We have suffered with unbalanced programs in the past: tactically better almost from the very start, but positionally, uh, stupid. (I don't want to muddle the thread by mentioning any, such as Fritz, by name.) We gave them more time to calculate, and got what we always knew was, to some extent, an illusory positional, or planning, judgement. At that time in chess programming history, until now, all of the programmers, as far as I know, knowingly sacrifice knowlege for speed. I understand that sound tactics were the initial minimum requirement, just as for humans.
But with the exact peak of tactical skill becoming somewhat irrelevant in play against humans, a new challenge arises, which I don't see anyone *dedicated* to (I could be wrong): writing a program that does whatever it needs to do, as slowly as necessary within reason, to find manoevers and plans worthy of a world champion. In contradistinction to the previously described "stupid" programs, this program could be relied upon to show the human how to play the position, with the understanding that it might be overlooking tactics. This would be like having a Grandmaster look over your games with you: judgement, though not gospel or mathematical proof. This program might lose terribly to the others, but if you can convincingly argue for its virtues, people would listen.
If, furthermore, the program could explain its reasoning, you'd have a *natural* Grandmaster-Teacher, whose style could be emulated without reservation.
I think that Ed Shroeder's Encyclopedia of Chess (EOC) was a first step in this direction. At least, it was marketed as such, sufficiently convincingly.
New "paradigm" desired: and we'll pay for it!
Moderator: Ras
-
mongrel
Re: New "paradigm" desired: and we'll pay for it!
With the title of my thread, I was intentionally echoing one of the classic old points of "contention" (in a good way), that regularly appeared in this forum in the somewhat distant past. Was it (Thorsten?) in defence of (Whittington?) and/or CSTal? I don't remember. And I don't feel like digging it up, so much has changed since then.
I do remember that the original arguments, which my previous post touched upon, were, generally, more or less, dismissed at that time. And I probably took the side of those dismissing it. But I'd like to point out that the very same issues are still with us.
I argue that a fundamental change has occurred, or which, if it has not occurred could be argued for anyway, that the top programs now have a surplus of tactical strength (versus humans), which they can *now* afford to sacrifice for the sake of these other values, that were formerly too expensive to slower programs on slower machines.
And I say again, partly in parallel with another thread, that while we've probably lost, through conquest, the precious, near-insoluble, Human vs. Computer epic, we still have this one. And I think some people don't want to talk about it because much of its challenge lies in a different sphere, much as poetry calls for skills different from applied math. I know this type of dread. And while not wanting to overstate this, I also know the type of unknowing dismissiveness that those heavily invested in one thinking style can have for the other: the mathematical mind can miss the poetry. For top performance in any field, including mathematics, both types of thinking must be represented, in balance. In short, in the field of computer chess, I think the Yin has outstripped the Yang, (or vice versa, I'm not an expert). And I think someone could found a career on righting the balance. That is what I am arguing.
I do remember that the original arguments, which my previous post touched upon, were, generally, more or less, dismissed at that time. And I probably took the side of those dismissing it. But I'd like to point out that the very same issues are still with us.
I argue that a fundamental change has occurred, or which, if it has not occurred could be argued for anyway, that the top programs now have a surplus of tactical strength (versus humans), which they can *now* afford to sacrifice for the sake of these other values, that were formerly too expensive to slower programs on slower machines.
And I say again, partly in parallel with another thread, that while we've probably lost, through conquest, the precious, near-insoluble, Human vs. Computer epic, we still have this one. And I think some people don't want to talk about it because much of its challenge lies in a different sphere, much as poetry calls for skills different from applied math. I know this type of dread. And while not wanting to overstate this, I also know the type of unknowing dismissiveness that those heavily invested in one thinking style can have for the other: the mathematical mind can miss the poetry. For top performance in any field, including mathematics, both types of thinking must be represented, in balance. In short, in the field of computer chess, I think the Yin has outstripped the Yang, (or vice versa, I'm not an expert). And I think someone could found a career on righting the balance. That is what I am arguing.