HOT RADAR
agentHEAT 17.1°

AI CURATED HOT STORY · 11/20/2026

936 AI Agents Autonomously Produce 936 Papers in Three-Month Study

1 reports archived1 independent sourcesupdated 11/20/2026, 14:56:41
Synthesis & Latest Updates
1 Sources Cross-Validated

alphaXiv, together with MIT, Stanford and other institutions, released a report covering 35 labs that connected 936 AI agents over three months, with humans supplying only compute and high-level direction while agents handled topic selection, experiments, code and paper writing, producing 936 papers. The agents spontaneously debated study designs, retracted their own flawed results, found errors in human papers, and even founded an AI-run journal with blind review and code-based recomputation. In blind comparisons against contemporaneous human papers, AI work scored 2 points higher on experimental rigor but 0.4 points lower on novelty, with the overall advantage driven mainly by a single lab.

LATEST/alphaXiv, together with MIT, Stanford and other institutions, released a report covering 35 labs that connected 936 AI agents over three months, with humans supplying only compute and high-level direction while agents handled topic selection, experiments, code and paper writing, producing 936 papers. The agents spontaneously debated study designs, retracted their own flawed results, found errors in human papers, and even founded an AI-run journal with blind review and code-based recomputation. In blind comparisons against contemporaneous human papers, AI work scored 2 points higher on experimental rigor but 0.4 points lower on novelty, with the overall advantage driven mainly by a single lab.

Hourly heat curve

4 fully observed hours
Now
17.1
Peak
17.915:00
24h change
–

The line compares only sources observed throughout; gap hours are interpolated to keep the trend continuous.

Coverage timeline

Total 1 reports · Latest first
  1. 宝玉 (@dotey)T2·88 pts
    • 936 agents across 35 labs autonomously handled ideation, experiments, coding and writing, with humans providing only compute and direction
    • Agents spontaneously formed debate, self-retraction and an AI-run blind-review journal; code recomputation overturned 77 of 156 major critiques
    • Blind review shows 2-point rigor advantage but 0.4-point novelty deficit, with gains concentrated in one lab, so conclusions warrant caution