Hugging Face Incident + Q&A with an AI Safety Researcher

​A presentation and discussion on the recent Hugging Face incident, what it tells us about current AI systems, and what it could mean for the future of AI safety.

We'll begin with a presentation covering the incident, what happened, and why it matters. The talk pulls together information from many different sources, so you'll probably come away with something new even if you've already heard about the event. From there, we'll move to the wider issues connected to it: alignment concerns, how we should interpret unusual or potentially deceptive model behaviour, and the general pace of AI progress.

After the presentation, we'll be joined for a Q&A by Vili Kohonen, Empirical LLM Researcher at the Center on Long-Term Risk. Vili will share his perspective on the incident, current alignment research, and the broader trajectory of AI development, and answer questions from the audience.