This episode unpacks the fierce economic and philosophical debates shaping Bittensor's future, revealing how mechanisms like minor burns and root sell pressure are intentionally designed to forge a resilient, incentive-driven market for decentralized AI.
The Case for Minor Burns and Economic Freedom
- Const, a core architect of Bittensor, opens by addressing the contentious topic of "minor burns," where subnet owners intentionally burn a portion of the TAO emissions designated for miners. He argues this is not a flaw but a feature of a system designed for economic experimentation and resilience.
- Historical Context: The previous "root network" system for allocating emissions failed because it was not incentive-compatible, leading to insider deals and collusion. This necessitated the creation of Dynamic TAO (DTAO).
- The Unstoppable Force of Incentives: Const asserts that any attempt by the protocol to technically prevent minor burns would be a "fool's errand." Subnet owners would simply find other, less transparent ways to control emission flow, leading to a game of "whack-a-mole" that punishes honest actors and rewards clever cheaters.
- Economic Experimentation: He frames minor burns as a tool for subnet owners to protect their token equity and manage their economies, especially in early, low-liquidity phases. "Don't we want to have subnets figuring out the token economics that drive the highest prices for their tokens and ultimately the most demand to be participating in DTAO? I think that is of higher importance."
- Strategic Implication: Investors and researchers should view minor burns not as a bug, but as a signal of a subnet owner's economic strategy. The key is to analyze whether this strategy is sustainable and attractive to miners in the long run, as a subnet where everyone burns ultimately fails.
Understanding the Dynamic TAO (DTAO) Mechanism
- The conversation details the core mechanics of DTAO, explaining the rationale behind its complex but deliberate design. Const emphasizes that the system is built to force market intelligence and long-term thinking.
- Root Sell Pressure: This mechanism, where a portion of a subnet's validator rewards are automatically sold for TAO, was designed to test the market. It prevents new subnets from artificially inflating their token price in a low-liquidity environment, thereby protecting investors from buying into unsustainable hype. Const notes that subnets less than a year old are still in a "long drawn out ICO process," and their markets will only mature as root proportion declines.
- The "Hotfix" Was a Restoration: A recent update was not a change but a restoration of a key DTAO feature. This feature automatically reduces the injection of new alpha tokens when a subnet's price falls below its emission value.
- zkML (Zero-Knowledge Machine Learning): While not discussed in this episode, this technology enables private verification of AI models, a concept relevant to building trust in decentralized AI systems.
- Actionable Insight: This restored mechanism acts as a downside buffer, slowing price drops and mechanistically pushing the total value of the DTAO market back towards equilibrium. For investors, this provides a degree of built-in price support, making subnet tokens more investable by reducing tail risk.
Market Participation and Investor Psychology
- The discussion shifts to the current state of the DTAO market, addressing concerns about low volume and falling prices.
- A Market in a "Lull": Dave from Sturdy asks about low participation rates. Const attributes the current low volume to investor fear and the natural learning curve of a novel, complex system. He points out that Bittensor's design injects significant liquidity, allowing the market to function even with low external participation.
- Value Investing vs. Short-Term Fear: Mark, who is building a Bittensor-focused hedge fund, expresses a preference for the current "depressed" prices over a speculative bubble. He sees the system as built for long-term value investors, not day traders.
- The Power of Yield: Const reminds listeners that despite price volatility, the high APRs for stakers are designed to reward long-term conviction. "If you're betting on a horse like that, you know, you you you can sit you can sit easy. You can sleep easy because that person is is that horse, right, is charging for you."
The True Solution to Miner-Owner Conflicts
- The conversation returns to minor burns, with a focus on the correct forum for resolving the conflict. Both Const and Distributed Tensor emphasize that the solution is social and political, not technical.
- The Failure to Listen: Distributed Tensor, a subnet owner, offers a candid perspective on the disconnect between miners and owners. He notes that miners often see burns as pure greed, while owners see them as a necessary survival strategy. He advocates for greater empathy and mutual understanding.
- Governance is the Answer: Const states unequivocally that the only way to stop burns is through stake-weighted governance. Miners who oppose burns must accumulate stake in a subnet and run validators that do not use the burn key.
- Proof-of-Stake Reality: In a Proof-of-Stake system, influence is measured by stake. Social consensus on platforms like Discord is immeasurable and cannot be used to enforce protocol rules. The power lies with token holders who actively participate in governance.
The Subnet Slot Dilemma: Scarcity vs. Expansion
- The group discusses the controversial decision to cap the number of subnet slots at 128 and the potential for deregistrations.
- Unintended Consequences: Limiting slots was meant to reduce chaos but created a secondary market for subnet "cold keys," leading to speculative trading around anticipated subnet takeovers.
- The Sanctity of Property Rights: Const is strongly against forcibly deregistering subnets, comparing a subnet slot to a token balance. He argues that violating these implicit property rights would damage trust in the network.
- Path Forward: The team is considering either increasing the number of slots to allow new teams to enter or developing a mechanism to cycle slots without liquidating the associated tokens. This remains an open and critical discussion for the ecosystem's growth.
The Aphine Vision: A Case Study in Integration
- Const concludes by detailing his new project, Aphine, a subnet designed to integrate the entire Bittensor technology stack.
- The Pinnacle Layer: Aphine aims to be the final layer that pulls together compute (from subnets like Targon), pre-training, and inference (from subnets like Shoots) to produce "pinnacle intelligence."
- Mechanism: It will use Reinforcement Learning (RL) to fine-tune models for specific, high-value tasks like advanced coding and mathematics. It is built directly on top of Shoots (SN1) for its inference layer, showcasing true interoperability.
- The End Goal: To create a full-stack, decentralized AI that can compete with top centralized labs and be used by anyone globally, often without them even realizing they are interacting with a decentralized network. This represents the ultimate ambition of the Bittensor project.
This discussion reveals DTAO as a sophisticated market designed to be self-regulating through raw incentives, not rigid rules. For investors and researchers, the key takeaway is that long-term success in Bittensor hinges on understanding these deep economic forces and participating in the stake-based governance that truly shapes the network's evolution.
