Bittensor（TAO）代幣經濟學

Issuance Mechanism

• Maximum Supply: 21 million TAO.
• Issuance Schedule: TAO tokens are mined on-chain following a halvening cycle, analogous to Bitcoin.
  • Initially, 7,200 TAO are emitted daily.
  • Once 50% of the total supply (10.5 million) is issued, the daily emission rate is halved.
  • Each subsequent halvening occurs whenever another 50% of the remaining unissued supply has been mined.
• Dynamic Emission Model: A proposed “Dynamic TAO” system (still conceptual as of early 2024) aims to dynamically adjust emissions to subnets based on price signals from subnet pools. This mechanism links emission rates to the value and demand for subnet resources, making emissions responsive to activity and utility within the protocol.

Allocation Mechanism

Subnet-Level Distribution

• Distribution to Subnets: Emissions are first distributed across the protocol’s subnet infrastructure, with allocation determined by validator votes.
• Within Each Subnet:
  • Miners: 41% of subnet emissions, rewarding computational work (e.g., AI models, data processing).
  • Validators: 41%, supporting governance and ensuring accurate task validation.
    • Of the validator portion, 82% is typically redistributed to delegators who stake with validators; validators retain 18%.
  • Subnet Owners: 18% for the entities that create and maintain subnet infrastructure.

Validator Pool Dynamics

• Subnet Token Pools: TAO is staked and traded within subnet pools using an AMM-inspired design (often compared to Uniswap V2).
• Validator Impact: Validators must purchase and stake TAO in these pools, directly affecting both token emission and market dynamics.
• Emissions Adjustments: Emissions to each subnet can be proportionally increased based on the subnet pool’s activity and perceived value; this further fine-tunes incentive alignment.

Usage and Incentive Mechanism

• Payments: TAO is used as payment for:
  • Creating new subnets.
  • Obtaining subnet user identification slots (UIDs).
  • Covering transaction fees within the Bittensor network.
• Staking/Delegating:
  • Staking: Prospective validators must stake (either self-stake or via delegated stake) sufficient TAO to participate and earn validator rewards.
  • Delegating: Holders can delegate TAO to validators, sharing in validator rewards minus the validator’s commission.
• Earning Rewards:
  • Miners: Earn rewards for computational work on subnets.
  • Validators: Rewarded for maintaining network consensus and governance.
  • Delegators: Receive a share of validator rewards relative to their delegation proportion.
• Task-Based Incentives: Specific subnets may apply custom reward schemes for distinct task types (e.g., AI model refinement, data labeling).

Lock-Up and Vesting Mechanisms

• Staked TAO: TAO staked by validators and delegators is typically subject to bonding and unbonding periods for protocol security and governance. The specifics of unbonding/lock-up periods may be set by subnet design, but are generally designed to prevent rapid in-out flows that could destabilize governance or consensus.
• Emission Vesting: No explicit time-locked vesting schedules are published for protocol rewards—emissions are distributed directly to miners, validators, and subnet owners on a continuous basis.
• Foundation/Investor Tokens: The Opentensor Foundation and early investors have acquired TAO through validator mining, but no public disclosure exists concerning vesting or lock-up arrangements for foundation or investor-held tokens.

Unlocking Timelines

• Network Emission: TAO emissions (mining) will gradually unlock the total supply over time, decreasing at each halvening. There is no fixed calendar-based end-date; unlocking follows a supply/milestone approach based on emission and halvening cycles.
• Foundation/Investor Unlock: There is no published vesting schedule for foundation/investor token allocations, but the Foundation’s tokens are obtained via validator mining rather than pre-mine, with no public lock-up schedule.

Summary Table

Analytical Insights

• Decentralized, Dynamic Allocation: Bittensor’s dynamic emission and allocation mechanisms serve to concentrate incentive where it is most valuable, determined algorithmically via subnet pools and validator governance.
• Proof-of-Intelligence Utility: Tokenomics are tightly coupled to network utility—mining, validating, and subnet ownership are all rewarded in TAO based on real-time performance and contribution, not static schedules.
• Open Participation, Network Security: Staking and delegation encourage wide participation, while lock-up/unbonding periods bolster security. Subnet-level design introduces flexibility, enabling distinct incentive models.
• Transparency Gaps: There are uncertainties around foundation and investor allocations/vesting due to lack of disclosed terms; this opacity warrants caution for those assessing long-term supply overhang risks.
• Future Evolution: The tokenomic system is in active development, with major proposals (like Dynamic TAO) still under consideration as of early 2025, indicating possible future changes to emission adjustments and incentive mechanics.

Conclusion

Bittensor employs a sophisticated, flexible token economy designed to incentivize the creation and validation of digital commodities, especially within AI and machine learning sectors. Its dynamic, on-chain emission mechanisms seek to efficiently direct rewards to the most valuable subnet activities, while staking and delegation ensure decentralized consensus and governance. Lack of full transparency around certain investor and foundation allocations is a limitation, but the open, modular approach of the protocol supports adaptability and ongoing optimization of its tokenomic incentives.