Simon Möller

This post is meant to summarize the difficulties of AI Safety in a way that my former self from a few months ago would have found helpful. I expect that it might be a useful framing for others, although I’m not saying anything new. I thank David Lindner for helpful...

I’m grateful to Bogdan Cirstea, Konstantin Pilz and Raphaël S for providing feedback on this post. This post tries to clarify the concept of situational awareness, in particular with respect to current large language models. What is situational awareness Not writing anything new here, just summarizing prior work. (It’s worth...