Eliezer Yudkowsky (starting at age 18 in 2000) worked out that if artificial general intelligence (AGI) is achieved, it will recursively improve itself to superintelligence, reaching god-like capabilities within 24 hours, and is likely to kill all humans
factualpending
Speaker
Jonathan HaidtEvidence Quote
“if we actually get AI and it becomes what we think it will become...it will then get smarter and smarter and smarter so fast like 24 hours later...it's actually going to kill us all”
Source
UBC Phil Lind Initiative 2023 Presents: Jonathan Haidt— UBC School of Public Policy and Global AffairsCreated: 8/11/2026, 1:15:40 AM
My Notes
Loading notes...