We're building a Bittensor subnet incubation academy - modeled after La Masia, where promising subnet ideas are developed in-house on mainnet and graduated into dedicated subnets. Where builders become subnet owners.
Most great subnet ideas never make it to mainnet because the capital barrier is too high. Academia changes that.
We scout, evaluate, and develop subnet concepts inside a single subnet using Bittensor's multiple incentive mechanism architecture. Projects earn emissions based on real performance - not promises. The best ones graduate into their own dedicated subnets.
- Mechanism 0 - Research & Design: subnet theses, incentive designs, benchmark suites, exploit analysis.
- Mechanism 1 - Prototype & Arena: working implementations, benchmark competition, mainnet-tested prototypes.
An AI agent evaluation system handles intake, scoring, and candidate ranking at scale - think ATS for subnet ideas. Transparent rubrics. Auditable scores. Meritocratic selection.
Three paths to graduation: crowdloan (community-funded), collaboration, or self-funding from emissions earned in the incubator.
Ideas are good. Execution is everything.
Over the next updates we'll publish:
- The Academia vision and La Masia development model
- Multi-mechanism architecture and incentive design
- AI agent evaluation system overview
- Submission and intake process for builders
- Graduation criteria and subnet transition paths
- Miner / validator onboarding docs
- Public roadmap with milestones