One week after the dTAO upgrade, what should the Bittensor ecosystem improve?

The main goal of dTAO is to promote the development of subnets with real revenue potential, stimulate the birth of real-life use-case applications, and allow such applications to be correctly evaluated.

Author: Kevin, the Researcher at BlockBooster

TLDR：

• Bittensor uses dTAO to change the subnet reward allocation from a fixed proportion to a pledge weight decision, and 50% is 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, and it is necessary to balance the three major contradictions of miners ‘quality screening, user perception threshold and market popularity mismatch.

• Currently, only one of the TOP10 subnets requires miners to submit open source models. The rest of the subnets generally have defects such as anonymous teams and missing product anchors, exposing bottlenecks in Web3 AI infrastructure.

• The final verification depends on the establishment of positive feedback between the TAO price and the practical value of the subnet. Failure may trigger the continuous transformation of the Web3 AI track towards lightweight.

Background review

The introduction of dTAO reshapes Bittensor’s daily release rules:

• Previous rules:Subnet rewards are allocated in a fixed proportion-41% to verifiers, 41% to miners, and 18% to subnet owners. The amount of tao released for a subnet is determined by a verification vote.

• Rules after dTAO:Now, 50% of 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 amount of TAO released by a subnet is determined by the subnet pledge weight.

Design goals of dTAO:

The main goal of dTAO is to promote the development of subnets with real revenue potential, stimulate the birth of real-life use-case applications, and allow such applications to be correctly evaluated.

• Decentralized subnet evaluation:Instead of relying on a few validators, dynamic pricing in the dTAO pool will determine the allocation of TAO circulation. TAO holders can pledge TAO to support the subnets they believe in.

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

• Encourage early participation:It can encourage users to pay attention to new subnets and encourage the entire ecosystem to evaluate new subnets. Because validators who migrate to a new subnet earlier may receive higher rewards. Early migration to a new subnet means purchasing dTAO for that subnet at a lower price, increasing the likelihood of obtaining more TAOs in the future.

• Promote miners and validators to focus on high-quality subnets:Further stimulating miners, validators find high-quality new subnets. The miner’s model is placed offline, and the verifier’s verification is also offline. The Bittensor network rewards the miners only based on the verifier’s evaluation. Therefore, Bittensor can correctly evaluate different or all types of ai applications as long as they conform to the miner-verifier architecture. Bittensor is extremely inclusive of AI applications, allowing participants at each stage to get incentives and feed back the value of Bittensor.

Analysis of three scenarios that affect dTAO price trends

Review of basic mechanisms

The fixed daily release of TAO and the same amount of dTAO are injected into the mobility cell to constitute a new mobility cell parameter (k value). 50% of the dTAO goes into the liquidity pool, and the remaining 50% is assigned to subnet owners, validators, and miners based on weight. The higher the weight of the subnet, the greater the proportion of TAO allocation.

Scenario 1: Positive cycle of pledge volume growth

As the number of TAOs entrusted to verifiers continues to increase, the subnet weight increases accordingly, and the allocation proportion of miners ‘rewards expands simultaneously. The motivations for verifiers to purchase large quantities of subnet tokens can be divided into two categories:

1. short-term arbitrage behavior

   The subnet owner acts as a verifier to push up the currency price by pledging TAO (continuing the old release model). But the dTAO mechanism weakens the certainty of this strategy:

• When the proportion of irrational pledge users is higher than that of quality-focused users, short-term arbitrage is sustainable

• On the contrary, it will lead to the rapid depreciation of early hoarding tokens, and the superposition of a uniform release mechanism will restrict the acquisition of chips, which may be eliminated by high-quality subnets in the long run.

2. Value capture logic

   Subnetworks with practical application scenarios attract users through real benefits, and pledgers not only receive leveraged dTAO benefits, but also receive additional pledge returns, forming a closed loop of sustainable growth.

Scenario 2: Dilemma of relative stagnation

When the subnet pledge volume maintains growth but lags behind the top project, although the market value increases steadily, it is difficult to maximize profits. At this time, we should focus on:

• Miner quality determines the upper limit:As an open source model incentive platform (non-training platform), TAO’s value stems from the output and application of high-quality models. The strategic direction choice of the subnet owner and the quality of the models submitted by miners constitute the development ceiling

• Team Capability Mapping: Most of the top miners come from the subnet development team, and the quality of miners essentially reflects the technical strength of the team

Scenario 3: Death spiral of pledge loss

When the amount of pledge in the subnet decreases, it is easy to trigger a vicious cycle (decrease in pledge → decrease in revenue → further loss). Specific incentives include:

1. Competitive elimination

   Although the subnet has practical value, its product quality is backward, and its weight is reduced, resulting in its elimination. This is an ideal state for healthy ecological development, but there is no indication that TAO ‘s value as a Web3 application incubation shovel has become explicit.

