James Altucher: “Here’s How Bittensor Prevents the Next Covenant”

The TAO Daily team chatted with James Altucher about what Bittensor should do differently after the Covenant exit, and what he thinks every subnet investor should be paying attention to going forward. Altucher is an entrepreneur, investor, ardent $TAO lover, bestselling author, and someone who’s been thinking about incentive designs and market structure for decades.

Here’s what he had to say.

On Fixing the Network

We asked Altucher what he’d change at the protocol level to prevent another Covenant-style exit. He gave us some ideas.

Structured sell-down schedules. “Make it more structured, like the SEC’s 10b5-1 plans, where an executive files a schedule by which they will sell shares,” he said.

In traditional markets, executives can’t just wake up and dump their stock. They file a plan in advance. It’s public. It’s spread over time. It’s designed so the market can absorb it without a shock.

Altucher thinks subnet owners should operate under something similar. An on-chain schedule showing when and how much they plan to sell. Investors see it coming weeks ahead, and nobody gets ambushed at midnight.

The Locked Stake proposal from Const is a good start, Altucher said, but he wants it taken further in this direction.

Learn more about Locked Stake below:

Sandbox the miners’ work. This one is about the actual AI work done on subnets. “Have miners contribute work only to a sandbox which is backed up into cloud,” Altucher told us. “This way if a subnet owner leaves, the work of the miners is still preserved, and a new subnet owner can take over the tech.”

Right now, when a founder walks, there’s a risk the technical contributions become inaccessible or fragmented. Altucher’s argument: the models, the training data, the code, all of it should survive independently of whoever holds the owner key. Build the infrastructure so the work outlives the person.

More subnets. “There’s always going to be an 80/20 rule,” Altucher said. “20% of the subnets will make 80% of the value. The more subnets there are, the less likely it is for one subnet to cause chaos like this.”

It’s a scale argument. When Covenant ran three of the most visible subnets on the network, its exit could shake the entire ecosystem. If Bittensor had 500 subnets instead of 128, the same departure would barely move the needle. More diversity means less concentration risk.

Build a flywheel, not a research project. This is the one Altucher was most emphatic about.

“When I spoke with Sam Dare last September, he didn’t really seem aware that he had to create value in the subnet tokens in order to incentivize subnets,” Altucher told us. “He sort of thought of Templar as a ‘research project’ that people would want to contribute to. That’s nice in theory but unhealthy in practice.”

The subnets that are thriving on Bittensor, he pointed to Chutes and Ridges specifically, have built real economic flywheels. Miners contribute because the economics make sense. Investors hold because the token captures real value. The whole thing reinforces itself.

What Every Subnet Investor Should Be Asking

Beyond the network-level fixes, we asked Altucher what a regular investor should look for before putting capital into any subnet. He broke it down into four things.

First: how dependent is the subnet on its miners? This is the survivability test. “Ridges, as an example, or Chutes, they cannot survive without their miners,” Altucher said. That’s actually a good thing. It means the miners are the real value, not the founder. If the founder walks away from a subnet where miners are essential to the product, the miners still have every reason to keep working. The economic engine runs without the person who built it.

If a subnet works fine without miners, that’s a red flag. It means the value lives in the founder’s head, and the founder can take it with them.

Second: how are subnet owners incentivizing miners? “Is it integral to their business model?” Altucher asked. “For Templar, it was not.” Miners contributed compute to a research mission. When the mission walked away, so did the rationale for mining. This was why they were burning 100% miner emissions after training runs.

Compare that to a subnet where miners earn because the product generates revenue and that revenue flows back through emissions. That’s a flywheel that can survive any founder exit.

Third: diversification. The simplest advice and the most ignored. The Covenant exit hurt most the people who were concentrated in SN3, SN39, and SN81. If that same capital had been spread across ten subnets, the damage would have been a portfolio event, not a personal financial crisis.

Fourth: invest in a basket of SOTA subnets. Altucher got specific here. “Invest in a basket of subnets that are state-of-the-art in their industries,” he said. “Bitmind, Yanez, Ridges come to mind, and many others.”

Below are the SOTA subnets on Bittensor:

The Takeaway

Everything Altucher said comes back to one idea: invest in machines, not people.

Structured sell schedules replace trust with transparency. Sandboxed work replaces hope with backups. More subnets mean less concentration risk. Flywheels make ecosystems self-sustaining. And as an investor, the question is always the same: if the founder disappears tomorrow, does the subnet still run?

If yes, you’re investing in infrastructure. If no, you’re investing in a person.

The Covenant exit taught the entire network the difference.