dTAO mechanism out of control? Bittensor is deviating from the AI track
Market incentives may reduce TAO releases to mere commercial subsidies, causing the project to lose its original significance.Original Title: Why dTAO is broken
Original Author: @tzedonn, crypto writer
Original Compiler: zhouzhou, BlockBeats
Editor's Note: The dTAO mechanism of Bittensor was originally designed to distribute the release of TAO more fairly, but issues emerged just a month after its launch. The SN28 subnet exploited a mechanism flaw, pushing the release of TAO into meme coin speculation, ultimately leading to centralized intervention by the foundation. As decentralization progresses, similar incidents may be unavoidable, and Bittensor could transform into a generalized incentive network rather than an AI project, with the core issue being a lack of unified goals.
Below is the original content (reorganized for better readability):
I am someone who is irresistibly drawn to novel token economics, watching as crypto protocols continuously adjust their incentive mechanisms, sometimes appearing quite clever—until they inevitably run into problems. This process itself is fascinating. So when Bittensor launched the dynamic $TAO (dTAO) system on Valentine's Day (was this a Valentine's gift from @const_reborn?), I was immediately intrigued.
The idea is simple: to provide a new, more "fair" way to distribute the release of TAO, allowing various subnets to obtain TAO more reasonably.
However, less than a month later, problems began to surface. It turns out that a design that seems reasonable does not necessarily survive in a free market.
How dTAO Works
A brief recap of how dTAO operates:
Each subnet has its own subnet token ($SN), existing as a TAO-SN based UniV2 type native liquidity pool (LP). However, "staking" TAO for SN is essentially the same as "exchanging" TAO for SN. The only difference is that you cannot add additional liquidity to the liquidity pool, nor can you trade directly between different SN tokens (e.g., SN1 → SN2), but you can exchange indirectly through TAO (SN1 → TAO → SN2).
The release of TAO is allocated based on the price of each subnet's SN token. They use a moving average price to smooth out price fluctuations and prevent price manipulation.
The release amount of SN tokens is also quite high, with a supply cap of 21 million, similar to TAO and BTC. A portion of this enters the TAO-SN trading pool, while the remainder is allocated to subnet stakeholders (miners, validators, subnet owners).
The number of SN tokens entering the TAO-SN trading pool depends on how many SN are needed to "balance" the TAO entering the pool, ensuring that the price of SN remains stable under TAO valuation while increasing liquidity.
However, if the calculated number of SN tokens obtained by the subnet exceeds its maximum release amount (determined by the SN release curve), then the SN release amount will be capped, leading to an increase in the price of SN under TAO valuation.
Core Assumptions of the dTAO Mechanism
Point (2) in the dTAO mechanism is based on a core assumption: that subnets with higher market capitalization create more value for the Bittensor network and therefore should receive more TAO release.
But the reality is that in the crypto market, the tokens with the highest prices are often those that are most hyped, most marketed, and even exhibit characteristics of Ponzi schemes. This is why the valuations of L1 public chains and meme coins are always relatively high.
The original intention of the mechanism design was good, as it assumed that subnets that truly create value would repurchase SN tokens through generated revenue, driving up the price of SN, thereby obtaining more TAO release. However, I find this logic somewhat naive.
Meme Coin Subnets & Broken Token Economics
Before the launch of dTAO, I discussed the obvious flaws in the dTAO token economic model with several crypto analysts—high market capitalization does not equal high revenue, nor does it mean that more value is truly created.
But I did not expect this theory to be validated by the market so quickly. The workings of the free market are always surprising.
Just before the dTAO upgrade, an anonymous user took over subnet 281 and directly transformed it into a meme coin subnet, naming it "TAO Accumulation Corporation," abbreviated as "LOL-subnet." This was clearly unrelated to AI.
On the (now deleted) GitHub page, it stated… no mining required, just hold the tokens—completely turned into a Ponzi scheme.
In the LOL-subnet (subnet 281), miners do not need to run any code, and the validator scoring mechanism is entirely based on the number of subnet tokens held by miners. The more SN28 tokens held, the more TAO release they receive.
What actually happened was: speculators bought SN28 tokens → SN28 price rose, SN28 price rose → received more TAO release, if TAO release exceeded the subnet's token release cap → SN28 price continued to rise, released SN28 tokens distributed to "miners" according to their holdings → the more SN28 held, the more received, in order to obtain more TAO, more people bought SN28 → higher prices → the Ponzi cycle continued.
