It's that time again. AI #108: Straight Line on a Graph. thezvi.substack.com/p/ai-108-strai…

please let the record state I quit before sama spoke …

Anthropic is hosting a recruiting social in NYC targeted at the quant trading industry! Signup in thread. I enjoyed trading systems, and Anthropic combines the technical depth of trading with being in the fastest most impactful area of tech.

Excited to have supervised these papers! EM was wild, with unclear implications for safety We answer how: there's a general evil vector. Boosting this is A solution to SFT on any narrow evil task We don't know WHY it's so general, but release better EM models to boost research

Oh, and my favourite part of this project is that Ed and Anna found the core results in a two week sprint!

Really awesome to see Ed and Anna's work on emergent misalignment covered in MIT Tech Review, alongside OpenAI's great new paper