A Crossroad in Computer Chess; Or Desperate Flailing for Relevance

[chrisw]

At least the people interested in NNUE are trying to do their own implementation of NNUE + trainer, as opposed to outright copying the NNUE backend code, so I do not see NNUE being much of an issue in the future. That cannot be said about Allie, which is still using Leela trained nets + Leela backends.

In regards to the wider picture, very few self-developed engines that is not a derivative of Stockfish or Leela would ever reach the heights of Stockfish or Leela, because individual efforts simply cannot compete directly against large distributed projects with massive computational power.

That would be to imply that neither SF nor LC0 are flawed in some way. Obviously they are flawed, LC0 is in trouble to way smaller nets running on AB (who predicted that?) and SF is flawed because evaluation not holistic (proven by LC0).
SF is not superhero and the kit needed to work on smaller nets and generate training games and do tuning is within reach of lone individuals or small teams.

So far, there is only one private engine, Stoofvlees, that comes close to Stockfish and Leela, and even Stoofvlees is weaker than Stockfish and Leela. Examples from Rubichess, Minic, and Igel simply show that it not enough to implement NNUE in your engine and take a sergio vieri net and expect to compete with Stockfish and Leela, the search needs to be good as well, and for most engines, their search is far weaker than Stockfish search to compete with Stockfish and Leela, or even Stoofvlees. So already at the highest levels of computer chess, as the OP said, 'The result is that there are two engines left; Stockfish, and Leela.'

Wild generalisations.
LC0 may well be in trouble to smaller nets using AB. It’s by no means clear that massive stacked towers and MCTS isn’t going the way if the dinosaurs.
SF doesn’t really exist anymore. That NNUE eval kind of belongs to everybody. The SF search is good but it’s not copyrightable and much of it is so well known as to be like Lego. There are other things that can be done and so on.
Anything can happen and usually does.

[Dann Corbit]

I think the largest problem is team size.
Stockfish and LC0 have giant teams.
Teams with highly skilled people who work for nothing.
Teams with sky-high piles of gleaming hardware worth millions of dollars, donating both hardware, time, electricity and effort in testing.

I guess that Andrew is as clever as the most clever person on either team.
But when you have 100 workers, it is hard to compete against that.
And when you have a million dollars of hardware on your team, how will a lone developer compete with that?

It is, in fact, the mega-teams that are squashing the smaller teams.
And the people taking things are taking them from the mega teams because that is the only way they can compete.

I guess that the only small entity who can beat a mega team is one who has a supremely clever idea (like the NNUE thing) and then hides what they did so that nobody can copy it.

If you even say what your approach was, someone else will try it.

[dkappe]

At least the people interested in NNUE are trying to do their own implementation of NNUE + trainer, as opposed to outright copying the NNUE backend code, so I do not see NNUE being much of an issue in the future. That cannot be said about Allie, which is still using Leela trained nets + Leela backends.

You are spreading your misinformation in several threads, I see. The discord handle of the NVIDIA developer who has contributed CUDA code to both Allie and LC0 is ankan. I assume that’s what you are referring to as “leela backend?” To top it all off, the code isn’t even chess specific.

Are you sure your last name isn’t “Besmirchfield?”

[maksimKorzh]

1. Are you making money on chess programming (I mean does the strength of Ethereal affect your income?)
2. If so - then why to get concerned by the fact the world is going to hell when this happens in other aspects but chess programming as well?
3. If not - then again why to get concerned? Your work would be inspiring other programmers for years!

I've not made any money off of Ethereal. In fact, the opposite, as I've spent thousands on testing rigs, and hundreds to host servers for OpenBench.

I believe you should continue your work as if Leela on the one hand and NNUE on the other never existed.
IGNORE modern fancy trends because they are just modern fancy trends.
Go your own way. Always.

You may be right. But refer to the top -- Ethereal is not a monetary thing. The pleasure I derive from Ethereal is seeing Ethereal compete and perform well. It makes me want to improve more. To move up the rankings. Or to play better and faster and more exciting. I see those motivations drying up due to NNUE.

I clearly see your point, Andy, but let me try to express my idea a bit more.
Obviously motivation is VERY important and without it you just can't make yourself keep working.
I want to emphasize that the matter of KEEP WORKING is REGARDLESS of nowadays fancy trends.
There's always a TRUE REASON WHY are you doing this, but quite often our consciousness shifts the emphasis from the true reason to false one.
I don't want to say that higher ranks are false, not at all, I just want to give an example - just imagine that by developing your own engine you're developing your own self and the fact that you LIVE is much more important than rankings. So in simple words I would say it's not the matter of WHAT to do or WHAT to achieve (e.g. top ranks) but HOW instead.

