§ Category
AI & ML
Model releases, benchmark passes, agentic-capability milestones, corporate actions.
§ Primer
Model releases, benchmark passes, agentic-capability milestones, corporate actions.
AI/ML markets are the fastest-growing category in the set. Manifold carries the densest public coverage: GPQA scores, FrontierMath passes, agentic-benchmark percentages, and speculative AGI-timeline markets. Polymarket has US-regulated counterparts on major lab milestones and corporate events like OpenAI IPO timing.
Canonical outcomes are typically binary with the benchmark and threshold in the event title. A specific subset uses date-range outcomes ("by Q2 2026", "by year-end 2026", "later") which the matcher handles with date-proximity signals.
Pitfalls: benchmark version drift (GPQA-diamond vs GPQA-full), model-family ambiguity ("GPT-5" before OpenAI formally names it), and markets that resolve on lab announcement vs on independent third-party verification. Storm surfaces the venue's stated resolution basis per event.
§ Matcher threshold
§ Active events
§ Recently resolved
No resolved or expired events in this category yet.