A Crossroad in Computer Chess; Or Desperate Flailing for Relevance

Discussion of anything and everything relating to chess playing software and machines.

Moderators: hgm, Dann Corbit, Harvey Williamson

Forum rules
This textbox is used to restore diagrams posted with the [d] tag before the upgrade.
dkappe
Posts: 845
Joined: Tue Aug 21, 2018 5:52 pm
Full name: Dietrich Kappe

Re: A Crossroad in Computer Chess; Or Desperate Flailing for Relevance

Post by dkappe » Mon Oct 05, 2020 3:42 am

AndrewGrant wrote:
Mon Oct 05, 2020 2:46 am
Could you please explain it to me then? If I'm missing crucial information, I'm perfectly happy to change my view when presented with additional information which brings my current view into question.
I’m happy to explain things to someone who listens and discusses in good faith. Your posts here and in other threads demonstrate amply that is not true in your case. The prospect of unpacking to you the history of that project and of that particular incident makes me feel indescribably tired and depressed. Now go ahead and have the last word, as is your habit.

AndrewGrant
Posts: 1023
Joined: Tue Apr 19, 2016 4:08 am
Location: U.S.A
Full name: Andrew Grant
Contact:

Re: A Crossroad in Computer Chess; Or Desperate Flailing for Relevance

Post by AndrewGrant » Mon Oct 05, 2020 4:16 am

dkappe wrote:
Mon Oct 05, 2020 3:42 am
AndrewGrant wrote:
Mon Oct 05, 2020 2:46 am
Could you please explain it to me then? If I'm missing crucial information, I'm perfectly happy to change my view when presented with additional information which brings my current view into question.
I’m happy to explain things to someone who listens and discusses in good faith. Your posts here and in other threads demonstrate amply that is not true in your case. The prospect of unpacking to you the history of that project and of that particular incident makes me feel indescribably tired and depressed.
:shrug: I'm here if you want to enlighten me. Otherwise, I'll stick to my story. I obviously don't have complete information. I know as much as that blog post, and the comments that were made public at the time. If there's more to it, you could convince not just me, but many others as well I reckon. I'm willing to entertain any idea, even if I don't agree with it, so long as its argued thoroughly.
dkappe wrote:
Mon Oct 05, 2020 3:42 am
Now go ahead and have the last word, as is your habit.
I would like our exchange to end with me saying "Okay, I see your point of view, I can concede some ground to you", or "I see what you are saying, but I don't see how this changes the original premise", after listening to your side of things, and asking questions to gather more insight.

Tony P.
Posts: 208
Joined: Sun Jan 22, 2017 7:30 pm
Location: Russia

Re: A Crossroad in Computer Chess; Or Desperate Flailing for Relevance

Post by Tony P. » Mon Oct 05, 2020 9:40 am

AndrewGrant wrote:
Mon Oct 05, 2020 1:36 am
Its a shame to see any project use and abuse the work that went into Leela.
There's a big rant below, beware... :mrgreen:

If you fear that your work will be abused, you have other options: 1) not to start, 2) keep the code private, 3) charge for it.

Out of the major deep learning libraries, the only one that's under a GNU license* is OpenNN. Who is still using it? :roll: * Namely, LGPL, the one that obliges devs to allow users to RE the entire code that calls the however small covered lib, to give the users the power to debug mods of the lib instead of simply rolling the breaking changes back and happily using the product as it was :lol:

Google has released Tensorflow under Apache 2.0, but that hasn't prevented it from making money off consulting businesses on the reuse of that lib and renting out TPUs, and indirectly off the reputation boost. It's been a win-win. Google has also released a lot of small codebases supplementing its papers (Alpha Zero was just too large and important to be donated). Likewise, the sponsors of MXNet likely don't regret releasing it under that license, and FB doesn't regret opening PyTorch under modified BSD.