Ultimately, the TAO release officially flowed into… meme coins! At one point, SN28 even became the 7th ranked subnet by market capitalization in Bittensor.
Why didn’t SN28 completely take over Bittensor? Centralized intervention saved Bittensor.
In the rapid expansion of SN28, the Opentensor Foundation directly used their root stake to run custom validator code, encouraging everyone to sell SN28. Ultimately, SN28 plummeted 98% within hours and was completely liquidated.
After the Opentensor Foundation took action, SN28 plummeted 98%.
Essentially, the foundation acted as a centralized entity, preventing the free market from operating according to the dTAO mechanism. But they were able to do this because we are still in a transitional phase—the release mechanism of TAO is gradually shifting from the old model to the dTAO mechanism.
Old TAO Release Mechanism & Transition to dTAO
Under the old mechanism, the 64 validators holding the most TAO staked in SN0 (the "root subnet") could vote on the allocation of TAO release.
However, this mechanism also had significant issues driven by interests, especially as large validators (such as Opentensor Foundation, DCG Yuma, Dao5, Polychain, etc.) held too much power.
For example, potential conflicts of interest include:
They could prioritize allocating TAO to subnets they invest in or incubate.
They could guide TAO release to subnets they operate validators for, which can earn TAO rewards.
These issues have long existed, and dTAO was intended to address this centralization problem, but the SN28 incident proved that the new mechanism still has significant flaws.
Moving in the right direction towards decentralization, but risks remain.
Breaking away from the old mechanism is indeed a step towards decentralization. Although this may mean the team will lose some control over TAO release, I still recognize that they have chosen a more decentralized reward mechanism.
However, when the SN28 incident occurred, the dTAO mechanism had only been online for a week, and SN0 (the root subnet) still controlled about 95% of the TAO release (as shown by the blue line in the chart below), which allowed the Opentensor Foundation to intervene quickly and prevent the Ponzi scheme from continuing to expand.
But the problem is:
About a year later, the power of SN0 will gradually decrease to around 20%, at which point it will no longer be able to directly control most of the TAO release.
If a situation similar to SN28 arises in the future, it is likely that no one will be able to intervene through SN0.
In this case, Bittensor may no longer be a "decentralized AI" project, but could completely transform into a meme coin incentive network.
Bittensor is still in the transitional phase of the release mechanism, with control gradually shifting from the old mechanism (SN0 or "root prop") to the new mechanism (dTAO or "alpha prop").
Not just meme coins, Bittensor could become a generalized incentive network.
Even if we assume that in a bear market environment, people will not rush into meme coins, Bittensor still has a high probability of becoming a completely AI-unrelated "generalized incentive network."
For example: if someone launches a decentralized subnet for mining Bitcoin (this concept is not new), it could incentivize more efficient BTC mining, then use the mined BTC as continuous income to repurchase SN tokens, thereby obtaining TAO release.
If this model holds, TAO will shift from a decentralized AI project to a generalized incentive project, with TAO release no longer used to promote AI development, but instead becoming a means to subsidize various operational costs (OpEx).
Technically, there is nothing wrong with this, as the Yuma consensus mechanism was originally designed to reach consensus on "subjective" work and is not necessarily limited to AI. But without a clear goal, the entire network may become… meaningless.
Conclusion: Cracks in the dTAO Mechanism Have Emerged
The dTAO mechanism has only been online for a month, and problems have already surfaced.
According to the incentive logic of the free market, without centralized intervention, Bittensor may no longer be an AI project, but rather an "attention network" dominated by meme coin subnets, or it could become a generalized incentive network, with various enterprises utilizing TAO releases to subsidize operational costs without promoting the development of the Bittensor ecosystem.
I believe Bittensor needs a true "objective function" that allows all subnets to move in the same direction. But the problem is that it is difficult to define an absolute goal in the AI field (AGI?), as we have already seen that even LLM evaluation frameworks struggle to achieve complete fairness… this is also why the Yuma consensus mechanism was initially designed to reach consensus on "subjective" work.
"Tell me the incentive mechanism, and I can tell you the outcome." Peace!
