AI Futures Project Updates AGI Timeline Forecasts to Mid-2028 for Automated Coding Milestone
Key Takeaways
- ▸Automated Coder (AC) timeline accelerated from late 2029 to mid-2028, with similar 1.5-year compressions for TED-AI milestones
- ▸Recent model evaluations (Gemini 3, GPT-5.2, Claude Opus 4.6) and faster METR v1.1 trends drove revised doubling-time estimates from 5.5 months to 4-4.5 months
- ▸Claude Code achieved $2.5B annualized revenue in 9 months, with Anthropic maintaining 10x year-over-year revenue growth, validating accelerated real-world progress
Summary
The AI Futures Project has released a Q1 2026 timelines update, significantly accelerating its forecasts for key AI milestones. The project's researchers have moved their median estimate for achieving Automated Coder (AC)—the point at which AI companies would prefer to replace human software engineers with AI systems—from late 2029 to mid-2028, representing approximately a 1.5-year acceleration. Similar shifts were observed for Top-Expert-Dominating AI (TED-AI) timelines, which also compressed by roughly 1.5 years.
The updated forecasts are driven by several converging factors: new model evaluations (Gemini 3, GPT-5.2, and Claude Opus 4.6), faster-than-expected progress on METR's coding time horizon metric, and revised doubling-time estimates from 5.5 months to 4-4.5 months. Real-world evidence has further supported these accelerated timelines, with Claude Code achieving over $2.5 billion in annualized revenue within nine months of its release, and Anthropic continuing its pattern of 10x annual revenue growth. The project's analysis suggests that if real-world progress continues at roughly 65% the pace depicted in the "AI 2027" scenario, the AC milestone will be reached in 2028.
The researchers emphasize that while their methodology has limitations, the METR coding time horizon represents the strongest available metric for forecasting automation progress. They remain cautious but note that respected AI researchers continue to predict even faster timelines for automated AI R&D capabilities, suggesting the field may be approaching transformative capabilities sooner than previously modeled.
- Project plans quarterly timeline updates going forward and maintains AC as primary metric for forecasting agentic AI capabilities
Editorial Opinion
The AI Futures Project's accelerated timelines reflect genuine progress in agentic AI systems, particularly in coding automation where real-world adoption metrics validate theoretical models. However, researchers appropriately acknowledge the inherent uncertainty in forecasting—their willingness to update "in both directions" suggests intellectual honesty about a rapidly evolving field. The convergence between theoretical modeling and commercial metrics like Claude Code's adoption provides stronger evidence than forecasts alone, though extrapolating from current trends to full AGI capabilities remains speculative.