Just by making Leela available to Caruana, Ding and the other super-GMs fully for free instead of ad-hominem freemium coaching services, the devs have missed out on a 5 (maybe 6) figure amount. (Well, that was a consequence of Glaurung's GPL, but they could still have done the Fat Fritz cloud trick, for example.)

Btw, you're absolutely right not to release the training code for the Ethereal nets.

And fwiw, SF isn't a villain, beating it didn't save the humanity :lol: Leela just improved the SOTA a little and gave us a different flavor of chess (the latter had also been done by Fizbo, iCE, Wasp, Rodent et al. without beating SF). Even Komodo isn't a villain, it's affordable and would provide decent chess knowledge even if SF didn't exist.

noobpwnftw
Posts: 467
Joined: Sun Nov 08, 2015 10:10 pm

Re: A Crossroad in Computer Chess; Or Desperate Flailing for Relevance

Post by noobpwnftw » Mon Oct 05, 2020 9:47 am

People like H shouldn't have done what they did, however, there is not so much for "good faith" in computer chess programming anyways, this forum is full of it and I'm gladly contributing my part.

User avatar
towforce
Posts: 10757
Joined: Wed Mar 08, 2006 11:57 pm
Location: Birmingham UK

Re: A Crossroad in Computer Chess; Or Desperate Flailing for Relevance

Post by towforce » Mon Oct 05, 2020 9:51 am

Tony P. wrote:
Mon Oct 05, 2020 9:40 am
Google has released Tensorflow under Apache 2.0, but that hasn't prevented it from making money off consulting businesses on the reuse of that lib and renting out TPUs, and indirectly off the reputation boost...

I'm a fan of Google, and they have released a huge amount of code in many areas into open source - and a lot of it is of very good quality. I don't know why they've done it (I suspect in some cases it stifles potential competitors), but I have personally benefited from using several of their open source libraries. A lot of the time, it just looks like generosity: they've made something good, and they're not going to sell it, so they give it away.
Writing is the antidote to confusion

Kiudee
Posts: 27
Joined: Tue Feb 02, 2010 9:12 pm

Re: A Crossroad in Computer Chess; Or Desperate Flailing for Relevance

Post by Kiudee » Mon Oct 05, 2020 10:25 am

Tony P. wrote:
Mon Oct 05, 2020 9:40 am
Just by making Leela available to Caruana, Ding and the other super-GMs fully for free instead of ad-hominem freemium coaching services, the devs have missed out on a 5 (maybe 6) figure amount. (Well, that was a consequence of Glaurung's GPL, but they could still have done the Fat Fritz cloud trick, for example.)
Interesting, how some people immediately think about how they can make money with something. The phrasing “missing out” somehow implying that it is your duty as a proper homo economicus to extract all the money you can.
One aspect which is overlooked here is that Leela Chess Zero and Stockfish are where they are because they are free and open. Leela would have never been a reality, if it wouldn’t have been supported by the large group of contributors.

User avatar
mvanthoor
Posts: 764
Joined: Wed Jul 03, 2019 2:42 pm
Full name: Marcel Vanthoor

Re: A Crossroad in Computer Chess; Or Desperate Flailing for Relevance

Post by mvanthoor » Mon Oct 05, 2020 2:08 pm

Rebel wrote:
Sun Oct 04, 2020 9:50 am
His (rightly) concern is that engine authors download the latest SF NNUE net and gain a free xx elo. Example, suppose I add NNUE (I won't, but suppose) to ProDeo and get and 150 elo gain on the rating lists because of that. 6-7 months later -- having changed not one bit on my engine -- I download the latest SF net, it gives me a free 30-50 elo, I release it and rub my hands seeing ProDeo climb and climb in the rating lists. I would say that's a very unhealthy situation and surely this (kind of things) is what's going to happen in the near future.

So while NNUE is a fantastic new development regarding increasing strength it has an unwished negative side effect and I can understand that some engine authors already now consider to stop.
The solution is simple. Split the rating lists:

1. One for 'classic' engines without any neural networks.
2. One for 'hybrid' engines that use classic stuff but also some sort of neural network add-on.
3. One for 'full' neural network engines such as Leela.

