One week after the dTAO upgrade, in what aspects should the Bittensor ecosystem be improved?

Author: Kevin, the Researcher at BlockBooster

TL;DR

• Bittensor has changed the allocation of subnet rewards from a fixed ratio to being determined by the staking weight through dTAO, with 50% injected into the liquidity pool, aiming to promote the development of high-quality subnets through decentralized evaluation.

• Early high volatility, APY traps, and adverse selection coexist, requiring a balance between miner quality screening, user cognitive thresholds, and market heat mismatch.

• Among the current TOP10 subnets, only 1 requires miners to submit open-source models, and the rest of the subnets generally have defects such as anonymous teams and lack of product anchoring, exposing the bottleneck of the Web3 AI infrastructure.

• The final verification depends on the positive feedback loop established between the TAO price and the practical value of the subnet, and failure may lead to the continuous transformation of the Web3 AI track towards lightweight.

Background Review

The introduction of dTAO reshapes the rules for daily release of Bittensor:

Previous rules: Subnet rewards were distributed in a fixed ratio - 41% to validators, 41% to miners, and 18% to subnet owners. The TAO release volume of the subnet was determined by validator voting.

After dTAO: Now, 50% of the newly issued dTAO tokens will be added to the liquidity pool, and the remaining 50% will be distributed among the validators, miners, and subnet owners based on the decisions of the subnet participants. The TAO release volume of the subnet is determined by the subnet staking weight.

The design goals of dTAO:

The main goal of dTAO is to promote the development of subnets with real income potential, stimulate the birth of real-use case applications, and ensure that these applications are correctly evaluated.

Decentralized subnet evaluation: No longer relying on a few validators, the dynamic pricing of the dTAO pool will determine the allocation of TAO issuance. TAO holders can support the subnets they believe in by staking TAO.

Increase subnet capacity: Removing the subnet cap promotes competition and innovation in the ecosystem.

Encourage early participation: Able to incentivize users to pay attention to new subnets and incentivize the entire ecosystem to evaluate new subnets. Because validators who migrate to new subnets earlier may receive higher rewards. Early migration to a new subnet means buying the subnet's dTAO at a lower price, increasing the possibility of obtaining more TAO in the future.

Encourage miners and validators to focus on high-quality subnets: Further stimulate miners and validators to find high-quality new subnets. Miners' models are placed off-chain, and validators' verification is also off-chain. The Bittensor network only rewards miners based on the evaluation of the validators. Therefore, for different types or all types of AI applications that fit the miner-validator architecture, Bittensor can correctly evaluate them. Bittensor has extremely high inclusiveness for AI applications, allowing participants at every stage to receive incentives and thus feed back into the value of Bittensor.

Three Scenario Analysis Affecting the Price Trend of dTAO

Basic Mechanism Review

The daily fixed release of TAO and an equal amount of dTAO injected into the liquidity pool constitute the new liquidity pool parameters (k value). 50% of the dTAO enters the liquidity pool, and the remaining 50% is allocated to subnet owners, validators, and miners according to the weight. The higher the weight of a subnet, the greater the proportion of TAO allocation it receives.

Scenario One: Positive Feedback Loop of Increased Staking

When the TAO delegated to validators continues to increase, the subnet weight will rise accordingly, and the miner reward allocation proportion will also expand. There are two types of motivations for validators to massively purchase subnet tokens:

1. Short-term arbitrage behavior

Subnet owners as validators can raise the coin price by staking TAO (continuing the old release model). But the dTAO mechanism weakens the certainty of this strategy:

• When the proportion of irrational staking users is higher than quality-focused users, short-term arbitrage can be sustainable

• Conversely, it may lead to the rapid devaluation of the hoarded tokens, coupled with the limited acquisition of chips due to the uniform release mechanism, and they may be eliminated by high-quality subnets in the long run

2. Value capture logic

Subnets with real application scenarios attract users through real yield, and stakers not only obtain leveraged dTAO returns but also additional staking rewards, forming a sustainable growth loop.

