LLMs predict my coffee - https://dynomight.net/coffee/
Coding, math, whatever.
#im
Why not benchmark with physical experiments?