Is chess programming now boring ???

[Dann Corbit]

From "Giraffe: Using Deep Reinforcement Learning to Play Chess" by Matthew Lai

"We also investigated the possibility of using probability thresholds instead of depth
to shape search trees. Depth-based searches form the backbone of virtually all chess
engines in existence today, and is an algorithm that has become well-established over
the past half century. Preliminary comparisons between a basic implementation of
probability-based search and a basic implementation of depth-based search showed that
our new probability-based approach performs moderately better than the established
approach. There are also evidences suggesting that many successful ad-hoc add-ons to
depth-based searches are generalized by switching to a probability-based search. We
believe the probability-based search to be a more fundamentally correct way to perform
minimax.
Finally, we designed another machine learning system to shape search trees within
the probability-based search framework. Given any position, this system estimates the
probability of each of the moves being the best move without looking ahead. The system
is highly effective - the actual best move is within the top 3 ranked moves 70% of the
time, out of an average of approximately 35 legal moves from each position. This also
resulted in a signficant increase in playing strength.
With the move evaluator guiding a probability-based search using the learned evaluator,
Giraffe plays at approximately the level of an FIDE International Master (top
2.2% of tournament chess players with an offcial rating)12."

1

[AlexChess]

What a Generative AI says about the use of AI to develop a chess engine:

How to use Generative AI to write a chess program:
Generative Artificial Intelligence (AI) offers various possibilities for writing a chess program:

1. Move generation:
AI can generate candidate moves using convolutional neural networks (CNNs) that analyze the chessboard and evaluate positions.
Search algorithms like the Monte Carlo Tree Search (MCTS) can be used to explore different move sequences and evaluate their effectiveness.

2. Position evaluation:
Deep neural networks (DNNs) can be trained on a dataset of chess games to evaluate the strength of a position.
DNNs can also be used to estimate the probability of winning for each player in a given position.

3. Developing playing strategies:
AI can be used to develop general or specific playing strategies for certain openings or playing styles.
Machine learning algorithms like reinforcement learning can be used to optimize playing strategies based on experience.

4. Game analysis:
AI can be used to analyze chess games and identify mistakes or weaknesses in a player's game.
AI can also be used to suggest improvements and variations to the moves played.
Tools and libraries:

There are several open-source libraries and frameworks that can be used to develop chess programs with generative AI:

PyTorch: https://pytorch.org/ : Machine learning library with support for neural networks and machine learning.
TensorFlow: https://www.tensorflow.org/ : Open-source machine learning library developed by Google.
JAX: https://www.jaxwritessongs.com/ : High-performance machine learning library for research and education.
Stockfish: https://stockfishchess.org/ : Open-source chess engine with support for position analysis and move search.

[laurietunnicliffe]

Damn you AI.
You think your taking over. Right!
Just remember, I can always pull out your plug !!!

[shadyantra]

I somehow agree with this post. A reality cannot be curbed or suppressed. Modern Chess has been bread winner for many players and even programmers.

Initially, I thought of reviving the oldest board game fully playable duing Kautilya Chanakya period) but due to work pressure and family commitment, I have stopped creating Youtube videos. People love to play fastest variant of modern chess.

Lack of funds is another big reason, but I still congratulate all programmers for managing talkchess.com forum and portals like chess.stackexchange.com.

Once, I am free with mortgage and liabilities, I will give time to dedicate the portal for amateur players.

[towforce]

...all the "juicy stuff" was being invented/tested/implemented all that time ago...

...I have missed the golden age of chess development...

We are ABSOLUTELY NOWHERE NEAR THE FINAL FRONTIER.

Two suggestions:

1. Proof machine for a given position

a. Generate a list of conditions that must exist for a win

b. Generate a list of conditions for each condition in list (a) above

c. Keep building this condition tree, terminating each branch when the new condition generated already exists in the tree

If this works, you will have built a machine to determine whether, in a given position, conditions for a win exist.

2. Find chess's deep underlying patterns. The NNs are finding a large number of shallow patterns - not a small number of deep ones. Likewise, books about chess tactics are full of shallow patterns. Find a deep pattern that can be used to accurately evaluate any chess position, learn this chess pattern, and become the first player to go from unrated to world champion in one year!