Steelmanning AI x-risk: Debate & sharpen your beliefs

A session designed to sharpen your inside views on AI existential risk by debating its core arguments and counter-arguments.

Maybe you've read the case for why advanced AI could pose an existential threat to humanity, and found it convincing. Or maybe you're skeptical that an AI could ever become autonomous, misaligned, and powerful enough to cause that kind of harm on its own.
Either way: this is a debate with abundant disagreement even among experts. We believe that arguing the case directly (rather than just reading about it) is one of the best ways to find out which side you actually stand on, and whether your reasoning holds up to scrutiny.
In this session, we'll present a set of core premises behind loss-of-control scenarios, where an advanced AI could cause extinction or large-scale harm. Each participant draws a premise at random and has to steelman the objection to it: argue, as a skeptic would, why you don't buy the premise. You'll then defend that objection against a partner arguing for the premise. Then you'll switch and debate their premise instead.
We'll rotate partners and premises a few times, so you get to interact with multiple arguments and people.
Afterward, we'll open it up to a group discussion to share takeaways and what beliefs shifted or didn't.
We hope this helps you sharpen your views and clarify your own uncertainties. The point isn't to land on a particular conclusion (you might leave more convinced of AI x-risk, less convinced, or unchanged), but to go through the argumentation and come out with a stronger sense for why you believe what you do.
We recommend at least some prior familiarity with AI safety and existential risk from AI. You don't need any deep background, just a basic idea of the scenario we're discussing.
The session will be held at UPF Campus Ciutadella, room 13.001.

The building is the following (green):

If you have any questions, feel free to contact us at aisafetybcn@gmail.com.