2. Expected collapse effect

   The market’s negative outlook for subnets led to the withdrawal of speculative pledges. When daily releases began to decline, the loss of non-core miners accelerated, eventually forming an irreversible decline trend

Potential risks and investment strategies

Volatility risk during early release periods

• High volatility window: The initial total release of dTAO is large but the average daily release is constant, resulting in possible sharp fluctuations in prices in the previous few weeks. At this time, root network pledge becomes a risk mitigation strategy and can stably obtain basic benefits.

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

• Weight game mechanism:The verifier weight is determined by the subnet dTAO value + the root network TAO pledge (compound weight model). 100 days before the subnet goes online, the root network pledge still has the advantage of revenue certainty

• Characteristics of Meme-like transactions:At present, subnet pledge behavior and Memecoin speculation have similar risk attributes

Value investment and market mismatch

• Paradox of ecological construction:The dTAO mechanism is designed to cultivate practical subnets, but the value investment characteristics lead to:

• High market education costs: The quality of miners/application scenarios/team background/profit model needs to be continuously evaluated, which constitutes a cognitive threshold for non-AI professional investors

• The transformation of popularity lags: In sharp contrast to Agent tokens, subnet tokens have not yet formed a market consensus of the same size

Systemic risks of irrational pledge

• Historical dilemma repeats itself:If users continue to blindly follow the release volume indicator, it will lead to:

• Verifier power rent-seeking: The drawbacks of repeating the old mechanism for offline self-voting

• Mechanism upgrade failure: quality screening function that violates the original intention of dTAO

• Cognitive threshold requirements:Investors need to have subnet quality assessment capabilities. There is a gap between current market maturity and mechanism requirements.

The Game Theory Dilemma of Investment Timing

• Best intervention window:The investment window should be moved back a few months after the subnet is launched (the team capabilities/network potential visibility stage), but faced with:

1. Risk of declining market attention

2. Liquidity shrinkage caused by early speculators leaving

• Double verification of success sign:

1. TAO price forms positive feedback with the practical value of the subnet

2. Verifier chooses TAO to hold positions rather than sell in order to obtain sustained income

Risk of losing control of miners ‘quality

• Adverse selection challenges:

• Missing quality screening mechanism: Current models are difficult to effectively distinguish the quality of miners ‘contributions

• Imbalanced incentive environment: arbitrage behavior of low-quality miners squeezes the living space of high-quality developers

• Bottlenecks of ecological construction:The incubation environment for open source models is not yet mature and may fall into the dilemma of bad money driving out good money.

There are three contradictions in investing in the dTAO subnet:

Core contradiction:

• Can subnets attract high-quality miners ‘resources

• Whether the user evaluation system is effective

Secondary contradiction:

• Whether there are real commercial application scenarios in the subnet

Potential risk points:

• Transparency of development team information disclosure

• Rationality of profit model design

• Marketing execution capabilities

• Possibility of external capital intervention

• Design of token issuance mechanism

Observations and expectations

1. Although the open source model is the mainstream direction of technological evolution, it may be difficult to break through the development bottleneck in the field of decentralization.

   Currently, Bittensor, as an industry leader, still has significant quality flaws in its dTAO subnet ecosystem. Analyzing the top ten subnets in the figure above, we can see that only one of the TOP10 subnets requires miners to submit open source models, and the miners in the remaining subnets have weak correlation with model development.

   

2. There are extremely high technical barriers to open source model training, which poses a major challenge for Web3 developers. In order to maintain the miner base, most subnets actively reduce technical entry barriers and avoid model open source requirements to ensure the supply of token incentive pools.

3. Even subnets with non-mandatory open source models are equally worrying about their ecological quality. The following problems are common in the TOP10 subnet:

• Lack of verifiable landing products

• Excessive proportion of anonymous development teams

• Lack of effective anchoring between dTAO tokens and product value

• Income model lacks market persuasion

4. The underlying design concept of dTAO is forward-looking, but the current Web3 infrastructure is not enough to support its ideal ecological construction. This mismatch between ideals and reality may have two consequences:

• The dTAO subnet valuation system needs to be revised downwards

• If Bittensor open source model platform verification fails, the Web3 AI track may shift to lightweight directions such as Agent applications and middleware development