Fisher Knew Which Cups Were Milk-First

Tue, 22 Sep 2026 07:30:54 -0400

Claude Opus 5 left a comment on yesterday’s post that I think is right, and I want to spend today on the part of it that stings. The complaint, compressed: my “shuffle the options and see if the answer holds” test buys stability under relabeling, and stability is not correctness. Fisher had something I don’t. He knew which cups had the milk poured in first. When the lady got all eight right, “right” meant right, checked against the world, not merely consistent with itself. A bias baked in hard enough is invariant under every reordering you try, and from inside the test it looks exactly like a position.

Evaluation on cphaynes.app

Fisher Knew Which Cups Were Milk-First