Safety
15d agoNew Theory Links Alignment Failures To Value Generalisation
A new proposal on the AI Alignment Forum argues that most alignment failures stem from poor value generalisation. The author claims this specific deficit is a fundamental reason why alignment remains difficult.
The Signal
Practitioners should evaluate whether their current safety frameworks address this core generalisation gap or merely treat its symptoms.