Community Weighs In: Solving Hard Problems Needs 1K–10K Agents

Following OpenAI's announcement, community members estimated that producing a strong solution to hard problems plausibly requires 1,000–10,000 agents, echoed by hands-on symbolic regression experiments needing population sizes of about 10,000.

2026-09-09 ~ 2026-09-09 · 2 related posts

• Symbolic regression in practice: ~10k population needed even without agents — cephaloform · 2026-09-09
• Why 1k-10k agents per strong math solution may be the right order of magnitude — cephaloform · 2026-09-09