Back to All Events

What Everyone in Technical Alignment is Doing and Why: Anthropic, OpenAI, DeepMind Safety, Conjecture

We’ll be discussing what everyone in practical safety is doing and why, using Thomas Larsen’s LessWrong article as a jumping off point. There’s a lot to get through, so we’ll be focusing just on the four biggest players: Anthropic, OpenAI, DeepMind Safety, and Conjecture. We’ll try to identify common themes (like LLM alignment) to see where the community is allocating its resources.

Previous
Previous
18 January

Agent Incentives: A Causal Perspective

Next
Next
1 February

What Everyone in Technical Alignment is Doing and Why: CHAI, CAIS, Sam Bowman and MIRI