LessWrong review slams Yudkowsky's AI doom book as vague and unconvincing
A viral critique argues 'If Anyone Builds It, Everyone Dies' fails to substantiate its catastrophic AI risk thesis.
Deep Dive
James Brobin's review on LessWrong criticizes Eliezer Yudkowsky and Nate Soares's book 'If Anyone Builds It, Everyone Dies.' The review argues the 2026 book offers vague, abstract warnings about AI superintelligence causing human extinction but fails to explain how modern LLMs or scaled systems actually create this risk. It calls the work a missed opportunity to ground public AI safety discourse in technical reality.
Why It Matters
Highlights the growing debate over whether AI existential risk arguments are substantiated or alarmist.