Will computers ever solve chess?

I am not sure. The problem likely comes down to the no cloning theorem, so you have to be very careful how you transform the state, containing superimposed information at every stage. I tentatively hypothesise that if you can express it in a way where every step is reversible - i.e. unitary transformations - that would suffice. I have not done that above (rather assumed that it would be possible instead).

I think the computer should always be able to find the best moves in a position by calculating all the possible responses and so on, but I think it would take a very long time for it to calculate everything.

Correct me if I have missed something though.

Just the content of the previous 7000+ posts.

#7006
Per GM Sveshnikov it takes 5 years of computer time with human input
No need to calculate all black moves: 1 is enough if it leads to a table base draw
No need to calculate all white moves: reasonable moves e.g. 4 is enough

None of these are true.  You've just decided to hitch your wagon to a dead guy that made a completely unsupported claim.

Here's the state of the thread:

Chess will not be solved in our lifetimes, nor in the foreseeable future of expected technology advances.  People can posit otherwise, but right now it's science fiction.  If they make some very big breakthroughs in quantum computing, then the long term future may change, but not the present/short term future.  Quantum computers can't even play Pong as things sit now, much less solve chess.

#7009
Your state is not the state of the thread.
Being dead is no reason to be wrong. Einstein, Newton... all dead.
The profession of GM Sveshnikov was to analyse chess, and in this century also the elder GM use engines to do so.
Facts and figures support the truth of his claim.
You can have a different opinion but you are the one who provides no support .

I have come to a different conclusion -  that the problems are practical rather than it being possible they are fundamental.

All you need (easy to say) is a reversible operators for each of the steps I illustrated and then to apply those steps to the massively superimposed state dealing with all positions repeatedly. While I only described the parts of the state that are of use, every destructive operation can be replaced by one that is reversible - it adds extra inputs and outputs to preserve enough to reconstruct the original state.

For example an NAND gate is destructive. What we want is a gate which provides NAND outputs on one line, but has other lines which make it reversible. Because NAND has output 1 for 3 different states of the inputs, this is impossible with a 2 output gate, but there is a 3 input, 3 output gate called the CCNOT or the Toffoli gate which achieves this.  This is the simplest universal, reversible quantum gate and suffices (in principle) to efficiently implement all quantum computations (including ours).

I think computers might solve chess at some point, not soon, but I believe chess is a draw with perfect play from both sides

Yes.

With all the self-appointed brainiacs in this thread you’d think it would have already been accomplished. I mean no offense, but all the arguments are very subjective…. And boring. 

Have no fear!  The site has had a few members that claim to have solved chess, and have solved how to beat the strongest engines.  The only problem is they dont disclose their knowledge, refuse to play anyone, and...well...you get it.

Very poor analogy.

If anyone not participating has been following the discussion they would know that there are no quantum computers big enough to solve chess. Indeed it is an open question whether there ever will be. But we (some of us) have been discussing the question of whether such a computer could be programmed to solve chess. I have my view, but would not claim the question is undeniably answered.

Which analogy?  Pong?

That one was not supposed to be a solid analogy, just a humorous statement that illustrates that quantum computing is still in its infancy, for all the potential it has.  I've been programming since I taught myself as a kid, and I am all for quantum computing, but it's not there, even with Toblerone gates (yes, that was more humor, and yes, I knew about the gates, in fact I posted the same table of gates in another thread where Blueemu and I were trying to educate tygxc...).

The existence of the Toffoli gate should actually make my point.  Quantum computing will have to do backflips to emulate conventional programming logic and to get around limitations. 

"Here's a new kind of abacus:  it is awesome, but you need to know how to use it...you can't flip the beads back and forth with your finger, you have to throw it in the air as fast as you can several times until it lands on the ground with the right answer...yes, it will let you know when it has the right answer..."

I guess one *could* formulate even more precise analogies, but once you incorporate a double slit experiment or something then we've lost 85% of the people reading.

