1. Tell me you trust only humans who always explain you in detail how they come to their beliefs. You are probably very lonely. 2. There is a lot of ongoing wo

1. Tell me you trust only humans who always explain you in detail how they come to their beliefs. You are probably very lonely. 2. There is a lot of ongoing work on mechanistic interpretability by e.g. antropic that shows we can understand LLMs better than we initially thought.