Use of words like “learning” leaves non-scientists with the impression that ML algorithms are doing something completely different from model fitting algorithms developed in other scientific fields.
— 🔥Kareem Carr | Statistician 🔥 (@kareem_carr) March 31, 2024
We don’t fully know what human learning entails at a mechanistic level. We’re just equating it to our current best metaphor (AI algorithms) like earlier thinkers equated the human mind to the behaviors of clockwork mechanisms.
— 🔥Kareem Carr | Statistician 🔥 (@kareem_carr) March 31, 2024
The Harmfulness Of Projection
