Q1 2026 AI Timelines Update: Faster Progress in Agentic Coding and Automated RD

In the latest update from the AI Futures Project, researchers Daniel Kokotajlo, Eli Lifland, and Brendan Halstead have revised their timelines for significant milestones in artificial general intelligence (AGI), particularly focusing on automated coding. The updates reflect a faster-than-expected pace of progress, driven by new model evaluations and refined time horizon metrics.

Key Changes

1. Shorter Timelines for Automated Coder (AC) Milestone
   • Daniel’s median forecast for the AC milestone has shifted from late 2029 to mid 2028.
   • Eli's forecast has also moved similarly, indicating a more optimistic outlook on when AGI systems will outperform human software engineers.

Technical and Methodological Updates

• Switch to METR Time Horizon v1.1

  • The team adopted the latest version of the Metaculus Estimated Time Remaining (METR) time horizon metric, which shows faster trends compared to its predecessor.

• Inclusion of New Models

  • Data from recently evaluated models such as Gemini 3, GPT-5.2, and Claude Opus 4.6 were incorporated into the analysis.

• Revised Doubling Time Estimates

  • Daniel and Eli adjusted their estimates for the present doubling time of the METR time horizon:
    • Daniel: From a 5.5-month median to 4 months.
    • Eli: From a 5.5-month median to 4.5 months.
  • These adjustments were based on:
    • The faster trend observed in METR’s new v1.1 data.
    • Continuation of the rapid trend seen in models released since 2024.
    • Further analysis of existing data points.

• 80% Time Horizon Requirement for AC

  • Daniel revised his median estimate for the 80% time horizon requirement for AC from 3 years to 1 year, largely due to the impressive capabilities of Claude Opus 4.6.

Real-World Impact

While these technical updates are significant, the real-world implications are even more striking. Coding agents have seen a dramatic increase in both usefulness and popularity:

• Claude Code’s Rapid Growth
  • Just nine months after its release, Claude Code achieved an annualized revenue of over $2.5 billion.
  • This success is part of Anthropic's ongoing trend of 10x-ing annualized revenue each year, which has now reached the $10 billion range.

Why It Matters

The METR coding time horizon trend, despite its limitations, remains a crucial piece of evidence for forecasting coding automation. The rapid growth in this metric suggests that the development and deployment of agentic coding systems are advancing more quickly than previously anticipated.

For practitioners and researchers, these updates highlight the need to stay vigilant and adaptive. The shift towards shorter timelines means that the industry is moving closer to a point where automated tools will not just augment but potentially replace human software engineers in certain tasks. This transition could have profound implications for both the tech industry and broader economic landscapes.

Source