If you think of an AI computing network as an “on-call service desk,” the real challenge isn’t how much computing power it can offer, but whether supply can meet demand when a task comes in. Who is willing to pay for results, what problem the service they buy solves, and whether they buy again later—those are what validate product demand. An increase in AI applications across the industry doesn’t mean these networks have already secured orders; without project-level data on customers, revenue, or utilization, I won’t claim they’re succeeding.
Over the rolling 24-hour period, $TAO -2.81%. Over the rolling 24-hour period, $RENDER -3.06%. Over the rolling 24-hour period, $FET -5.48%. These are changes in the respective spot prices, not task volumes, customer counts, or computing power utilization; the declines don’t prove that demand weakened at the same time, much less that progress on any particular product caused the drops.
To assess the supply and demand for computing power, you need to look at both sides: whether customers’ tasks can be matched and reliably delivered, and whether the compensation providers receive covers their resource costs. Network incentives may attract supply, but rewards are not service fees paid by customers. Who funds the rewards, whether the mechanism is merely proposed or already implemented, and whether supply remains after rewards are reduced—all of these questions require specific rules and data to answer. Devices being online only shows that resources are available; it doesn’t mean anyone is paying for them.
Even if customers pay for a product, that doesn’t automatically establish the value of its token. Does using the service require the token? Where do the fees go? Can increased demand translate into sustained demand for the token? Each of these steps needs to be verified separately; product usage, network delivery, and benefits to token holders can’t be collapsed into a single conclusion. Nor can growth in a company’s AI business be counted directly as orders for these networks.
I’d rather see a specific customer need go from payment to delivery and then to repeat purchases, while also making clear who pays the service fees and who funds the network rewards. What we can confirm now is the changes in three spot prices; the available information still doesn’t tell us who already has sustained paying demand or how value flows to the tokens.
Over the rolling 24-hour period, $TAO -2.81%. Over the rolling 24-hour period, $RENDER -3.06%. Over the rolling 24-hour period, $FET -5.48%. These are changes in the respective spot prices, not task volumes, customer counts, or computing power utilization; the declines don’t prove that demand weakened at the same time, much less that progress on any particular product caused the drops.
To assess the supply and demand for computing power, you need to look at both sides: whether customers’ tasks can be matched and reliably delivered, and whether the compensation providers receive covers their resource costs. Network incentives may attract supply, but rewards are not service fees paid by customers. Who funds the rewards, whether the mechanism is merely proposed or already implemented, and whether supply remains after rewards are reduced—all of these questions require specific rules and data to answer. Devices being online only shows that resources are available; it doesn’t mean anyone is paying for them.
Even if customers pay for a product, that doesn’t automatically establish the value of its token. Does using the service require the token? Where do the fees go? Can increased demand translate into sustained demand for the token? Each of these steps needs to be verified separately; product usage, network delivery, and benefits to token holders can’t be collapsed into a single conclusion. Nor can growth in a company’s AI business be counted directly as orders for these networks.
I’d rather see a specific customer need go from payment to delivery and then to repeat purchases, while also making clear who pays the service fees and who funds the network rewards. What we can confirm now is the changes in three spot prices; the available information still doesn’t tell us who already has sustained paying demand or how value flows to the tokens.
