TLDR;

• Protocol Upgrade Adds Uniswap v3-Style Pools: A new upgrade enables optional user-supplied liquidity with a 0.3% protocol fee distributed to LPs—potentially deepening Alpha markets. However, LPs currently forgo Alpha emissions, creating an opportunity cost that may hinder adoption.

• Revenue is the New Narrative: Projects like Targon and Gradients are pioneering Alpha token buybacks, while initiatives like DSV’s Revenue Search are pushing the ecosystem to prove real-world viability. Subnets with maturing business models and early revenue traction are gaining mindshare.

• Templar Ships v1.0.0 + CCLoco: Templar launched its first protocol version alongside CCLoco, a communication layer reducing training overhead by 50x. It’s a serious technical unlock for scalable, decentralized AI and represents a classic underpriced infrastructure play for investors.

• Infra & Tooling Improvements Go Live: Alpha trading infrastructure continues to mature with new features like TWAP orders, taostats limit orders, and Backprop’s Alpha feed—all aimed at making subnet markets more accessible and liquid.

• Subnet Highlights: Quantum (SN63) launched with its first benchmark, HashTensor (SN16) is planning a token burn using mined KAS, and SN42 unlocked TikTok as a new data source. Gradients, Synth, and Dojo continue shipping high-signal use cases across training, prediction, and frontend generation.

Highlights of the Week

⚡️The latest protocol upgrade introduces optional support for Uniswap v3-style liquidity pools, enabling user-supplied liquidity within defined price ranges. A flat 0.3% protocol swap (and move stake) fee is now applied to all staking transactions and distributed pro-rata to LPs in the active range. While subnet owners must opt-in, this addition is suspected to unlock deeper liquidity, stabilizes Alpha pricing, and create a new yield path for TAO and Alpha holders beyond mining and validation. A few subnets have implemented the pools such as Celium (SN51), and you can currently add positions through taostats.io.

Important consideration: LPs will not currently receive Alpha delegation rewards (emissions) on their LP positions, presenting a meaningful opportunity cost relative to simply holding spot Alpha tokens.

My take: Uniswap v3 is an important technical innovation for the chain, however, the fee also applies to other actions such as movestake, which is quite egregious (expecting this to be fixed soon). Ignoring the desire that this turns into a more dynamic market-driven fee mechanism (unimportant for now), I do believe that there needs to be a greater incentive for LPs. No one is going to sacrifice their alpha emissions in order to LP into these pools.

I am expecting this implementation to evolve throughout the course of the year and for it to take a different shape as more community discussion over the concept and it’s economics increase upon the release of more information and experimentation in production.

⚡️ The search for revenue is top of mind. The community has come together to encourage Alpha token buybacks, explore subnet revenue models, and maintain focus on proving that the network and its subnets can begin to sustain themselves independently of TAO emissions.

Initiatives like DSV Fund’s Revenue Search live streams, Neuromancer’s latest article on revenue flywheels, and Targon announcing their first Alpha buybacks are the early stages of a positive movement that forces the community to stay disciplined and ask themselves if these existing subnets are truly viable in a free market.

My take: These initiatives are definitely lighting a fire up everyone’s a** to be much more scrappy and intentional about what they are building, albeit the quest for revenue shouldn’t be a broad stroke painted on all subnets. There is obviously nuance across all subnets with varying levels of PMF-maturation.

Take action: Unless you are investing in moonshot ideas that fit a “venture” style return profile AND are confident in the team’s ability to execute, I would narrow my focus on identifying subnets that are already putting up revenue numbers or are operating in a space where there’s a proven business model across web2. I have an inclination that these subnets will significantly outperform the field.

⚡️ New and enhanced alpha token trading features, such as limit orders on taostats, TWAP on tao.app, and an alpha feed feature on Backprop Finance are now live. I have tested all of them and they work like charm! Try them out, let me know what you think.

⚡️ Templar announced the launch of Protocol v1.0.0 alongside CCLoco—a new communication layer designed to drastically reduce bandwidth requirements in distributed training. According to the team, CCLoco reduces communication overhead by 50x—down to just 2.8 GB per training run—enabling scalable, permissionless training while preserving model quality.

My Take: This is an important milestone that’s flying under the radar. As discussed in this week’s TAO Talk, Templar is one of the only live, incentivized distributed training systems in crypto. It’s highly technical, under-marketed, and led by deeply committed engineers, making it a classic information asymmetry play for investors. With state-of-the-art optimization happening in real time and zero effort spent on narrative-building, it represents a frontier bet closer to early-stage venture than simple trenches rotations.

Take Action: If you’re allocating in the subnets, it’s worth it to get in the weeds: read Templar docs, join the Discord Channel, and ask questions.

More below… ⬇️