Don't Worry About the Vase Podcast Book Review: If Anyone Builds It, Everyone Dies
Sep 19, 2025
Dive into a gripping review that unveils the chilling thesis of how superintelligence could threaten humanity. Learn about the alarming pathways to artificial superintelligence and why current AI alignment efforts might not be enough. Discover the pressing need for robust policies to mitigate risks, including global bans and hardware monitoring. The discussion wraps up with pragmatic advice for governments and individuals on advocating for safety measures. Can we steer the future before it spirals out of control?
AI Snips
Chapters
Books
Transcript
Episode notes
Clear, Stark Thesis
- If anyone builds superintelligence under current conditions, the authors claim everyone dies and mean it literally.
- Zvi adds a "probably" but agrees the book gives the clearest long-form explanation for non-experts.
Rhetorical Clarity For New Audiences
- The book presents familiar arguments in a new, clearer rhetorical style aimed at non-experts.
- Zvi finds it highly informative and persuasive for general audiences.
Warnings Can Backfire
- Early MIRI work warned of AGI risks and unintentionally spurred AI investment.
- The authors note those warnings later helped fund labs like DeepMind and OpenAI.
