Large Language Models and the Critical Brain Hypothesis — AI Alignment Forum