An interactive 1.5-hour workshop focused on AI forecasting and safety awareness held in multiple locations across the U.S. on Apr 27th. Find your own local location below! If you don't see your location on here and would like to run a workshop, please see the signup form at the bottom of this page.

Our goal is to help participants understand the urgency of AI safety through interactive demonstrations and group exercises aimed at helping people forecast what might be coming over the next year or two with AI and AI safety.

Participants meet up with their local in-person groups -- the first portion will be a nation-wide virtual intro and demo, and then local groups will break out into their own sessions.

A guided programming session where we'll go over how to programmatically access modern LLMs and how to design evaluations for them.

Our goal will be to replicate a paper that Anthropic put out earlier this year called "Alignment Faking in Large Language Models".

We'll have a central virtual call for everyone to dial into, but individual locations may have separate small groups where people will be programming together.