Bittensor’s most valuable asset is shifting from “stories” to “invoices”: revenue-generating subnets have reached $70 million in annualized revenue, doubling in two months. The number itself doesn’t change the valuation, but it changes the way people talk about TAO—from “believing in the AI narrative” to “whether the numbers add up.”
First, the growth rate. Going from around $35 million to $70 million in two months is 100% growth. The absolute figure isn’t particularly remarkable in the AI sector, but the direction is worth noting. Revenue curves for any crypto project rise steeply from a low base in the early stages, so you can’t extrapolate linearly and assume it will “double several more times over the next year.” That’s the first dose of cold water. The genuinely valuable information is that, for the first time, this curve has a metric that can be tracked over time, rather than relying on a new narrative to stay alive each time.
The second point is the shift in the basis for pricing. In the past, TAO’s price anchor was roughly a combination of expected emissions and an AI narrative premium. Buying it was essentially buying an option on a “future compute network,” with no denominator for valuation. When subnets start generating real payments, the anchor shifts to discounted cash flows—even if the discount rate is very high and the denominator is still pretty rough, the nature of the valuation is entirely different. The more important difference is in the flow of funds: emissions are subsidies paid to miners through inflation, and they naturally create persistent selling pressure; paid revenue comes from external capital flowing in, the opposite direction. This isn’t a change in sentiment; it’s structural.
The third point—and the one I think is most important to watch going forward—is where the money ultimately lands. What is the revenue denominated and settled in? If it remains in the Alpha tokens of individual subnets, it improves the subnets’ balance sheets, but TAO itself captures very little. For $70 million to matter to TAO, there needs to be a transmission mechanism—for example, participating in a subnet might require staking or locking up TAO, or some portion of subnet revenue might be used to buy back, burn, or collateralize TAO. Until this mechanism is in place, $70 million is subnet revenue, not cash flow for TAO. This is where many AI-themed tokens fall short: the ecosystem is bustling, while the token itself remains neatly outside the flow of funds.
The fourth point is revenue quality, the risk most easily overlooked. When revenue doubles in two months, three questions need to be asked: How much comes from real external payments, and how much from internal ecosystem circulation—subnets paying one another, or “usage” driven by incentives? Is revenue highly concentrated in two or three subnets? If so, it’s closer to project-based revenue than a network effect. And emissions subsidies are still in place: if subsidies for a subnet are reduced, will its demand disappear as well? Only demand that can withstand declining subsidies and still keep paying is genuine.
My inclination is to put this on the watchlist rather than treat it as a conclusion. Just watch two things: whether revenue steps up again next period or flatlines, and whether the project team clearly explains the settlement path—how revenue translates into buying pressure for TAO. If either of these two things materializes, the $70 million figure will carry a different weight.
Risk disclaimer: Early-stage revenue data is based on a small sample, and reporting definitions may change. Don’t use the growth rate from a single quarter as a substitute for verifying the settlement mechanism.
#TAO #Bittensor #Crypto
First, the growth rate. Going from around $35 million to $70 million in two months is 100% growth. The absolute figure isn’t particularly remarkable in the AI sector, but the direction is worth noting. Revenue curves for any crypto project rise steeply from a low base in the early stages, so you can’t extrapolate linearly and assume it will “double several more times over the next year.” That’s the first dose of cold water. The genuinely valuable information is that, for the first time, this curve has a metric that can be tracked over time, rather than relying on a new narrative to stay alive each time.
The second point is the shift in the basis for pricing. In the past, TAO’s price anchor was roughly a combination of expected emissions and an AI narrative premium. Buying it was essentially buying an option on a “future compute network,” with no denominator for valuation. When subnets start generating real payments, the anchor shifts to discounted cash flows—even if the discount rate is very high and the denominator is still pretty rough, the nature of the valuation is entirely different. The more important difference is in the flow of funds: emissions are subsidies paid to miners through inflation, and they naturally create persistent selling pressure; paid revenue comes from external capital flowing in, the opposite direction. This isn’t a change in sentiment; it’s structural.
The third point—and the one I think is most important to watch going forward—is where the money ultimately lands. What is the revenue denominated and settled in? If it remains in the Alpha tokens of individual subnets, it improves the subnets’ balance sheets, but TAO itself captures very little. For $70 million to matter to TAO, there needs to be a transmission mechanism—for example, participating in a subnet might require staking or locking up TAO, or some portion of subnet revenue might be used to buy back, burn, or collateralize TAO. Until this mechanism is in place, $70 million is subnet revenue, not cash flow for TAO. This is where many AI-themed tokens fall short: the ecosystem is bustling, while the token itself remains neatly outside the flow of funds.
The fourth point is revenue quality, the risk most easily overlooked. When revenue doubles in two months, three questions need to be asked: How much comes from real external payments, and how much from internal ecosystem circulation—subnets paying one another, or “usage” driven by incentives? Is revenue highly concentrated in two or three subnets? If so, it’s closer to project-based revenue than a network effect. And emissions subsidies are still in place: if subsidies for a subnet are reduced, will its demand disappear as well? Only demand that can withstand declining subsidies and still keep paying is genuine.
My inclination is to put this on the watchlist rather than treat it as a conclusion. Just watch two things: whether revenue steps up again next period or flatlines, and whether the project team clearly explains the settlement path—how revenue translates into buying pressure for TAO. If either of these two things materializes, the $70 million figure will carry a different weight.
Risk disclaimer: Early-stage revenue data is based on a small sample, and reporting definitions may change. Don’t use the growth rate from a single quarter as a substitute for verifying the settlement mechanism.
#TAO #Bittensor #Crypto