Will Quantum Computers solve Chess?

I have a close friend who is a top engineer at a quantum computing manufacturer and he told me that by 2026-7 they will have the worlds first large (several hundreds of logical qubits) commercially viable quantum computer.

I’m worried that such a computer could solve chess and that this would destroy chess as a sport. Some autistic guy could just memorize 40 moves and become the world champion when his real chess strength is at 1900-level.

Perhaps only Chess960 will survive the quantum era.

@Nytorpet said in #1:
> I’m worried that such a computer could solve chess and that this would destroy chess as a sport.

That's a ridiculous thing to worry about. Solving chess would mean nothing, at least in terms of playing the game. Computers are already so much stronger than humans, that compared to humans, they might as well have solved chess.

> Some autistic guy could just memorize 40 moves and become the world champion when his real chess strength is at 1900-level.

That's also ridiculous, and kind of offensive too. By memorizing 40 moves, I assume you mean memorizing 40 moves deep. To do that and consistently be successful, you would need to memorize probably millions (maybe more or less, not sure, but a large number) of positions. It is true that GMs are sometimes able to play their prep for a long time, but again, solving chess won't make much of a difference to humans. In fact, if we are able to solve chess, it would probably be better for humans to prep based on the old fashioned chess engines (using their own judgement as well) because God might play a move that is still technically drawn, but gets into a very difficult position, whereas the engine would likely avoid that (although again, humans would have to use their own judgement as well).

> Perhaps only Chess960 will survive the quantum era.

Solving Chess960 seems to me like a very small step up from solving regular chess.

@Nytorpet said in #1:
> [...] I’m worried that such a computer could solve chess and that this would destroy chess as a sport. Some autistic guy could just memorize 40 moves and become the world champion when his real chess strength is at 1900-level. [...]

It seems to me that we should already be making a clear distinction between chess played by humans and chess played by software regardless of what quantum computing may achieve.

Even if it reaches the point where chess is as good as solved, that should not in itself affect the human game.

The passage I quote above is not a realistic scenario. No-one can master chess by memorizing moves. The number of possible games multiplies exponentially with each move.

EDIT: Sorry, cross-posted with AsDaGo who makes the same points (and others) better than I did.

@Nytorpet

If we use them to store every position we would need more storage than our naming system for it can count

5 pieces db is 939mb the small Syzygy version
6 pieces db is 150.2 GB the small Syzygy version
7 pieces db is 18.4 TB the small Syzygy version

every time it goes up a unit of measurement
If we follow up that tree

8 pieces Pentabytes
9 pieces exabyte
10 pieces zettabyte acording to www.quora.com/How-much-data-can-be-stored-on-all-the-computers-in-the-world#:~:text=Estimates%20suggest%20that%20all%20the,music%20ever%20created%206%20times.
we have 2 of these for all current devices

11 pieces yottabyte

and thats the end of our naming system
so its safe to say it wont happen soon

@AsDaGo said in #2:
”Solving chess would mean nothing, at least in terms of playing the game.”

It would destroy the beauty of the game. Let’s say that we know for certain that the Marshall Attack in Spanish is unsound and how to defeat it. Or if we know for certain that 7.g3. Is a refutation of the Najdorf in the Sicillian. That would make chess a lot more boring.

There are estimated to be more possible chess games than atoms in the universe.
Even if computers could greatly reduce that number by not counting "senseless" moves and games it would still be way, way, way too much information to store. Chess is simply not solvable in practice

@Finnfinity said in #6:
> There are estimated to be more possible chess games than atoms in the universe.
> Even if computers could greatly reduce that number by not counting "senseless" moves and games it would still be way, way, way too much information to store. Chess is simply not solvable in practice
The quantum world is strange and quantum computers derive their power from this strangeness. Nobody knows if quantum computers can solve chess but it’s not an outrageous idea. Solving chess is a mathematical problem so basically all you have to do is to set up an equation and solve it on a computer which is much more powerful than any present day super computer.

@Nytorpet said in #5:
> ”Solving chess would mean nothing, at least in terms of playing the game.”
>
> It would destroy the beauty of the game. Let’s say that we know for certain that the Marshall Attack in Spanish is unsound and how to defeat it. Or if we know for certain that 7.g3. Is a refutation of the Najdorf in the Sicillian. That would make chess a lot more boring.

That seems unlikely, because if there were a refutation to the Najdorf, we likely would have found it by now. I think it is much more likely that openings we currently consider to be unsound (like the Englund gambit for example) are actually holdable with absolutely perfect play. But even if this were found to be the case, it would still be nothing more than an interesting bit of trivia at the human (and perhaps even conventional chess engine) level.

Solving chess would certainly be a highly impressive feat if it is possible, but it would be a much smaller shock to the chess world than the first strong chess engines were.

@Nytorpet said in #7:
> The quantum world is strange and quantum computers derive their power from this strangeness. Nobody knows if quantum computers can solve chess but it’s not an outrageous idea. Solving chess is a mathematical problem so basically all you have to do is to set up an equation and solve it on a computer which is much more powerful than any present day super computer.
They may have better computing power but that's not the issue, it's disk space. You still need to store the data for all these different positions and moves somehow. And I don't think you can find a universal equation to find the best move for every one of these countless, vastly different scenarios.

@AsDaGo said in #2:
> That's a ridiculous thing to worry about. Solving chess would mean nothing, at least in terms of playing the game. Computers are already so much stronger than humans, that compared to humans, they might as well have solved chess.
>
>
>
> That's also ridiculous, and kind of offensive too. By memorizing 40 moves, I assume you mean memorizing 40 moves deep. To do that and consistently be successful, you would need to memorize probably millions (maybe more or less, not sure, but a large number) of positions. It is true that GMs are sometimes able to play their prep for a long time, but again, solving chess won't make much of a difference to humans. In fact, if we are able to solve chess, it would probably be better for humans to prep based on the old fashioned chess engines (using their own judgement as well) because God might play a move that is still technically drawn, but gets into a very difficult position, whereas the engine would likely avoid that (although again, humans would have to use their own judgement as well).
>
>
>
> Solving Chess960 seems to me like a very small step up from solving regular chess.
one of my friends told me if chess is solved, white might have an obvious winning chance. if this happens, chess as a sport would not be played by as many people