1
OK, I can partly explain the LLM chess weirdness now
(“make LLMs play better with one weird trick”)
science and existential angst
Visit dynomight.net → RSS feedOn Hacker News Feb 2021 – Jul 2026
Life, Learning & Writing 35Science & Space 29Society, Policy & Culture 18Tools & Productivity 7History & Retro Computing 3AI & Machine Learning 3
(“make LLMs play better with one weird trick”)
Why not benchmark with physical experiments?
are they good or bad?