Author: Momir Amidzic

Translation: Shenchao TechFlow

Shenchao Introduction: Momir Amidzic, managing partner at IOSG Ventures, conducted a calm dissection of Bittensor. His core argument is straightforward: TAO is essentially a funding program for AI research and development without any obligation to return profits, and subnetworks can take the money and leave at any time. In an optimistic scenario, the eternal thirst of AI for resources could keep subnetworks around; in a pessimistic case, this is a wealth transfer from token speculators to AI developers. The article is not long, but it clearly articulates the structural contradictions of Bittensor.

Bittensor has a very sophisticated narrative: a decentralized AI intelligence market that allocates funds to the most influential research through market forces. TAO is the coordination layer, subnetworks are the labs, and the market is the grant committee.

Peeling away the narrative, what remains is not so pretty.

Bittensor is a funding program where crypto speculators fund AI research and development, while the recipients have no obligation to return any value to TAO.

Imagine TAO as Elon Musk, the first investor in "non-profit" OpenAI. The subnetworks are Sam Altman, who took the money, made the product, with not a single word in the contract requiring them to share profits. In the end, they might privatize the profits and return nothing to the original funders.

Bittensor releases TAO to subnet operators and miners based on the price of subnet tokens. Once a subnetwork receives a TAO allocation, there are no mandatory mechanisms requiring the resulting AI models, datasets, or services to remain in the Bittensor ecosystem. Subnet operators can create valuable outputs, withdraw TAO emission rewards, and then deploy the products elsewhere, moving to centralized clouds, independent APIs, or packaging them into a regular SaaS company.

TAO has no equity, no licensing rights. The only tether is the subnet token, the token price needs to perform well to continue attracting resources. But this only works until the subnet reaches escape velocity; once a product is good enough to operate independently outside of Bittensor, that tether breaks. The relationship between Bittensor and subnetworks is more akin to research funding rather than venture capital.

From this perspective, Bittensor represents a wealth transfer from token speculators to AI researchers. To put it more bluntly, it’s a shift from speculators to technically skilled farmers.

The mechanism is simple:

• TAO investors provide capital by supporting the market price of TAO
• Subnetwork operators exchange "proof of performance" for TAO inflation rewards, essentially maintaining the price of subnet tokens
• AI products built with this capital can leave at any time; the only constraint is whether they still need resources

This is a nightmare scenario for VCs: you invest money, the company produces something, and then it owes you nothing. All that’s left is a token release plan and a prayer.

Optimistic Interpretation

Now let's flip the perspective. The optimistic argument is built on two pillars:

1. Continuous resource thirst. AI companies are always short on cash. Computing power, data, and talent are expensive. If Bittensor can stably provide these resources at scale, subnetworks have rational incentives to stay, not because they are locked in, but because leaving means losing access to the resource pipeline. The soft assurance is that AI will always need more resources, and TAO can provide a scale that separate financing cannot match. In this logic, subnetwork teams will proactively maintain their token valuation, creating a positive feedback loop for TAO’s economy without any compulsory mechanisms.
2. Crypto has a unique ability to aggregate resources. Bitcoin aggregated massive computing power solely through token incentives. Ethereum's proof of work is an extremely successful magnet for computing power. Bittensor is applying the same playbook to AI. The "compulsive mechanism" is the token game itself; as long as TAO has value, there is incentive for participation.

If we run 1000 simulations of Bittensor’s future paths, the distribution will be severely skewed.

In most paths, Bittensor remains a niche funding program. Subnetwork outputs marginal AI outcomes. Those that perform best gain some traction and, after cashing out rewards, shift to closed modes, returning nothing to TAO. Emissions exceed the value created, and the token price drops.

In a few paths, something clicks. A certain subnetwork produces a genuinely competitive AI service. Network effects begin to compound. TAO becomes a significant coordinating layer for decentralized AI infrastructure, capturing value not through compulsion but through the gravitational pull of being a reserve asset in an operational AI economy.

In very few paths, TAO becomes a defining asset.

What Could Go Wrong

The bearish argument is quite simple:

No Stickiness. Once a subnetwork doesn’t need emission rewards, it will leave. Bittensor is a stepping stone, not the destination.

Centralized AI is Winning. OpenAI, Google, and Anthropic control a much larger magnitude of computing power and talent. TAO cannot compete with the depth of the VC and PE markets. The best talent will follow conventional routes.

Inflation is a Tax. The TAO emission schedule dilutes holders to fund subnetworks. If a subnetwork doesn’t create corresponding value, this is chronic bleeding disguised as a growth mechanism.

The optimistic argument frankly seems more like wishful thinking than a realistic path to success.

Conclusion

Most of the funds invested in TAO will ultimately support development work that will never give back to token holders. But the crypto industry has repeatedly proven that token incentive coordination mechanisms can yield results that any rational model cannot predict. Bitcoin shouldn't work, but it does. However, this is a weak argument in itself, the industry uses it to endorse countless ideas, many of which cannot stand in the face of first principles.

The problem with TAO is not the lack of compulsion; there is none, and dTAO’s efforts won't change that. The issue is whether the game-theoretic incentives are strong enough to keep the best subnetworks on track. If you're buying TAO, you're betting on a soft assurance that can hold up in a hardcore world.

This is either naive or visionary.