Give subnets runway, not endless subsidy Emissions should help subnets prove themselves, then increasingly reward the value they create I propose two complementary mechanisms for how Bittensor funds subnets: Emission Taper Subnets start with access to their full available emissions As they mature, that share gradually tapers Revenue-funded alpha burns or direct distributions can slow or reverse the taper when they return verifiable economic value to alpha holders If that value doesn't develop, the subnet gradually receives a smaller share of its otherwise available emissions Tapered-out TAO emissions are burned 🔥 That means less dilution and less unproductive sell pressure This matters even more as $TAO appreciates At $1,000 TAO, a smaller TAO allocation can fund the same dollar costs, so mature subnets shouldn't automatically consume the same amount of TAO forever This adds another control beyond market allocation: emissions can adapt over time to subnet maturity and verifiable economic performance With Emission Taper setting the protocol-side ceiling and Fractal Emissions letting owners opt lower, market support, verifiable value return, maturity, subnet-specific needs and TAO purchasing power can all shape how much subsidy is actually consumed, without relying on external oracles or subjective performance reviews The goal: Early subnet: Bittensor funds growth and experimentation Growing subnet: real demand starts carrying more of the load Mature subnet: real economic value exceeds network subsidy Fractal Emissions Let subnet owners choose how quickly they consume their available emissions: 100% 50% 25% Taking less means slower dilution and more runway through a proportional deregistration-score adjustment A subnet that needs extra time can accept fewer emissions now in exchange for a longer runway to prove itself Declined TAO emissions are burned 🔥 It also gives subnet owners direct control when TAO appreciates faster than their actual funding needs Fractal Emissions lets the owner reduce the rate rather than consume an unnecessarily large subsidy The chosen rate should be visible to the market, signalling how much network support the subnet believes it can productively use A subnet taking 100% while failing to turn that subsidy into real value would face pressure from both sides: the Emission Taper would keep reducing its available support over time, while the market could see that it is taking more from the network than it can justify and factor that into its own allocation decisions The rate itself becomes a signal: take more, and the expectation to produce value rises with it Both mechanisms could materially reduce the amount of newly emitted TAO reaching circulation by burning tapered-out and voluntarily declined emissions Together, they solve different parts of the same problem Emission Taper determines how much subsidy remains available as a subnet matures Fractal Emissions lets the subnet choose how quickly it consumes that available subsidy More runway to prove value, less subsidy when that value doesn't materialise These ideas may not be perfect in their current form, but they're worth discussing, and Bittensor's engineers could likely take the underlying mechanisms much further Full write-ups below ↓

Established subnets need capital to scale. Research-stage subnets need time to mature. Fractal emissions could let subnets choose how quickly they consume emissions.