Chess AI: Competing Paradigms for Machine Intelligence

TL;DR

This paper compares two leading chess engines, Stockfish and LCZero, using a classic endgame puzzle to evaluate their problem-solving approaches and explores broader implications for AI and AGI.

Contribution

It introduces a novel comparison of chess engines using endgame studies and discusses theoretical applications of Bellman's equation for AI optimization.

Findings

Stockfish outperforms LCZero on Plaskett's Puzzle

Engine differences reflect distinct problem-solving methods

Discussion on AI implications for imagination and general intelligence

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.