#6998
"not necessarily more than 147 qubits"
#7015
"there are no quantum computers big enough to solve chess"

Google operates a 72 qubit quantum computer at Bristlecone
https://thenextweb.com/news/google-reclaims-quantum-computer-crown-with-72-qubit-processor 
DWave is already at 5000 qubit annealing
https://www.hpcwire.com/2019/02/27/d-wave-previews-next-gen-platform-debuts-pegasus-topology-targets-5000-qubits/ 

So...in your world, a preview of a plan (not even a working test model mind you) is the same as "already at 5000 qubits"?  Or maybe you aren't even reading what you post.  Hint:  it's a fluff marketing piece with no actual content or real achievement being announced of any kind.  A couple pages that all say "this company is going to do something they haven't even started yet, invest some money!"

This kind of massive overreaching is right in line with all your claims, so I guess it's not really surprising.

#7015
The method you describe is overambitious: an attempt to strongly solve chess, i.e. generate a 32 men table base. Weakly solving like checkers is enough and requires far less positions.
However, the 72 qubit computer of google at Bristlecone suffices to generate a 9 men or even 10 or 11 men tablebase.
First initiate all positions as U. Then seed the known 7 men table base results W/D/L. Then apply your method to generate the 8 men table base. Then use this to generate the 9 men table base.
men            positions   qubits
32               4,41E+33   112
31               9,13E+38   129
30               2,52E+42   141
29               4,98E+44   148
28               6,32E+45   152
27               1,73E+45   150
26               1,63E+44   147
25               1,18E+43   143
24               7,73E+41   139
23               4,60E+40   135
22               2,51E+39   131
21               1,26E+38   127
20               5,79E+36   122
19               2,46E+35   118
18               9,59E+33   113
17               3,45E+32   108
16               1,14E+31   103
15               3,42E+29    98
14               9,37E+27    93
13               2,32E+26    88
12               5,18E+24    82
11                1,03E+23   76
10                1,81E+21   71
9                  2,77E+19   65
8                  3,64E+17   58
7  4037164893744720  52
6      36641451949418  45
5          261516365480  38
4              1376431148  30
3                    4750192  22
2                          8064  13
Total              8,73E+45 153

Please note that 94.67% of these positions is illegal, less than 1 in a million is sensible, and are not yet compressed for symmetry by a factor 8 like syzygy.

Seems to be a misunderstanding. I was responding to the immediately previous post, not yours. Now edited to include quote.

Note: my understanding of the issues is at a level where you should check everything I say.

The 147 qubits was the absolute lower limit for the representation of a general chess position. It is the nature of quantum devices that the state of such a device could be the superimposition of all positions. While it is easy for me to provide such a lower limit based on simple counting (done by others) this is merely the starting point.

Unfortunately the computational power of a real quantum computer is not determined merely by the number of qubits. This is because they are not perfect: noise reduces their capabilities in a way that IBM has quantified using "quantum volume", combining the number of qubits and the noise level in a way which indicates computing power (by indicating the size of the quantum circuit that can be implemented successfully).  This is analogous to the way the capacity of a communication channel and the noise level can be combined to calculate the maximum bit rate of communication.

Here is how to measure quantum volume for real!

As I write, the largest quantum volume achieved is 1024, and I believe this means the maximum size of quantum circuit that can be implemented is a square array of 10 x 10 two qubit gates.  This is more relevant to the limits on quantum computing that the impressive qubit counts from DWave (undoubtedly with a fair amount of noise that doesn't matter so much for some applications).

UPDATE: IonQ has claimed it will achieve a quantum volume of 2^22 which (by I infer from the simpler examples) means a quantum circuit of 22 x 22 quantum gates.  However, there has been no confirmation of this - they are probably more interested in their IPO as the first publicly traded quantum computing company. Their roadmap is ambitious, aiming to get to quantum volume of 2^1024 by 2028.

How big solving chess would require is far from clear to me, but it seems highly likely it would be around that level or higher. Both in terms of the capacity for representation and the capacity for computation seem in the right ball park. I suspect there will be other higher priority tasks queued up to harness the worlds most powerful computer.

#7021
In 2022 IBM plans to release its 433-qubit Quantum Osprey system.