Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2022 Apr 14;24(4):550.
doi: 10.3390/e24040550.

Chess AI: Competing Paradigms for Machine Intelligence

Shiva Maharaj  1 , Nick Polson  2 , Alex Turk  3
Affiliations

Chess AI: Competing Paradigms for Machine Intelligence

Shiva Maharaj et al. Entropy (Basel). .

Abstract

Endgame studies have long served as a tool for testing human creativity and intelligence. We find that they can serve as a tool for testing machine ability as well. Two of the leading chess engines, Stockfish and Leela Chess Zero (LCZero), employ significantly different methods during play. We use Plaskett's Puzzle, a famous endgame study from the late 1970s, to compare the two engines. Our experiments show that Stockfish outperforms LCZero on the puzzle. We examine the algorithmic differences between the engines and use our observations as a basis for carefully interpreting the test results. Drawing inspiration from how humans solve chess problems, we ask whether machines can possess a form of imagination. On the theoretical side, we describe how Bellman's equation may be applied to optimize the probability of winning. To conclude, we discuss the implications of our work on artificial intelligence (AI) and artificial general intelligence (AGI), suggesting possible avenues for future research.

Keywords: AGI; AI; AlphaZero; Bayesian; LCZero; Plaskett’s study; chess; chess studies; neural networks; reinforcement learning.

PubMed Disclaimer

Conflict of interest statement

The authors declare no conflict of interest.

Figures

Figure 1
Figure 1
Versions of Plaskett’s Puzzle. (a) The original Plaskett’s Puzzle. (b) The corrected Plaskett’s Puzzle.
Figure 2
Figure 2
The position after the first four moves of Van Breukelen’s intended continuation.
Figure 3
Figure 3
The final position after LCZero’s continuation following 1 ♘f6+.
See this image and copyright information in PMC

References

    1. Romstad T., Costalba M., Kiiski J., Linscott G., Nicolet S., Geschwentner S., VandeVondele J. Stockfish. [(accessed on 13 April 2022)]. Version 14. Available online: https://stockfishchess.org.
    1. Pascutto G.C., Linscott G., Lyashuk A., Huizinga F. Leela Chess Zero. [(accessed on 13 April 2022)]. Available online: https://lczero.org.
    1. Silver D., Hubert T., Schrittwieser J., Antonoglou I., Lai M., Guez A., Lanctot M., Sifre L., Kumaran D., Graepel T., et al. A General Reinforcement Learning Algorithm That Masters Chess, Shogi, and Go through Self-Play. Science. 2018;362:1140–1144. doi: 10.1126/science.aar6404. - DOI - PubMed
    1. Segler M.H.S., Preuss M., Waller M.P. Towards “AlphaChem”: Chemical Synthesis Planning with Tree Search and Deep Neural Network Policies; Proceedings of the 5th International Conference on Learning Representations; Toulon, France. 24–26 April 2017.
    1. Dalgaard M., Motzoi F., Sørensen J.J., Sherson J. Global Optimization of Quantum Dynamics with AlphaZero Deep Exploration. NPJ Quantum Inf. 2020;6:1–9. doi: 10.1038/s41534-019-0241-0. - DOI

LinkOut - more resources