This category contains 6 tracked predictions. Each page includes the original claim, current evidence, counterevidence, and what would change our assessment.

200,000 AI coding agents run in parallel by March 2027 Not Yet Testable
March 2027 Updated 2026-11-07 40% confidence in assessment

OpenBrain runs 200,000 Agent-3 copies in parallel by March 2027 (page 12). Appendix I, linked from June 2027, gives 250,000 copies for a later date.

Coding agents begin transforming professional work Confirmed
Mid 2025 Updated 2026-11-07 85% confidence in assessment

Specialized coding agents begin to transform professional work, providing substantial time savings in some settings despite continuing unreliability.

Coding fully automated but research taste still requires humans Emerging
March 2027 Updated 2026-11-07 80% confidence in assessment

Research taste remains difficult to train even when coding is fully automated; the two separate scenario excerpts appear below in source order.

RE-Bench score reaches 1.3 Confirmed
Early 2026 Updated 2026-11-09 70% confidence in assessment

1.3 on RE-Bench matching top expert humans given 8 hours at well-defined AI research engineering tasks.

Superhuman coder emerges Not Yet Testable
March 2027 Updated 2026-11-07 60% confidence in assessment

A superhuman coder (SC): an AI system that can do any coding tasks that the best AGI company engineer does, while being much faster and cheaper.

SWE-bench-Verified score reaches 85% Confirmed
Mid 2025 Updated 2026-11-09 90% confidence in assessment

We forecast that mid-2025 agents will score 85% on SWEBench-Verified. (page 3, footnote 3. Note: the 72% starting point is tracker-added context, not from the source.)