Scenario Two: The Dilemma of Relative Growth Stagnation

When the subnet staking volume maintains growth but lags behind the leading projects, the market capitalization steadily increases but it is difficult to maximize returns. At this time, the focus should be on:

• Miner quality determines the ceiling: TAO as an open-source model incentive platform (not a training platform), its value comes from the output and application of high-quality models. The strategic direction chosen by the subnet owners and the quality of the models submitted by the miners jointly constitute the development ceiling

• Team capability mapping: Top miners often come from the subnet development team, and the quality of miners essentially reflects the technical strength of the team

Scenario Three: The Death Spiral of Staking Loss

When the subnet staking volume declines, it is easy to trigger a vicious cycle (reduced staking → reduced returns → further loss). The specific causes include:

1. Competitive elimination: Although the subnet has practical value, the product quality lags behind, and the weight decline leads to elimination. This is the ideal state of ecological health development, but so far there is no sign of the value of TAO as the "Web3 application incubation shovel" becoming apparent

2. Expectation collapse effect: Market pessimism about the subnet's prospects leads to the withdrawal of speculative staking. When the daily release volume begins to decline, non-core miners accelerate the loss, ultimately forming an irreversible downward trend

Potential Risks and Investment Strategies

• High volatility window period: In the early stage of dTAO, the total release volume is large but the daily release is constant, which may lead to violent fluctuations in the first few weeks. At this time, staking on the root network becomes a risk mitigation strategy, which can stably obtain basic returns

• APY trap: The short-term temptation of high APY may mask the long-term risks of insufficient liquidity and lack of subnet competitiveness

• Weight game mechanism: The validator weight is jointly determined by the subnet dTAO value and the root network TAO staking (compound weight model). In the 100 days before the subnet goes online, the root network staking still has the advantage of return certainty

• Meme-like trading characteristics: At the current stage, the subnet staking behavior has similar risk attributes to Memecoin speculation

Value Investment and Market Mismatch

• Paradox of ecosystem building: The dTAO mechanism aims to cultivate practical subnets, but the value investment characteristics lead to:

• High market education cost: It is necessary to continuously evaluate miner quality/application scenarios/team background/profit model, which poses a cognitive threshold for non-AI professional investors

• Slow conversion of heat: In contrast to Agent tokens, subnet tokens have not yet formed a similar scale of market consensus

Systemic Risks of Irrational Staking

• Repetition of historical dilemmas: If users continue to blindly follow the release volume indicator, it will lead to:

• Validator rent-seeking: Repeat the subnet self-voting flaws under the old mechanism

• Mechanism upgrade failure: Violating the quality screening function of the dTAO design

• Cognitive threshold requirement: Investors need to have the ability to evaluate subnet quality, and the current market maturity and mechanism requirements have a gap

The Dilemma of Investment Timing Game

• Best entry window: The investment window should be postponed to a few months after the subnet goes online (the stage where team capabilities/network potential are visible), but facing:

1. Market attention attenuation risk 2. Liquidity contraction caused by the withdrawal of early speculators Dual verification as a sign of success: 1. Positive feedback between TAO price and subnet utility value 2. Validators choose to hold TAO positions rather than sell them to obtain continuous returns Miner quality control risk - Adverse selection dilemma: - Lack of quality screening mechanism: the current model is difficult to effectively distinguish the contribution quality of miners - Imbalance in the incentive environment: the arbitrage behavior of low-quality miners squeezes the survival space of high-quality developers - Bottleneck in ecological construction: the open-source model incubation environment is not yet mature, which may fall into the "bad money drives out good money" dilemma Three contradictions in investing in dTAO subnets: Core contradictions: - Whether the subnet can attract high-quality miner resources - Whether the user evaluation system is effective Secondary contradictions: - Whether the subnet has real business application scenarios Potential risk points: - Transparency of development team information - Rationality of profit model design - Execution capability of marketing - Possibility of external capital involvement - Design of token issuance mechanism Observations and expectations Although the open-source model is the mainstream direction of technical evolution, it may be difficult to break through the development bottleneck in the decentralized field. At present, Bittensor, as the industry leader, its dTAO subnet ecosystem still has significant quality defects. By analyzing the ranking of the top ten dTAO reward release subnets, it can be known that only 1 of the TOP10 subnets requires miners to submit open-source models, and the correlation between the miner group and model development in the remaining subnets is relatively weak. The high technical threshold of open-source model training poses a major challenge for Web3 developers. Most subnets actively lower the technical access threshold to maintain the miner base, and avoid the requirement of open-source models to ensure the supply of token incentive pools. Even for subnets that do not require open-source models, their ecological quality is also worrying. Common problems in the TOP10 subnets: - Lack of verifiable commercial products - High proportion of anonymous development teams - Lack of effective anchoring between dTAO tokens and product value - Lack of market persuasiveness in the revenue model The underlying design concept of dTAO is forward-looking, but the current Web3 infrastructure is not sufficient to support its ideal ecological construction. This mismatch between ideals and reality may lead to two consequences: - The valuation system of dTAO subnets needs to be adjusted downward - If the Bittensor open-source model platform fails to be verified, the Web3 AI track may turn to lightweight directions such as Agent applications and middleware development