Questions about secrecy in AI alignment research

A user argues that AI safety has long faced dismissal despite serious questions about evaluating intelligence beyond human ability, how incentives and correction could support alignment, and how to maintain reliable, scalable oversight. They ask why these questions have been studied “in hiding” and claim alignment conferences are “secret” and invitation-only.