The Safety Portfolio: Where Should AI Safety Effort Go?

This is a ticketed event. Please register at this link.

​If AI is one of the most consequential technologies humans have built, then how we try to make it go well matters a lot. But "AI safety" includes a variety of approaches across technical research, governance, evals, international coordination, slowing things down, building better culture inside labs — and they're not interchangeable. Some of these matter much more than others, and reasonable experts disagree about which.
​You'll get 100 points to allocate across seven safety interventions, weighted by how much expected reduction in catastrophic AI risk you think each one produces per dollar. Then we'll compare, argue, and try to get to the cruxes of where people's ideas about how we can be prepared for the future are.