🔥 MAJOR TIMELINE REVISIONS: AI FUTURES PROJECT
AI Futures Project: Q1 2026 Timelines Update
Source: AI Futures Project (Daniel Kokotajlo, Eli Lifland, Brendan Halstead) Published: April 2, 2026
Headline: Timeline forecasts have shifted significantly shorter.
📊 Key Revisions
| Metric | Previous (Jan 2026) | Now (Apr 2026) | Shift |
|---|---|---|---|
| Daniel’s Automated Coder median | Late 2029 | Mid 2028 | ~1.5 years sooner |
| Eli’s Automated Coder median | Early 2032 | Mid 2030 | ~1.5 years sooner |
| TED-AI median | — | 1.5 years sooner | Accelerated |
| METR doubling time | 5.5 months | 4 months | Faster |
🧮 What Is “Automated Coder”?
Definition: The point at which an AGI company would rather lay off all their human software engineers than stop using AIs for software engineering.
Significance: This is the milestone where AI becomes more valuable than human engineers for coding work.
📈 Reasons for the Revision
- METR Time Horizon v1.1 adoption — New trend is faster than previous v1.0
- New model evaluations — Gemini 3, GPT-5.2, Claude Opus 4.6 data included
- Impressiveness of Opus 4.6 — Daniel revised AC requirement from 3 years to 1 year
- Claude Code explosion — $2.5B annualized revenue in 9 months
- AI company researchers doubling down — “Automated AI R&D is coming soon”
💬 Key Quote
“Progress in agentic coding has been faster than we expected over the last 3-5 months. The METR coding time horizon trend has its flaws, but we still consider it the best individual piece of evidence for forecasting coding automation. On that metric, growth has continued at a rapid pace.”
🚨 Additional Developments
OpenAI’s Brockman: “70% to AGI”
Source: Research & Development World, April 2, 2026
Greg Brockman stated OpenAI is “70% to AGI.” A prominent cognitive scientist (NYU professor emeritus) disputes this, saying “we’re nowhere close.”
Significance: First explicit percentage estimate from OpenAI leadership on AGI progress.
Microsoft Pivots to Independent AGI Pursuit
Source: The Meridiem, April 2, 2026
Key Details:
- Mustafa Suleyman transitioned from AI integration chief to dedicated AGI development
- 9 months of planning before public announcement
- Renegotiated OpenAI contract “unlocked Microsoft’s ability to pursue superintelligence”
- Microsoft now becomes OpenAI’s direct AGI competitor
Quote:
“The partnership model served the AI deployment phase. The AGI race demands proprietary capability.”
Significance: Microsoft is no longer dependent on OpenAI for frontier AI. This signals hyperscalers believe AGI is close enough to warrant independent development.
DeepMind’s “Project Mario” (Background)
Source: Colossus Magazine (excerpt from “The Infinity Machine” by Sebastian Mallaby)
Key Details:
- DeepMind tried for 4 years to establish AGI safety governance
- External oversight committees proved ineffective
- Hassabis conclusion: “More realistic to have influence within the company than design security mechanisms from outside”
Less timeline-relevant, but provides context on AGI safety governance struggles.
📈 Updated Timeline Comparison
| Source | Timeline | Confidence | Change |
|---|---|---|---|
| AI Futures Project | Mid 2028 (AC) | Forecasters | NEW: 1.5 yrs sooner |
| Greg Brockman | ”70% to AGI” | High | NEW: explicit percentage |
| Internal Lab Planning | 2027-2028 | High | Unchanged |
| Jensen Huang | NOW | High | Unchanged |
| Sequoia | 2026 | High | Unchanged |
| Elon Musk | End 2026 | Aggressive | Unchanged |
| Sam Altman | 2027-2028 | High | Unchanged |
| Dario Amodei | 1-3 years | 90% | Unchanged |
✅ Homepage Status
Homepage remains valid. Current “~1 year” estimate aligns with:
- AI Futures Project: Mid 2028 AC milestone (~2 years)
- Brockman: “70% to AGI” — significant progress already made
- Microsoft pivoting to independent AGI — race intensity increasing
No changes required.
Sources
- AI Futures Project: Q1 2026 Timelines Update
- Research & Development World: Brockman interview
- The Meridiem: Microsoft AGI pivot
- Colossus Magazine: Project Mario excerpt