Keep pulling the thread on Daniel Kokotajlo.
Daniel Kokotajlo's median forecast for the arrival of an Automated Coder (AC) has shifted from late 2029 to mid 2028.
Eli Lifland's median forecast for the arrival of an Automated Coder (AC) has shifted from early 2032 to mid 2030.
Claude Code reached an annualized revenue of over $2.5 billion in early February 2026, nine months after its release.
Anthropic's annualized revenue has continued to grow 10x each year, reaching the $10 billion range.
Daniel Kokotajlo's median forecast for the arrival of Top-Expert-Dominating AI (TED-AI) has shifted approximately 1.5 years sooner.
Eli Lifland's median forecast for the arrival of Top-Expert-Dominating AI (TED-AI) has shifted approximately 1.5 years sooner.
The AI Futures Model forecasts were updated to include data from newly evaluated models Gemini 3, GPT-5.2, and Claude Opus 4.6.
Daniel Kokotajlo revised his estimate for the doubling time of the METR time horizon from a 5.5-month median to a 4-month median.
Eli Lifland revised his estimate for the doubling time of the METR time horizon from a 5.5-month median to a 4.5-month median.
Daniel Kokotajlo revised his median estimate for the 80% time horizon requirement for an Automated Coder down from 3 years to 1 year, influenced by the performance of Claude Opus 4.6.
The METR coding time horizon trend is considered the best individual piece of evidence for forecasting coding automation.
An Automated Coder (AC) will be achieved in 2028 if real-world AI progress continues at approximately 65% of the pace depicted in the AI 2027 scenario.