x
Towards formalizing universality — AI Alignment Forum