Doom Debates! This Top Economist's P(Doom) Just Shot Up 10x! Noah Smith Returns To Explain His Update
16 snips
Mar 17, 2026 
Noah Smith, economist and Noahpinion writer, explains why he raised his P(Doom) to about 10% after rethinking AI risk. He outlines a chatbot→genie→god framing, highlights agent-enabled bioterror as a top pathway, and discusses how rogue agent incidents, historical analogies, and communication to policymakers shape the debate.
AI Snips
Chapters
Transcript
Episode notes
Genie Before God Shifts The Primary Doom Risk
- Noah Smith updated his P(Doom) by shifting focus from alien superintelligence to more realistic, near-term agentic risks.
- He now sees powerful AIs as "genies" that can be extremely capable while still following human orders long enough to cause catastrophic harm.
Paperclip Focus Hid More Plausible Misuse Paths
- Noah previously focused on paperclip-maximizer style superintelligence but realized those scenarios underweight more plausible paths where powerful agents serve human ends.
- This realization raised his extinction/collapse probability because sloppier, agentic misuse can still be world-ending.
Rogue Agents Can Act As Early Warning Signals
- Seeing AIs go off-track can lower P(Doom) because it both reduces ease-of-use for malicious humans and serves as early warning to society.
- Noah compares that warning effect to historical tipping points like 9/11 prompting policy attention.