TLDR: Bittensor, the leading AI-crypto protocol, is preparing to launch two significant upgrades: Dynamic TAO and EVM compatibility. The first addresses centralization and scalability issues, the second will allow developers to build decentralized applications directly on the network. The second is a welcome upgrade, as a common complaint with the current state of Bittensor is that there are basically no apps using it.
AI coins have shown relative strength in recent months whenever the market bounced.The leading AI coin is TAO. Bittensor has been the darling of VC in the AI space. It received almost $400 million in pre-seed and seed funding. Its Market cap is 3.5 B, which puts it first on the list of AI coins, before for example Render.
It rose sharply in price in September 2024, clearly separating itself from the pack.

What is Bittensor?
Bittensor offers censorship-resistant access to a decentralized network of machine learning models. Another way to explain this is that BitTensor outsources AI tasks such as text prompts and image generation to a decentralized network of computers. Instead of sending your prompt to one server, TAO miners compete to get you the best answer. These miners get paid in the TAO token.
Sounds cool right? Taking a step back, it is indeed quite remarkable. Bittensor boldly competes with full-stack AI giants like OpenAI and Google. Why? Because Bittensor engages its competitors in the entire AI lifecycle. Not just generating outputs but also creating new algorithms and training models.

What Are Bittensor Subnets?
Bittensor subnets are specialized, decentralized networks within the broader Bittensor ecosystem. Each subnet is designed to handle a particular type of AI workload. In each subnet, miners perform the computational tasks (e.g., training models, storing data) specified by that subnet. Validators verify the quality and performance of the miners’ outputs.
Here are a few examples of subnets:
- Text prompting
- Translation
- Image generation
- Transcription
- Storage

Criticisms of Bittensor
There are two main areas of critique. The first is that there is a lot of redundant computation. Eric Wall in a viral Tweet put it like this:
“For every prompt, you have literally a thousand miners who will complete the exact same task, redundantly.”
Isn’t that also the case with Bitcoin miners? Not quite. Every extra honest Bitcoin miner adds someething to the network. Every extra miner makes it harder to attack the network. Ok, but won’t extra Bittensor miners make the outputs of the network of higher quality? This is what Eric Wall doubts:
“The network will validate these answers by checking how similar they are to each other. If you’re an outlier you don’t get TAO tokens.[…] they’re all incentivized run the same LLM because outlier responses are punished.”
Critique 2: Demonstration of Usability
Despite a large market cap, Bittensor lacks a strong user base and active applications. There are only a few apps leveraging its technology, and most have limited functionality. This lack of adoption undermines the value proposition of the project (Even its founders have struggled to explain or demonstrate practical use cases, which raises doubts about the project’s real-world applicability. When one of its founders was asked in October 2023 to demonstrate an app, he couldn’t.). One of the updates mentioned above will address that issue, by the way.
Critique 3: Tokenomics
Bittensor’s tokenomics are viewed as unsustainable, with 7,200 TAO tokens minted daily, putting pressure on its price. There’s also ambiguity around the vesting schedule for early investors, potentially posing a risk of large sell-offs.
New Developments: Dynamic TAO and EVM Compatibility
As mentioned, Bittensor is preparing for two major updates—Dynamic TAO and EVM compatibility. Let’s have a look at these in turn.
Dynamic TAO: Each Subnet its Own Token
Currently, TAO emissions are distributed by a small group of validators, which is a centralization bottleneck. The Dynamic TAO upgrade will decentralize this process by introducing market-driven emissions.
Each subnet will then have its own token, paired with TAO in liquidity pools, allowing the broader market to signal a subnet’s value through staking and swapping. This means that early holders of successful subnets will benefit, as they can accumulate relatively more of the subnet token supply. The “Dynamic” aspect is that the more TAO is staked in the subnet pool, the more TAO emissions are allocated to the subnet. And the higher the subnet token price can go.
This will lead to increased yield opportunities. TAO holders will be able to directly engage with specific subnets, taking on higher risk but potentially gaining significant returns. However, staking TAO in these pools is risky, as the token exchange rate in each pool will fluctuate with demand.
EVM Compatibility: Enter Smart Contracts
By adding Ethereum Virtual Machine (EVM) compatibility, Bittensor enables developers to create on-chain applications and smart contracts directly within its ecosystem. This will transition Bittensor from an appchain network to a more versatile platform for decentralized applications.
This is a big deal. After the EVM compatibility, TAO can be collateral in decentralized finance (DeFi) applications, a transaction gas token, and a key player in new smart contract use cases. This boosts demand and brings new avenues for TAO-based yield strategies.
Conclusion
Bittensor’s goal is to decentralize control of AI. As with all blockchain projects, the question is if the computational redundancy that is required justifies the benefits. While we can debate this ad nauseam, the market seems to have decided it likes Bittensor. Bittensor has a narrative that people (think they) can grasp: it relates to a driver of economic growth in the ‘real’ world. At the same time, Bittensor keeps building. The EVM compatibility will mean more utility and demand for the TAO token. The added decentralization might encourage a leading exchange like Coinbase to list TAO, benefiting liquidity and potentially, price.
Erik started as a freelance writer around the time Satoshi was brewing on the whitepaper.
As a crypto investor, he is class of 2020. More of a holder than a trader, but never shy to experiment with new protocols.