philosopher of science: statistics, machine learning. leader of research group "epistemology and ethics of machine learning" at u. tübingen. banner: alex katz.
My paper with @se_zezulka "From the Fair Distribution of Predictions to the Fair Distribution of Social Goods" was accepted at @FAccTConference. You can find a preprint here:
https://t.co/NrWtVTRoqo
@no_earthquake I tried lecturing in an full "anatomical theater" type lecture hall thinking it would be no big deal, just like other lectures ... Nope. Completely unexpected wave of anxiety as soon as I saw all those people stacked on top of me.
@peligrietzer People who witnessed the last wave of AI hype talked this way too. My PhD advisor: it's possible to put out a fire with gasoline, if you use enough gasoline.
Lot of cool books continuing to come out on the history and ideas of logical empiricism.
Among them: Alan Richardsons' Logical Empiricism as Scientific Philosophy, Sahotra Sarkar's The Vienna Circle: The Story of Logocal Empiricism, and Adam Tuboly's Otto Neurath in Britain.
@peligrietzer But Shannon info can't be about coding, or anything cognitively interesting, since it's just a transformation of probabilities. (Unless you give a subjective interpretation to the probabilities)
The probing/interpretability research assumes, reasonably, that LLMs must have internal organs. But what if they didn't? What would be the empirical signature of a body without organs?