Future progress of chess software in chess

[Dann Corbit]

AlphaZero's code is written by humans. There was human input everywhere: general design and concept, net format, search algorithm, training game parameters, and so on. Having some part that can improve through automated training is not the same thing at all as humans being useless in the process and contributing nothing at all that would yield more strength than if they didn't intervene.

I think that the problem with #2 and #3 is that nobody is working on it.

You could put a million AI engineers on the issue, and they wouldn't solve these problems any time soon. These problems are much harder than what's currently the bleeding edge in AI sophistication.

I agree with you.
But I also think that software and hardware are both improving exponentially.
So absurdly difficult things happen faster than we imagine that they will.

[towforce]

If I may be permitted to extend the topic slightly, I remember a quote from the first episode of "Humans", a British drama about humanoid robots (link), a girl struggling to get motivated to study for college saying something like, "What's the point of me doing all this work to learn this when this device can answer any questions about the subject (or anything else) in one second?"

The big picture (chess and everything else): in the near future, IMO, the big winners will be those who work with the AI daily, and then do what the AI recommends them to do.

Here's a simplified version of my reasoning:

* the AI will have more time for you than most people

* the AI will be more accepting of you as you are than most people

* the AI will be more resourceful than most people

* the AI will likely have more data about you than most people

* the AI will have more data about people who are similar to you than most people

* the AI will be at least an order of magnitude less expensive than most people

* above all, the AI will have less agency issues (link) than most people

We might be about to experience a world in which a higher proportion of people than ever before are healthy and have good social skills. If that happens, it will be a big win!

[Uri Blass]

1)Chess machines are equal to human+machines in playing correspondence chess
2)Chess machine are equal to union of humans+chess machines in developing better chess engines.
3)Chess machines are equal to union of humans+chess machines in teaching humans to play chess better in OTB chess.

I think that we are very close to 1.
When do you expect 2 and 3 to happen so we need no humans to make engines or to teach chess?

I think that the problem with #2 and #3 is that nobody is working on it.
As for #1, I do not know if anyone has measured that, but I also guess it depends on the player.
Some players will surely contribute to the match in a positive way, some in a neutral way (as described in your #1) and some in a negative way.
So the alternative question:
1)Chess machines are equal to (the world's best human correspondence player)+machines in playing correspondence chess
is really a much more difficult question in my view.

That's an interesting and completely wrong point of view.
The reality is that humans are useless in correspondence chess on top level because no amount of human intervention can score a win against SF on decent hardware on ICCF TC. Ergo, human contribution to engine performance is exactly 0 (ZERO).

I am not sure that no amount of human intervention can score a win against SF on decent hardware on ICCF TC(maybe some human intervention can score a win in 1 out of 20 games) but even if it is correct that humans are uselesee in this task it does not mean that humans are useless in correspondence chess because it is possible that human intervention can do better than SF against some human who use some weaker software than SF or weaker hardware than SF or some bad opening that stockfish by itself is not strong enough to beat without human intervention.

[Milos]

I am not sure that no amount of human intervention can score a win against SF on decent hardware on ICCF TC(maybe some human intervention can score a win in 1 out of 20 games)

No amount of human intervention can score a win against SF on decent hardware on ICCF TC ever.
Regular ICCF standard TC is equivalent to 5 days per move (even without doubling). SF on decent hardware in opening reaches depth of 80 in 5 days. In middle game easily over 100 in endgame 120+.
You really think a human can steer engine to a position where there is a trap that is more than depth of 80??? Sorry but that's just delusional.
One can ofc construct such a position (these days even that is becoming almost impossible except for some endgame positions) but steer an engine there, sorry simply impossible.

but even if it is correct that humans are uselesee in this task it does not mean that humans are useless in correspondence chess because it is possible that human intervention can do better than SF against some human who use some weaker software than SF or weaker hardware than SF or some bad opening that stockfish by itself is not strong enough to beat without human intervention.

Sure there are retards and oldies in ICCF that can't use SF and have a machine that is having 1Mnps. 95% of active ppl with rating above 2100 on ICCF today use the latest SF and hardware that makes at least 15Mnps from a starting position. So your strawman is quite weak there .

[Dann Corbit]

There are wins in correspondence chess, and everyone is using Stockfish and other strong engines to help with the analysis.
There are also losses, and a large number of draws. What part of the losses and draws is the influence of humans? It is hard to say.
I think that very strong humans can add more than you imagine.
They can (of course) be wrong or have a negative influence.
But I do not think the value of human inputs is zero.
For most humans, probably yes, it is zero or negative influence.
But not for all.

[Cornfed]

There are wins in correspondence chess, and everyone is using Stockfish and other strong engines to help with the analysis.
There are also losses, and a large number of draws. What part of the losses and draws is the influence of humans? It is hard to say.
I think that very strong humans can add more than you imagine.
They can (of course) be wrong or have a negative influence.
But I do not think the value of human inputs is zero.
For most humans, probably yes, it is zero or negative influence.
But not for all.

With 'wins' there are 'losses' by definition....

Much of the w/l's are going to be due to...sub optimal opening choices.

[EroSennin]

The big picture (chess and everything else): in the near future, IMO, the big winners will be those who work with the AI daily, and then do what the AI recommends them to do.

And very soon there will be no need for humans to exist. AI might easily exterminate humans just out of boredom.

[EroSennin]

Sure there are retards and oldies in ICCF that can't use SF and have a machine that is having 1Mnps. 95% of active ppl with rating above 2100 on ICCF today use the latest SF and hardware that makes at least 15Mnps from a starting position. So your strawman is quite weak there .

Nowadays with the help of a correspondence book you only need to reach like depth 30 to comfortably draw every game. You can even play different openings.

[Cornfed]

Sure there are retards and oldies in ICCF that can't use SF and have a machine that is having 1Mnps. 95% of active ppl with rating above 2100 on ICCF today use the latest SF and hardware that makes at least 15Mnps from a starting position. So your strawman is quite weak there .

Nowadays with the help of a correspondence book you only need to reach like depth 30 to comfortably draw every game. You can even play different openings.

Clearly the 'future of chess' is...OTB or online one-on-one chess. "Correspondence" is dead.

[mclane]

I would like to talk about the progress of chess software of the past.