What makes you say that? Everything I've seen says AlphaZero beat SF8 easily: https://en.wikipedia.org/wiki/AlphaZero#Chess_2

Because that was several years ago. SF has improved since then.

To elaborate, in the Top Chess Engine Championship (which is also used as a benchmark in the AlphaZero vs Stockfish comparison) Stockfish has won all seasons since May 2020[1,2]. A common runner up for those seasons is LCZero[3], which is an engine that is derived from/a reimplementation of AlphaZero. Stockfish is also ahead of LCZero in Fischer random chess (as hosted at TCEC) winning 4 out of the 5 FRC tournaments.

As for how close Stockfish 8 is to a current version of Stockfish, Stockfish 15 has ~400 more ELO than Stockfish 8 as measured in their own self-tests[4].

However, when making these comparisons between AlphaZero and Stockfish it might be tempting to frame it as machine-learning-engine vs classical-engine, which is not true. Stockfish incorporates a neural network when evaluating chess positions[5].

[1]: https://en.wikipedia.org/wiki/Top_Chess_Engine_Championship#...

[2]: https://en.wikipedia.org/wiki/Stockfish_(chess)#Top_Chess_En...

[3]: https://en.wikipedia.org/wiki/Leela_Chess_Zero

[4]: https://github.com/glinscott/fishtest/wiki/Regression-Tests#...

[5]: https://stockfishchess.org/blog/2020/introducing-nnue-evalua...

Great summary! Also, at [1], one can hear Deepmind hinting at that state-of-the-art engines (Stockfish, Leela Chess Zero) will beat Alphazero.

[1] https://www.youtube.com/watch?v=N2MncpRMnFA&feature=youtu.be...

Stockfish 15 is much stronger than SF8.

Not the point.

But that's just the latest Stockfish, after LeelaZero implemented AlphaZero strategy and beat Stockfish, and then Stockfish had to add neural networks to its code to stay competitive.