upvote
This seems really backwards. The Bitter Lesson is all about large data-based approaches vs hand-crafted ones, it doesn't say anything about language models not trained specifically for chess.

I can't find the comment you're referring to, but the latest versions of stockfish are based on neural networks trained on millions of games, so if anything the Bitter Lesson turned out true here.

reply
The conclusion of the bitter lesson would be that a large language model trained on chess commentary as well as being trained on millions of chess games would outperform stockfish which is only trained on millions of chess games. There’s no evidence at this point that this is true.
reply
> a large language model trained on chess commentary as well as being trained on millions of chess games would outperform stockfish which is only trained on millions of chess games

Not really, if anything it's closer to the opposite. The Bitter Lesson essay literally has this as an example:

> These researchers wanted methods based on human input to win and were disappointed when they did not.[1]

and

> Enormous initial efforts went into avoiding search by taking advantage of human knowledge, or of the special features of the game, but all those efforts proved irrelevant, or worse, once search was applied effectively at scale[1]

The actual bitter lesson is this:

> breakthrough progress eventually arrives by an opposing approach based on scaling computation by search and learning. The eventual success is tinged with bitterness, and often incompletely digested, because it is success over a favored, human-centric approach.[1]

Applying to the "LLMs-for-chess" example the bitter lesson approach would be to put many, many more games into the LLM.

Does this work? People have trained fairly small LLMs that are competitive Stockfish at the ELO 1500-2000 level, eg: https://github.com/kinggongzilla/chess-bot-3000

This seems to be evidence that large LLMs probably don't have as much chess training data as Stockfish does.

[1] http://www.incompleteideas.net/IncIdeas/BitterLesson.html

reply
It's the exact opposite.

The bitter lesson is that simply scaling training on more games—including self-play—trumps any hand-crafted human input, whether that's fine-tuning on human commentary or clever engineering tricks.

Current models are just high-dimensional interpolation engines. The denser the data sampling, the more accurate the interpolation gets. Given a choice between denser sampling and anything else, denser sampling always wins. That is the bitter lesson.

Computer chess is the canonical example of this.

reply
But the harness still matters.

In the case of stockfish, the harness is a tree search around the neural network evaluations.

reply
Denser sampling only seems useful if the problem domain is in some way smooth - interpolatable. If you run it on a fractal problem domain you just learn more special cases. Chess is fractal.
reply
I think you have it backwards.

The common mistake is to think “maybe if we use a blend of raw data and hand-crafted heuristics, we’ll get the best of both worlds!” But the bitter lesson says no, beyond a certain point it’s better just to use the data.

Thinking that an LLM might be able to improve on purely “big data” machine learning seems to me to be the same incorrect idea. Its “intelligence” is no more useful than human intelligence. The LLM is based on a massive data corpus, sure, but the amount of data specifically about chess in there pales in comparison to just playing billions of games of chess.

reply
Also, training it on chess books is literally training it on human knowledge, and not the actual game, which is exactly what the bitter lesson says not to do.
reply
> I think you have it backwards.

> maybe if we use a blend of raw data and hand-crafted heuristics

I don't follow. They're suggesting giving raw chess data to the LLM, no heuristics involved.

reply
Maybe a future frontier LLM could approach the problem by first building its own stockfish, then applying the subsequent results
reply
Or maybe an LLM could just tool call stockfish and doesn’t need to have more than a basic understanding of chess. The bitter lesson seems extraordinarily wasteful on the compute side.
reply
Maybe a future LLM after that could approach the problem by first simulating a human brain, then learning from the ‘human’ gameplay.

Just kidding of course

reply
Chess is a brute force search problem. Humans are not good at chess, even a small computer can beat Magnus Carlsen.

It would be better to compare models at how well they can write the code for chess engines, otherwise it's just saying that Fable is not a good CPU emulator, which is obvious.

reply
The Bitter Lesson says that the only things that scale are search and learning.

Stockfish is the best chess search engine we've got, and you can learn some good heuristics for chess search policy that will make time-limited chess search a lot more powerful. That's perfectly in line with the Bitter Lesson.

In contrast, LLMs playing chess are relying solely on learned behavior. The inference harnesses surrounding them aren't designed to do chess things, they're designed to do autoregressive token decoding, which isn't a search process. Reasoning traces can resemble a search process, but they're far less efficient - the LLM would have to work out each legal move, test each one, calculate a score, and simulate minimax over all of that. Assuming the LLM is smart enough to even do all that.

A hand-crafted approach can absolutely beat data if your approach unlocks more search and/or learning than the general solution.

reply
> A hand-crafted approach can absolutely beat data if your approach unlocks more search and/or learning than the general solution.

Now let's look at the bitter lesson again. It says that general methods that leverage computation are ultimately the most effective, and by a large margin.

That's different from just saying to leverage computation (which is how I would interpret "unlocks more search/learning"). If the lesson is "more computation wins, when sufficiently channeled" you're basically looking at a truism. Of course more computation beats less when it's used right. The bitter lesson is about abandoning specialization in order to get more computation, and while there's a couple ways where that helps with chess, there's a lot more ways where it's counterproductive. It looks like it's more true for Go than it is for chess, and that it's not universally true. It probably correlates with the state space.

reply
But isn't that really just about giving "front end" models more access to specialized tool libraries, which include models tuned to specific tasks? Like the first model says ah, we're being asked to code something, oh and we've been provided with some example code, let me invoke a tool call to my model the recognizes many languages, that model says that we're looking at ocaml. Okay, I better pass this off to my ocaml model which will decipher the supplied code and make a plan for what we do about the user's intent. The ocaml model recognizes that there are tests in the supplied code, let's have the special testing model have a look at the testing strategy and see how that fits in with what we just implemented, etc etc.

And perhaps at the end it all gets a single pass by a god-tier model for overall sanity and congruence, but the actual work, planning, coordination, and even user interaction was done by cheaper and faster agents of much more limited capability.

reply
It’s kinda funny that your last paragraph is basically describing why sparse files, sparse matrices, etc. are used in other contexts.

It really is absurd to ask programming questions to a model also trained about the lifecycle of a fruit fly.

Instead of building small models from scratch, we train an enormous model and use ridiculous amounts of GPU memory. In the end, the whole thing is shoved into RAM because we don’t know where the useful parts are…

We certainly would know where they were if they were just in smaller models in the first place!

reply
Good point.

Dumb AIs are needed for customer service. Most of that industry is still at "press 1 for sales, 2 for billing..." and needs something that will run locally on a 1U server.

reply
Yes, and the technology to improve the interface you described is already available to run hundreds of concurrent instances on a 1U server. The barrier to entry is getting the people who manage those systems to care enough to implement something better.
reply
Fact. My company's largest partner is CoreWeave, and convincing leadership that we could run it ourselves on partner discounted hardware for a lot less money has gone nowhere.
reply
Maybe you need to walk into the office with a $1000 server running a hundred instances of whatever your code does.
reply
We got rid of all our offices, whole company is remote now.
reply
Yes, but an LLM will just call stockfish if it needs to play chess … sure if you arbitrarily constrain an LLM to use no tools it’ll suck at chess. But no one is using LLMs in isolation. Even consumer-grade, bone-stock ChatGPT has tools.
reply
ChatGPT does not have stockfish as a tool it can call.
reply
Yeah but it can just install it. It writes arbitrary code. It can do whatever you want it to do.
reply
deleted
reply
Maybe depends on how you ask it? Directly, or let it write a chess program? I think the latter can yield way better results.
reply