x
Emergent Alignment Framework for Current Reasoning Models and a Scalable Architecture for Future AGI // Draft / seeking feedback — LessWrong