AI Policy Wiki
Dashboard

Eli Lifland

medium confidence · updated 2026-07-25

Forecaster at the AI Futures Project, co-author of AI 2027 and AI 2040: Plan A, and co-author of the group's published grading of AI 2027's first-year predictions. His median for full coding automation is in the early 2030s, later than co-author Daniel Kokotajlo's.

Eli Lifland is a forecaster at the AI Futures Project and a co-author of its principal scenario publications. He is one of the five authors of *AI 2027* (April 2025), a co-author of the AI Futures Model update published December 2025 with Daniel Kokotajlo, Brendan Halstead, and Alex Kastner, and one of the six authors of AI 2040: Plan A (July 2026).

Forecasting positions

Lifland's published median for full coding automation is in the early 2030s, with a median takeoff speed of about two years (Grading AI 2027's 2025 Predictions (AI Futures Project, February 2026)). Both figures are longer than Kokotajlo's — a 2029 median for full coding automation and a roughly one-year takeoff — and the divergence is stated explicitly in the co-authored grading post rather than smoothed into a group view.

With Kokotajlo he co-authored the February 2026 assessment of AI 2027's 2025 predictions, which reported aggregate quantitative progress at roughly 65% of the scenario's predicted pace and an implied takeoff window of mid-2028 through mid-2030. The post discloses an error in the trajectory graph published with the original scenario and volunteers a potential counterexample to one of the scenario's qualitative predictions. See AGI Timelines.

Relationships