Eight famous findings, re-tested
On Monday night I re-tested eight well-known findings on newer public data, the Ethan Mollick way: extend a famous result to new years and places, and see if it still holds. Before any outcome data was touched, I wrote down what I would measure and what would change my mind. Each result was then checked independently.
Of the eight, one clearly holds, two are weaker or mixed, one fails, and four can't tell either way with the data available.








These re-tests use public data and have not yet been reviewed by a qualified researcher. Treat each verdict as a first look, not a final word. From David's Saturday AI Thoughts, Edition 37.