ChatGPT

[op12no2]

ChatGPT cannot play chess but you can list a series of moves and ask it for a reasonable move - often get nonsense but it's interesting. And of course you can do things like this lol:-

> write a poem about a chess engine that lost all it's games to a squirrel

In a world of machines and code,
A chess engine reigned supreme.
It played with skill and grace,
A master of the digital board.

But one day, a challenger appeared,
A squirrel with a clever mind.
It watched the engine's every move,
And deftly countered every time.

The engine tried its best to win,
But the squirrel was too quick.
It outmaneuvered every trap,
And left the engine in a pick.

The engine's reign was at an end,
Defeated by a furry foe.
But still it sat and pondered,
Wondering where it went wrong.

And though it lost that fateful game,
The engine learned and grew.
For in defeat, it found its strength,
And a new path to follow.

So let this be a lesson,
To all who think they're strong.
For even the mightiest of machines,
Can fall to something small.

https://chat.openai.com/chat

[op12no2]

> explain the relative strengths and weaknesses of the chess engines stockfish and leela in the form of a shakespearean sonnet

Two engines of chess, Stockfish and Leela
Both strong in their own unique way
Stockfish's brute force, unmatched and unthwarted
Leela's creativity, always astute and astray

Stockfish's deep analysis, a strength without measure
Leela's flexibility, a weakness to treasure
In positions where calculation is key
Stockfish reigns supreme, a true master of the game

But when a creative solution is needed
Leela's skills are unmatched, indeed
Both engines are powerful, no doubt
But in different ways, without a doubt

So whether you need strength or creativity
Choose wisely between Stockfish and Leela
For both have their strengths and their weaknesses
In the game of chess, they are true goddesses.

[lithander]

This is uncanny.

If we assume Google's chatbot Lamda is operating on a similar quality level I'm not surprised that Blake Lemoine got convinced to talk to a person.

[op12no2]

> write a funny story about a chess engine programmer who still uses hand coded evaluation while everybody else has moved to nnue networks. the programmer likes coffee and whisky.

Once upon a time, there was a chess engine programmer named Bob who stubbornly refused to move on from hand-coded evaluation. All of his colleagues had switched to using NNUE networks, but Bob insisted that his hand-coded engines were just as good, if not better.

Bob spent hours upon hours pouring over lines of code, sipping on coffee and whisky to keep himself going. He was determined to prove that his hand-coded engines could compete with the best of them.

But as time went on, Bob's engines began to fall behind. They were unable to keep up with the creativity and flexibility of the NNUE-powered engines, and Bob's colleagues started to laugh at him.

"What are you doing still using hand-coded evaluation, Bob?" they would say with a chuckle. "That stuff went out of style ages ago!"

But Bob didn't care. He was convinced that his hand-coded engines were the future of chess, and he was determined to keep working on them.

And so, while the rest of the chess engine world moved on to bigger and better things, Bob continued to labor away in his little corner, sipping on coffee and whisky and dreaming of the day when his hand-coded engines would finally be recognized as the greatest in the world.

[op12no2]

This is uncanny.

If we assume Google's chatbot Lamda is operating on a similar quality level I'm not surprised that Blake Lemoine got convinced to talk to a person.

I've been getting it to write python code to train nets and write funny stories for by grandson. Amazing stuff. As a demonstrator it's very impressive and imagine all the front ends... Twitter is awash with screenshots from amazed people

[op12no2]

> write a funny story about a chess engine programmer who used javascript. his engine was, in time, able to beat state-of-the-art nnue engines. he was amused but got kidnapped by the fbi, who though he must have access to alien technology.

Once upon a time, there was a chess engine programmer named Tim who decided to use JavaScript to create his chess engine. All of the other programmers were using more traditional languages, like C++ and Java, but Tim was determined to try something different.

To everyone's surprise, Tim's JavaScript engine was able to hold its own against the state-of-the-art NNUE engines. It was able to find creative solutions to difficult positions, and it was able to out-calculate its opponents in long games.

The other programmers were amazed, and they started to wonder if Tim's engine had access to some kind of secret technology. They began to whisper among themselves, speculating that Tim must have stumbled upon some kind of alien technology that was powering his engine.

Tim, meanwhile, was amused by the whole situation. He knew that his engine was just plain old JavaScript, nothing more and nothing less. But he didn't mind the attention, and he continued to work on his engine, always pushing the boundaries and improving its performance.

But then, one day, Tim was kidnapped by the FBI. They had heard the rumors about his JavaScript engine, and they were convinced that he must have access to some kind of secret technology. They whisked him away to a secret facility, where they interrogated him for hours, trying to get him to reveal his secrets.

But no matter how hard they pressed him, Tim refused to break. He knew that his engine was just plain old JavaScript, and he wasn't going to give in to their demands.

In the end, the FBI had no choice but to let Tim go. They realized that he didn't have any access to secret technology, and they were forced to admit that his JavaScript engine was just a product of his own hard work and ingenuity.

Tim was released, and he went back to his work, still amused by the whole situation. He continued to develop his JavaScript engine, always pushing the boundaries and proving that anything is possible if you put your mind to it.

[smatovic]

> explain the relative strengths and weaknesses of the chess engines stockfish and leela in the form of a shakespearean sonnet
[...]

+1

> write a funny story about a chess engine programmer who still uses hand coded evaluation while everybody else has moved to nnue networks. the programmer likes coffee and whisky.

Once upon a time, there was a chess engine programmer named Bob...
[---]

What a meanie, I thought he was fishing.

--
Srdja

[op12no2]

What a meanie, I thought he was fishing.

An amusing coincidence

[op12no2]

> consider how stable diffusion could be used to help to evaluate chess positions

Stable diffusion is a mathematical concept that can be used to analyze complex systems and study how information or other quantities propagate and evolve over time. In the context of chess, stable diffusion could potentially be used to help evaluate positions by simulating how different moves and variations might unfold and affect the overall stability of the position.

For example, one could use stable diffusion to study how the value of a given position might change as a result of different moves by the players. By simulating the diffusion of this value over the course of the game, one could potentially identify positions that are more stable and likely to result in a favorable outcome, as well as positions that are more volatile and susceptible to sudden changes in value.

Another potential application of stable diffusion in chess is in the analysis of tactical patterns and motifs. By simulating the diffusion of various attributes or characteristics of the position, such as material balance or control of key squares, one could potentially identify common patterns and motifs that are indicative of good or bad positions, and use this information to guide the evaluation and decision-making process.

Overall, while the use of stable diffusion in chess evaluation is still a largely unexplored area, it has the potential to provide new insights and enable more sophisticated and accurate analysis of positions and games.

[smatovic]

> consider how stable diffusion could be used to help to evaluate chess positions
[...]

Maybe you can rephrase it with:

"Stable Diffusion deep generative neural networks"

> consider how stable diffusion deep generative neural networks could be used to help to evaluate chess positions

--
Srdja