Then everyone can compete in exactly the space they want; people who want to compete with classic engines with hand-crafted evaluation (with eval tuning as the only automated option) can do so without getting frustrated of seeing other engines pass their own just because they got a NNUE added; if they do, those engines go to the 2nd rating list. People who are into generating and researching different networks can compete with other networks on the third list.
Author of Rustic.
Releases | Code | Docs

fabianVDW
Posts: 146
Joined: Fri Mar 15, 2019 7:46 pm
Location: Germany
Full name: Fabian von der Warth

Re: A Crossroad in Computer Chess; Or Desperate Flailing for Relevance

Post by fabianVDW » Mon Oct 05, 2020 2:50 pm

mvanthoor wrote:
Mon Oct 05, 2020 2:08 pm

The solution is simple. Split the rating lists:

1. One for 'classic' engines without any neural networks.
2. One for 'hybrid' engines that use classic stuff but also some sort of neural network add-on.
3. One for 'full' neural network engines such as Leela.

Then everyone can compete in exactly the space they want; people who want to compete with classic engines with hand-crafted evaluation (with eval tuning as the only automated option) can do so without getting frustrated of seeing other engines pass their own just because they got a NNUE added; if they do, those engines go to the 2nd rating list. People who are into generating and researching different networks can compete with other networks on the third list.
Where is the line? A PSQT is an in NNUE fashion incrementally updated Perceptron. Or do we only start calling it NN when it has multiple layers? Or perhaps when the inputs are more sparse, like a KingxPiece ([64][64]) table(which I've recently added to FabChess)?
Author of FabChess: https://github.com/fabianvdW/FabChess
A UCI compliant chess engine written in Rust.
FabChessWiki: https://github.com/fabianvdW/FabChess/wiki
fabianvonderwarth@gmail.com

Tony P.
Posts: 208
Joined: Sun Jan 22, 2017 7:30 pm
Location: Russia

Re: A Crossroad in Computer Chess; Or Desperate Flailing for Relevance

Post by Tony P. » Mon Oct 05, 2020 5:35 pm

Kiudee wrote:
Mon Oct 05, 2020 10:25 am
Interesting, how some people immediately think about how they can make money with something. The phrasing “missing out” somehow implying that it is your duty as a proper homo economicus to extract all the money you can.
It's rather about bridging the gap between the work we love to do and the work we have to do to maintain ourselves. It's rather about making our hobby profitable so that we can afford to reduce the hours of daytime slavery.

Dann Corbit
Posts: 12000
Joined: Wed Mar 08, 2006 7:57 pm
Location: Redmond, WA USA
Contact:

Re: A Crossroad in Computer Chess; Or Desperate Flailing for Relevance

Post by Dann Corbit » Mon Oct 05, 2020 6:43 pm

Tony P. wrote:
Mon Oct 05, 2020 5:35 pm
Kiudee wrote:
Mon Oct 05, 2020 10:25 am
Interesting, how some people immediately think about how they can make money with something. The phrasing “missing out” somehow implying that it is your duty as a proper homo economicus to extract all the money you can.
It's rather about bridging the gap between the work we love to do and the work we have to do to maintain ourselves. It's rather about making our hobby profitable so that we can afford to reduce the hours of daytime slavery.
I think trying to make money on computer chess is a mistake.
That is because people who are talented enough to write a winning chess engine can make far more money programming something else
Secondly, and more importantly, when you change chess programming from a hobby into a profession, it goes from fun to work
Now you must support it, now you must document it, now you must continue to make progress.
Sucks all the fun right out of it.

Now, there are a special few who can make money at chess programming. But I strongly suspect that they are driven by the intellectual challenge of chess programming and not the money in it, because there's not much money in it
Taking ideas is not a vice, it is a virtue. We have another word for this. It is called learning.
But sharing ideas is an even greater virtue. We have another word for this. It is called teaching.

Post Reply