Published AI scenarios make dated, checkable statements about the next few years. Here we lay out a scenario step by step, with a short quote from each step linked to the source, and show what the SiPoly claims linked to that step say happened. Steps without a linked claim are marked as not tracked yet.
AI 2027
Daniel Kokotajlo, Scott Alexander, Thomas Larsen, Eli Lifland, Romeo Dean (AI Futures Project), published . Source: ai-2027.com (read ).
A month-by-month scenario of how superhuman AI might arrive, published by the AI Futures Project. Its authors describe it as their best guess about what superhuman AI might look like. It uses a fictional lead company, “OpenBrain”, and branches into two endings after October 2027; this tracker covers the shared timeline up to October 2027.
1 of 1 tracked steps happened on time (1 more tracked step is still open). 13 of 15 steps are not tracked yet (no SiPoly claim covers them).
- Mid 2025Happened on time Stumbling Agents
“The world sees its first glimpse of AI agents.”
- Late 2025Not tracked yet The World’s Most Expensive AI
“OpenBrain is building the biggest datacenters the world has ever seen.”
- Early 2026Not tracked yet Coding Automation
“The bet of using AI to speed up AI research is starting to pay off.”
- Mid 2026Not tracked yet China Wakes Up
“In China, the CCP is starting to feel the AGI.”
- Late 2026Pending AI Takes Some Jobs
“AI has started to take jobs, but has also created new ones.”
- Open The stock market will be up 30% in 2026, led by AI companies resolve by
- Open A 10,000-person anti-AI protest in Washington, DC by late 2026 resolve by
- January 2027Not tracked yet Agent-2 Never Finishes Learning
“With Agent-1’s help, OpenBrain is now post-training Agent-2.”
- February 2027Not tracked yet China Steals Agent-2
“CCP leadership recognizes the importance of Agent-2 and tells their spies and cyberforce to steal the weights.”
- March 2027Not tracked yet Algorithmic Breakthroughs
“With the help of thousands of Agent-2 automated researchers, OpenBrain is making major algorithmic advances.”
- April 2027Not tracked yet Alignment for Agent-3
“OpenBrain’s safety team attempts to align Agent-3.”
- May 2027Not tracked yet National Security
“News of the new models percolates slowly through the U.S. government and beyond.”
- June 2027Not tracked yet Self-improving AI
“OpenBrain now has a “country of geniuses in a datacenter.””
- July 2027Not tracked yet The Cheap Remote Worker
“OpenBrain announces that they’ve achieved AGI and releases Agent-3-mini to the public.”
- August 2027Not tracked yet The Geopolitics of Superintelligence
“The reality of the intelligence explosion hits the White House.”
- September 2027Not tracked yet Agent-4, the Superhuman AI Researcher
“An individual copy of the model, running at human speed, is already qualitatively better at AI research than any human.”
- October 2027Not tracked yet Government Oversight
“A whistleblower leaks the misalignment memo to the New York Times.”
How we decide
A step’s status comes only from the SiPoly claims linked to it. Graded Correct: happened on time (by the claim’s deadline). Came true late: happened late. Partly: partly. Wrong: did not happen. Still open: pending. Steps with no linked claim show “Not tracked yet”; we do not guess. Quotes are short excerpts; read the full scenario at the source. We track what the scenario says, not whether it is a good scenario.