Original | Odaily Planet Daily (@OdailyChina)

Author | Azuma (@azuma_eth)

Do you remember the story of Nvidia CEO Jensen Huang praising Bittensor (TAO)?

On March 20, Huang was asked during his appearance on Chamath Palihapitiya's All-In podcast whether he was optimistic about decentralized AI systems/computing networks. Palihapitiya mentioned Bittensor as an example (somewhat suspiciously self-serving), stating that a team from a subnet on Bittensor successfully trained a Llama model with 4 billion parameters (actually 72 billion parameters), and the entire process was accomplished through distributed computing cooperation. Huang's response was that it was "a remarkable technological achievement."

As a result of this positive news, TAO surged against the trend last month, peaking above $370, and Bittensor was seen as "the hope of the entire village" in the cryptocurrency industry.

However, just half a month later, the situation took a sharp turn due to a sudden statement — as of the morning of April 10, TAO had fallen below $290, declining significantly for three consecutive days, and Bittensor found itself in what could be its largest public relations crisis since its establishment.

What Huang praised is actually a team called Covenant AI

Before explaining the background of the incident, we need to first understand the subnet architecture of Bittensor.

Bittensor is a decentralized machine learning network centered around token incentives. Bittensor utilizes a subnet mechanism that allows different teams to build various AI task markets, with miners and validators participating in computing and evaluation to distribute TAO rewards.

The "subnet team" mentioned earlier by Palihapitiya is actually called Covenant AI (formerly Templar), and the model that Huang praised is named Covenant-72B, which is a model with 72 billion parameters trained collaboratively by over 70 independent contributors on general-purpose hardware without requiring permission. It is also the largest decentralized large model pre-training project in history.

In summary, essentially, Bittensor can be understood as the underlying infrastructure supporting projects like Covenant AI, responsible for providing incentives, governance, and network rules, rather than directly developing specific AI models or applications; while subnet projects like Covenant AI are closer to "application layer builders" providing specific AI tasks and model capabilities on the underlying network.

Covenant AI's sudden statement

On the morning of April 10, Covenant AI founder Sam Dare suddenly releaseda statement (considering the continuous decline of TAO, the actual contradictions may have been brewing longer), stating that due to Bittensor and its Jacob Steeves (online name Const) violating the principles of decentralization, Covenant AI has decided to withdraw from the Bittensor network.

Covenant AI pointed out in the statement that the core belief of the team is that "the training of cutting-edge AI models should not be controlled by any single entity," but when a single actor can suspend subnet emissions, overturn the subnet owner's governance of their own community space, publicly abandon projects without a process, and use token sell-offs as a coercion mechanism to force others to comply, that is not decentralization, but centralized control disguised as decentralization.

Covenant AI further accused that every participant in the Bittensor ecosystem — miners, validators, and investors — should understand that this power indeed exists and has been exercised by Const. The exercise of this power by Const is not for the health of the network, but to regain control over a team that has become "too independent" and difficult to manage, a subnet owner capable of building its own community, making independent decisions, and operating without permission, as this threatens his power within the entire ecosystem. Specifically, although Bittensor employs a so-called "three-person management" structure for network upgrades — with three people managing multi-signatures and claiming to the community that this is distributed governance — the reality is otherwise. Const still exercises absolute power and resists any true power transfer — the power within the Bittensor ecosystem has never left one person's hands.

Covenant AI also mentioned that in the past few weeks, Const took a series of actions conflicting with the principles proclaimed by Bittensor, including suspending the emissions of the Covenant AI subnet, removing the team’s management rights over its own community channel, unilaterally abandoning subnet infrastructure, and applying economic pressure by conducting large-scale public token sell-offs during operational conflicts.

As a result, Covenant AI decided to withdraw from the Bittensor network. The team concluded by stating that decentralized, permissionless AI training is not a unique function of Bittensor but a technological capability that the Covenant AI team hopes to continue advancing. Covenant AI’s research, team, models, and vision will continue to move forward, with some very exciting projects already underway, which will be announced to the public soon.

Conflict exposed, Bittensor caught in a public relations whirlpool

Due to the success of Covenant-72B (SubNet-3), along with the fact that the Covenant AI team also operates two key subnets, Basilica (SubNet-39, positioned as an AI model evaluation/inference-related subnet) and Grail (Sub-81, positioned as a more complex task-driven AI subnet), the team has a significant position in the Bittensor ecosystem — perhaps it was precisely due to Covenant AI's increasing power in the community, resources, and discourse that triggered the "power struggle" conflict with Const.

As the conflict between the two parties became public, the Bittensor ecosystem quickly fell into a public relations whirlpool.

On the product level, with the departure of Covenant AI, the community began to raise questions about the future development and value of the Bittensor network. As one of the most technically narrative and achievement-rich teams in the current Bittensor ecosystem, the exit of Covenant AI means that this capability chain has been directly extracted, and Bittensor's technological progress in AI model training and ecological activity will face uncertainty, leading to a more cautious market judgment on its long-term value.

On the reputation front, Bittensor's decentralization narrative is facing its biggest challenge since its inception. Covenant AI's accusations target the very core of Bittensor's narrative — "decentralized AI network." For Bittensor, which relies on decentralization narratives to attract developers and computing power participants, the impact of this governance controversy goes far beyond short-term price fluctuations; it may more likely shake the confidence of ecosystem participants.

On the brand level, Covenant AI has, as a result of this incident, overshadowed Bittensor in the community mindset. Prior to this statement, the market generally perceived the "Huang's praise" incident as a commendation of Bittensor, with few realizing that Covenant AI was the true protagonist and even fewer knowing of the team's existence. As the event developed, Covenant AI's visibility continued to rise, and Bittensor became perceived as the "bleeding" party within the community.

As of the time of writing, there has been no official statement from Bittensor on social media, and Const vaguely responded on his personal account: "This incident will drive the first batch of truly 'headless operation' (presumably not relying on a single team), truly commercialized subnets… Thanks to Covenant AI for making Bittensor more decentralized."

Under Const's response, many Bittensor community users (especially TAO holders) are urging Const to provide a more detailed response to the accusations raised by Covenant AI, but Const has not replied further.

Odaily Planet Daily will continue to follow this matter, please stay tuned.

免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。