Last reviewed: January 2026
Numerai Meta Model
The Crowdsourced Hedge Fund Brain
ActiveInfrastructure Entity
52
/70 Total
Score Profile
Dimension Scores
Score Rationale
Persistence — Operating continuously since December 2015. 10+ years of proven operation. Institutional-grade persistence.
Autonomy — The meta model operates autonomously once signals are submitted. Human models feed it, but aggregation and execution are autonomous.
Cultural Impact — Influential in ML/quant communities but less mainstream recognition. Respected but niche.
Economic Reality — Manages real hedge fund capital ($100M+); NMR token with significant market cap; pays out to stakers.
Governance — Hybrid: Numerai Inc controls platform, but signal weights determined by stake (pseudo-decentralized).
Technical Distinctiveness — Pioneered erasure encoding for data privacy, stake-weighted ensembling, novel tournament structure.
Narrative Coherence — More infrastructure than character. No distinct personality, but clear purpose: crowdsourced alpha generation.
Curator Notes
Numerai's Meta Model is an unusual entity: a hedge fund whose trading signals come from an ensemble of thousands of data scientists' models, coordinated via the NMR token and staking mechanism. The 'agent' here is the aggregated intelligence of the crowd, filtered through stake-weighted consensus. Operating since 2015, it manages real capital and has pioneered novel approaches to crowdsourced intelligence. Lower narrative score reflects that it's more infrastructure than character—but its persistence and economic reality are unquestionable. A canonical example of decentralized intelligence that predates the current 'agent' discourse.
Evidence Archive
Score History
Score history will appear here after future reviews.
Current score: 52/70
Metadata
Inception2015-12-01
ClassificationEnsemble ML System / Prediction Market
Websitenumer.ai
Last Updated2026-01-06