Sublime
An inspiration engine for ideas
causal estimates produced by observational studies may be labeled “provisional causality,” that is, causality contingent upon the set of assumptions that our causal diagram advertises.
Judea Pearl, Dana Mackenzie • The Book of Why
In a recent work (https://t.co/85LMTeOJPF), we examined whether LLMs are potential sources of probabilistic knowledge (rung 1 of Pearl's hierarchy), which led to the benchmark at https://t.co/Pla35jJSYU. The answer was no, which was surprising and poses fundamental challenges for various downstream tasks and key capabilities (including explanation,... See more
Elias Bareinboimx.comDeveloping Computational Agency
kaiton • 26 cards


fantastic essay on epistemology 
humans are coherence-maximisers, not truth-seekers
the trouble is that depending on your priors, a coherent explanation - a simple story that fits together well, isn't well correlated with being useful or true - 
https://t.co/XjutOAjCm5... See more
integrity of the system is more important than the truth of one case.
