One year ago, Bittensor launched Dynamic TAO (dTAO), an upgrade that felt like a new paradigm was coming to the network. As Const put it, the rollout was “insane,” but it worked. In a single move, Bittensor went from being a one-token blockchain into a system with 128 (and potentially more) subnet tokens, instantly creating open markets around emissions, incentives, and value discovery.
That transition didn’t just change the network’s structure; it changed its behavior. Bittensor moved from a low-liquidity environment into a high-liquidity market regime, with millions of dollars in daily volume and tens of thousands of TAO traded every day. But that growth came with a cost: low MEV resistance, increased volatility, and the inevitable chaos of a permissionless system.
The Ugly Part: Scams, Rugpulls, and Ponzies Were Guaranteed
The community didn’t sugarcoat it: almost every worst-case scenario predicted before dTAO actually happened. Ponzies, rugpulls, exploitative tokenomics, and financial schemes appeared across the subnet ecosystem. Const specifically referenced early examples Tenex, which put the whole ecosystem in chaos.
But instead of pretending this was unexpected, Const argued it was simply the price of permissionless markets. If you want a system driven by the wisdom of the crowd, you also have to accept that bad actors will show up and try to take advantage of it.
The difference is what happened next.
The Market Adapted: Subnets Learned to Fix Their Tokenomics
The ecosystem didn’t collapse under the chaos; it learned. Subnets quickly discovered that reckless emission design creates extreme sell pressure, and many teams began refining their economics to survive. A major lesson: subnets had to burn part of miner emissions or restructure incentives to avoid becoming unsustainable.
Metanova (SN68) was mentioned as a case study: its incentive mechanism initially pushed miners into adding more and more compute, which increased sell pressure.
From “Digital Commodities” to Real Infrastructure
Const highlighted how far Bittensor has come in terms of actual utility. In 2023–2024, some of the “best” subnets were essentially just spending emissions on OpenAI credits. That was the standard.
Today, the ecosystem looks completely different.
He pointed to Chutes as a major turning point, describing it as a general-purpose inference backbone serving more open-source model queries than any provider on earth, built in a decentralized and economically sustainable way.
The Community Became the Network’s Intelligence
One of the strongest points in the discussion was that the market didn’t improve because validators or foundation leadership controlled it. It improved because the community did.
Const credited the “weird but aligned” mix of dTAO traders, stakers, researchers, and subnet supporters for making Bittensor more efficient than it ever was under the old Root network model.
He argued that a functioning market made TAO allocation dramatically smarter. Compared to the old system, he said the network is now 100x more efficient at distributing emissions.
Integration Is Becoming the New Normal
A key shift is happening inside the ecosystem: subnets are no longer just competing in isolation. Const emphasized that even though Bittensor doesn’t directly reward subnets for integrating with each other, the ecosystem is naturally moving in that direction anyway.
He noted that Chutes is now the most integrated subnet in the network and also the most rewarded. That’s not written into the protocol. It’s an emergent outcome of what the market values: subnets that become foundational infrastructure win.
He described Bittensor as a stack, ranging from raw compute at the bottom to agentic workflows at the top, and said the most valuable subnets are the ones that connect those layers.
Const’s Major Announcement: Stepping Down as CEO
The biggest headline from the Novelty Search was Const’s announcement that he has stepped down as CEO of the OpenTensor Foundation.
He made it clear he’s not leaving Bittensor; he’ll be more involved, but now from inside the ecosystem rather than controlling it. His message was that Bittensor cannot claim to be unstoppable or permissionless if the chain itself is still controlled by a small group.
Right now, he admitted the chain isn’t fully decentralized. The Foundation still funds infrastructure, runs key nodes, and maintains development at a cost of millions per year. That centralization is a vulnerability, especially as Bittensor becomes visible to governments and large institutions.
So the plan is to remove that weakness completely.
The Next Step: Decentralizing Chain Nodes
The Foundation plans to decentralize chain nodes within the next six months. The first rollout will likely use a nominated proof-of-stake model, gradually distributing node operation across multiple organizations before moving toward a fully incentivized network.
Const also confirmed that transaction fees will be introduced, and those fees will help fund decentralized infrastructure. He suggested the chain could generate a few million dollars in transaction fees based on current activity.
The end goal is a chain that is harder to shut down, harder to control, and impossible to manipulate through centralized actors.
Governance Is Coming: A Real DAO-Like Structure
The long-term goal is decentralized governance and Const says it’s coming before the end of 2026.
He referenced the governance framework described in the original dTAO paper, including a Triumvirate model where subnet owners and validators gain real power over upgrades. Members would be selected through randomized selection mechanisms rather than campaign-style elections, reducing the risk of political capture.
Governance proposals would go through a multi-step process:
- Proposals get ratified by a Triumvirate.
- Validators and subnet owners can vote to delay or stop upgrades.
- A 50% quorum vote can block changes.
- “Nay” votes double the delay window, slowing down contentious upgrades.
The idea is to balance speed with decentralization: hot fixes can ship fast, but major upgrades face scrutiny.
First Governance Target: Cutting Emissions for Exploit Subnets
Const also revealed that one of the first governance actions could be allowing votes to stop emissions for exploitative subnets.
He directly mentioned Subnet 104, accusing it of manipulating incentives and extracting value from the ecosystem. Under the new governance structure, the community could vote to remove emissions from such subnets, redirecting rewards toward real builders.
This would represent a major shift: instead of relying on Foundation authority, the network would defend itself through transparent governance.
Final Take: dTAO Worked
The message from the call was clear: dTAO worked, but it exposed everything including the good, the bad, and the ugly. It attracted scams, forced tokenomics evolution, created real infrastructure winners, and turned the community into a market-driven governance layer.
Now Bittensor is entering its next era of decentralizing chain operations, decentralizing decision-making, and building a governance system capable of defending the network without centralized authority.