A Conversation with Nate Soares on the Case for Treating AI as Existential Risk

A Conversation with Nate Soares, President of the Machine Intelligence Research Institute and Author of If Anyone Builds It, Everyone Dies

If anyone builds superintelligent AI, everyone on Earth will die. That’s the thesis of the New York Times bestseller by Eliezer Yudkowsky and Nate Soares. Named to the New Yorker’s and Guardian’s best books of 2025, endorsed by figures from Stephen Fry to the former senior director of the White House NSC, and cited in the U.S Congress and House of Lords, the book has had a significant impact on the policy conversation around AI.

But how compelling is the argument? In this moderated discussion, Soares will lay out the case for why the current trajectory of AI development poses an existential threat — and then, we’ll have an opportunity to push back. We’ll try to identify the key assumptions, surface the strongest counterarguments, raise the objections that skeptics and proponents alike are grappling with, and stress-test the reasoning in a way that goes deeper than you’d get from just reading the book. The goal is to come away with a more informed and nuanced understanding of one of the most important questions of our time.

• When: Thursday, March 12, 2026 | 12:20–1:20 PM
• Where: 105 Jackson Meeting Room, Hauser Hall (1st floor)
• Lunch: Provided