The process of a desire to drop something you've been working on for years is similar regardless of the LEVEL you're doing this at.
Look at my engine for instance - I was happy to beat TSCP))) Now you might wonder why am I talking about this - now that's very important:
THE PROGRESS HAPPENS WHEN YOU BIT YOUR PREVIOUS VERSION, literally when you beat your own weaknesses.
When we die the only thing that would matter is HOW BETTER WE BECAME COMPARED TO WHEN JUST ENTERED THIS WORLD.
So true motivasion should be to GROW, grow against your own self regardless of ranks.

Your win is to CONTINUE, you lose if you quit. If it doesn't make sense for your own self anymore than continue for others. Help people to grow, to reach your level.

Sorry if this was a bit too much, but try to get my intent - it's to support you mentally and to say:
everything is passing through, NN and NNUE is not an exception

[Dann Corbit]

To top it all off, the code isn’t even chess specific.

AH, but it's math.
HOW DARE YOU USE THE SAME MATH AS ME!
THAT MATH IS MINE!

[maksimKorzh]

To top it all off, the code isn’t even chess specific.

AH, but it's math.
HOW DARE YOU USE THE SAME MATH AS ME!
THAT MATH IS MINE!

Ahhh!!!!

[AndrewGrant]

There's always a TRUE REASON WHY are you doing this, but quite often our consciousness shifts the emphasis from the true reason to false one.

You might actually be over estimating how primitive my motivations are. Again I might be wrong and not understand myself. But....

Before working on Chess, I played League of Legends in the top few hundred players. Sometime between Chess and now, I played World of Warcraft in one of the top 10 guilds in North America. I think, a general understanding of myself, is that I do things to satisfy competitive urges. I tend to pick one thing at a time in life, and ruthlessly dedicate myself to it.

I quit League of Legends when the game lost its competitive edge -- when the maker started making choices to benefit the average player, as opposed to maintaining the components of the game that set people apart at the upper echelons of play. I quit World of Warcraft for the same reason.

If I come to the conclusion that Computer Chess is no longer a viable competitive pastime, I may just quit.

[towforce]

I think, a general understanding of myself, is that I do things to satisfy competitive urges. I tend to pick one thing at a time in life, and ruthlessly dedicate myself to it.

The major way to become successful right now is to pick something and apply AI to it. Whatever your next adventure, you should DEFINITELY make big money from it!

[AndrewGrant]

It is, in fact, the mega-teams that are squashing the smaller teams.
And the people taking things are taking them from the mega teams because that is the only way they can compete.

I guess that the only small entity who can beat a mega team is one who has a supremely clever idea (like the NNUE thing) and then hides what they did so that nobody can copy it.

You make a really good point. One which Alayan and I had not stumbled upon in our talks. I'll have to ponder this for some time.

[maksimKorzh]

There's always a TRUE REASON WHY are you doing this, but quite often our consciousness shifts the emphasis from the true reason to false one.

Before working on Chess, I played League of Legends in the top few hundred players. Sometime between Chess and now, I played World of Warcraft in one of the top 10 guilds in North America. I think, a general understanding of myself, is that I do things to satisfy competitive urges. I tend to pick one thing at a time in life, and ruthlessly dedicate myself to it.

I quit League of Legends when the game lost its competitive edge -- when the maker started making choices to benefit the average player, as opposed to maintaining the components of the game that set people apart at the upper echelons of play. I quit World of Warcraft for the same reason.

If I come to the conclusion that Computer Chess is no longer a viable competitive pastime, I may just quit.

You might actually be over estimating how primitive my motivations are. Again I might be wrong and not understand myself. But....

Never meant to offend you. Sorry if that has happened.

Well, if something you're doing doesn't make any more sense without a competitive aspect then you're totally right - it's time to quit because chess would be getting even more monopolistic. It's hard to understand for me that position though because the only thing I compete with is my dumbness)))

Thanks you so much for sharing your previous experience - it makes more sense than initial post in order to understand what is truly bothering you.
Now the very last thing I can say is that life is not only competition and there are many other aspects to grow within other but competitions.
But I respect your position and clearly see your point, hence I wish you to FIND YOUR PATH and take the right decision on your future.
If I had only 1/4 of your skills how many people I could help... omg...

P.S. I'm just wondering why youtube tutorials dedicated to chess programming are done by noobs like and not by guys like you, this is so pity...