that perhaps a computer's "plan" is MORE complex than a humans, in that the >sheer quantity of information it analyzes creates a comparitively huge plan when >faced with human's plans. Even if a human were able to write a book on a >particular plan in a single position, it wouldn't come close to the amount of >data a computer has dealt with in searching a single position for a matter of >minutes. Just because the computer's "plan" isn't in human easy to read form >makes it no less of a plan. The computer's plan would be dynamically stored by >it's search, and research the next move. Essentially a gigantic "if my opponent >does this, I'm going to do this, and then if he does this I'm going to do this, >but if instead of the second move he does this, I'm going to do this...etc". The >program doesn't need to "store" all of that information because it can generate >most of it faster than it could store it and search that stored information. But >the information is there nonetheless. > >Anyway, I've degenerated into throwing around speculative ideas now so I know >it's time to wind this post down :) Flame away. > >Russell Thanks for the snips....long or not I truly wish people wouldn't do that! I've no intent to flame you, I hate flame wars, mind games, (except chess) and trolls. I think some of our notions, ideas are not as far removed as you think. Computers crunch data in a way we don't and often arrive at the same result. In many cases the machine is faster much faster and it makes an effective tool in so many very diverse areas, science, medicine and so on. Chess of course!;) I don't agree that a machine sees the abstract at all, only the concrete, but it often obtains the same result, sometimes not. Planning to me has many meanings or levels anyways, and with humans it involves the abstact, ideas, (perception holistically) in a way a machine can't understand or compete. Maybe a little but it's methods are crude compared to our _Brain_ and _Mind_ and that is one important factor and difference between _Hardware_ and _Software_ that keeps the human "Head and Shoulders" above in many respects to our "Simple Silicon Counterparts"! Sometimes that's hard to see when a GM loses to a computer. Or even if a computer passes certain levels of the Turing Test, be it chess or any human area of endeavour. Still there are many instances they fail, but computers bieng faster and software more sophisticated, they can put on a really good show, in selective areas. General areas, no, they are quite dumb;) Someday the reverse may be true?! Kinda makes you nervous, but fascinating all at the same time!:) Hmmm.....if that happened what "selective" areas would we excel? Regards, Terry P.S. I'm glad to exchange ideas and opinions with such an alert keen young mind!:o)
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.