AGI by 2027 is "strikingly plausible"

In the essay "Situational Awareness", former OpenAI researcher Leopold Aschenbrenner extrapolates the growth of compute and algorithmic efficiency and concludes that artificial general intelligence by 2027 is "strikingly plausible".

Forecast: Leopold Aschenbrenner, 2024. Source: Situational Awareness

Illustration: AGI by 2027 is "strikingly plausible" — Illustration: lastday.at. This photo has not been taken yet.

See it on the timeline

lastday.at tells the story of the machines that gradually took over human work, from ancient Egypt to artificial intelligence, and counts down to the last day of work on Earth: March 20, 2030.

Open the full timeline