2 August 20262 h 41 min
Behind AI 2027: superexponential horizons and a patient misaligned AI
With Eli Lifland
Eli Lifland says AI 2027 was meant to be as plausible as possible at the level of detail it commits to, getting the big calls like timelines, misalignment and government involvement right while accepting that specific details would be unlikely.
24 sections · 1329 lines