Bittensor co-founder Jacob Steeves has acknowledged that the protocol is not yet fully decentralized, saying the network still relies on core-team control in key areas while outlining a roadmap to complete decentralization within the next 18 months. Steeves, known in the crypto community as Const, said the current structure was not a design failure but a strategic decision made during the rapid development phase of artificial intelligence.
Bittensor has become one of the most prominent crypto-AI networks, using its TAO token to reward participants that contribute useful digital commodities, including machine intelligence, compute, storage and other services across specialized subnets. The project has attracted strong investor attention because it attempts to create an open market for AI resources outside the control of large technology companies.
However, its decentralization claims have faced growing scrutiny. Critics have argued that while Bittensor has open participation and distributed token ownership, important parts of the protocol still depend on a small group of engineers and core contributors. Steeves’ roadmap appears to directly address that criticism by acknowledging that Bittensor is not yet comparable to Bitcoin in terms of decentralization.
Centralization was a strategic trade-off
Steeves said Bittensor’s centralization reflects the need to move quickly in a fast-changing AI market. Unlike Bitcoin, which was designed primarily as a censorship-resistant monetary system, Bittensor is trying to build an adaptive intelligence marketplace. That has required frequent upgrades, rapid error correction and active protocol design.
The key issue is the economic incentive layer. Reports summarizing Steeves’ roadmap say Bittensor remains directionally guided by the core team, particularly around emissions, validator behavior and protocol-level incentives. That matters because Bittensor’s value proposition depends on whether the network can fairly reward useful intelligence production without excessive control from insiders or dominant validators.
The network has expanded significantly, with active subnet teams and validators competing to produce and evaluate different digital services. But decentralization is not only about the number of participants. It also depends on who controls upgrades, who determines incentives, how emissions are allocated and whether governance can function without informal founder authority.
Steeves’ admission may therefore be important for credibility. Rather than defending the protocol as already fully decentralized, he is framing decentralization as a process that must now become the project’s main priority.
Roadmap aims to reduce founder control
The 18-month roadmap includes several mechanisms intended to shift Bittensor away from core-team dependence. Planned changes include stronger validator competition, new liquidity pools that could help balance market dynamics, a conviction mechanism that allows token holders to signal long-term commitment, and steps to remove value extractors from the ecosystem.
The conviction mechanism is especially important because it could give committed TAO holders more formal influence while making short-term manipulation harder. Liquidity pools and shorting mechanisms could also help create more efficient markets around subnet assets and reduce the risk that attackers manipulate network growth or emissions.
If successful, the changes would move Bittensor closer to a model where validators, subnet operators and token holders collectively govern the system. That would help answer one of the biggest questions facing crypto-AI networks: whether they can scale without becoming dependent on the same centralized decision-making they claim to replace.
The challenge is execution. Decentralizing too quickly could slow development or expose the protocol to governance attacks. Moving too slowly could strengthen criticism that Bittensor is decentralized in branding but centralized in practice.
For investors, the roadmap adds both opportunity and risk. TAO’s long-term value depends heavily on whether Bittensor can become credible infrastructure for decentralized AI. Full decentralization would strengthen that thesis, but failure to deliver could undermine one of the protocol’s core narratives.
Steeves’ message is ultimately a reset of expectations. Bittensor is not yet fully decentralized, but its founder is now putting a timeline on when it should become so. The next 18 months may determine whether Bittensor can evolve from a founder-led crypto-AI network into a genuinely decentralized intelligence market.
