Will computers ever solve chess?

Those who are familiar with solving difficult combinatorial problems understand that the highly arbitrary nature of the rules of chess mean a small proof is utterly infeasible.

No. You use the words "easy" and "possible" in ways that are incompatible with their meanings.

It is not easy to fly over Everest by flapping your arms. Indeed, the stronger statement that it is not possible to do so is also true. And the same applies to solving chess using conventional computing.

Is it possible for this thread to die?.... Oh, if it were only that easy...  

I've never followed but that doesn't stop it from continually appearing on the first page of topics. [BTW, I wasn't trying to be snarky. I'm just impressed how long and how active it is. I'll go back and add a smilely.]

Which is why just because something is impossible, doesn't mean it can't be done. People have always done things that are impossible. I also think that the person or people responsible for solving chess wont be particularly good chess payers, or even know how to play. The answer might come as a byproduct of doing something else.

Agreed. That's one reason why it is unlikely that this haphazard collection of comments contains anything that resembles a proof.

Here's the problem with that...sometimes there are gems among the debris.  The less debris you allow to build up, the easier it is to find the gems.

It would be just as easy to start a new thread titled "The fantasy of solving chess in our lifetimes...let's all dream together of unimaginable technologies", and move the discussion there.

The OP's question is asked and answered.

Side thought to the side thought:

Sitting on top of a mountain (in Chile or no) pretending to understand the world does not make one a guru.  Nor does being a fountain of surface-level platitudes.

...probably 50 times with answers that contradict each other.

You.  You are.  You, who does not understand the world and has demonstrated such repeatedly.

Draw by repetition! Applies to threads as well as Chess. 

Ummm...okay, let's try replace-and-review:

Me, the one who has repeatedly demonstrated that I don't [know how to make spaghetti] am somehow pretending to [know how to make spaghetti]?

This is what doesn't make sense to you?  Past your ability to ken?  Un-grok-able?

s3bog, you said "I think that a viable approach to solving chess using existing hardware, might be realized here."

This implies either that you haven't been following the more informed parts of this discussion or you have ignored them entirely.

Not sure what you mean by "demonstrated something to be the case"

What "something to be the case" have you demonstrated?

Do you recall the part where I inferred that even if storage and time constraints were not fatal, the physical limits of computation imply that solving chess using a conventional computer would require more power than the human race uses for an impractically long time?

Of course, that's based on known algorithms. It hasn't been proved generally for all algorithms.

The discovery of a new shape which tiles a plane by Mann, McLoud, and Von Derau is a good example where a problem was solved not by an increase in computing power, but by the development of a new algorithm.

I don't think it's been proven that all algorithms must fail in solving chess. Just as chess has many games, there are many algorithms. Proving that all algorithms will fail in solving chess could be difficult.

Which is to be trusted in selecting the best move?

The best program making it's analysis given any amount of time?

Or, the best GM's given unlimited time at a consensus?

I think herein lies much of the controversy. How to trust a computer program that is programmed by humans.

ICCF Correspondence chess has pretty much proved that a chess engine plus a human plays better than a chess engine alone.

A computer program must be programmed by humans--humans are the only ones who know how to do a computer program [unless there are aliens from another planet who can do this] 

By the way, the answer to your question is that the best GMs given unlimited time for a consensus would have the edge over the best program given the same amount of time. This is because the best humans with a computer play better than a computer alone.

However, even though the best GMs would have the edge--both the humans and the chess programs would come to the same conclusion--that chess is a draw with optimum play.  [the best humans have already concluded this]  

In a few years, a computer will prove 1.f3 and 2.g4 is the best opening for white!!!

I think:
(human + engine) > (engine) > (human)