The Creator of Taostats Has Stopped Checking TAO Price Every Morning. Here’s What He Watches Instead

The man behind Taostats, the most popular explorer for the Bittensor network, has largely stopped checking its most-watched metric. For the better part of a year, price has had no place in Marcus ‘Mog’ Graichen’s morning routine.

Speaking with Alex Richardson at Exploit Summit, the Taostats creator explained why his attention has shifted toward a more demanding question.

Is the network turning its resources into products people actually need?

Mog came to Bittensor through media, web design and product development, with experience spanning the BBC, ticketing and pharmaceutical businesses.

As AI began reshaping everyday tasks, he saw an opportunity to help bridge technical development and commercial use. The network needed people who could explain its capabilities, understand customers and turn complex technology into accessible products.

That background shapes his interest in teams that adapt as they discover where their work is useful. He highlighted Score (SN44), which progressed from sports outcome prediction into football vision modelling and broader computer vision applications.

Mog also addressed criticism that many subnets receive more in emissions and token rewards than they earn in revenue. He regards that subsidy as a deliberate mechanism for funding decentralized AI and its supporting infrastructure. This gives teams time to develop without accepting restrictive venture deals, while requiring them to compete continuously against measurable performance criteria.

He also mentioned SayGM (SN28), an inference subnet positioned as a competitor to OpenRouter. He watches its incentive burn for the point at which genuine demand consistently exceeds the daily dollar value of the subsidy paid to miners. That would provide evidence that customer usage can support the service.

Wide adoption brings expectations that go beyond community support. Mog has previously argued that many subnets are not yet ready for the scrutiny accompanying higher attention and capital. Early supporters may accept experiments and resets, but paying customers expect a service to work reliably. A product’s decentralized foundations provide no exemption from that standard.

That challenge extends to Hippius (SN75), which Mog co-founded to develop decentralized storage infrastructure, with virtual machines planned. Its ambitions include alternatives to Amazon S3 and consumer storage services such as Google Drive, Dropbox and iCloud. Giving users control over where their data and model weights reside is part of the proposition, but winning their business requires competing with services already embedded in their routines.

For founders, execution also depends on understanding how Bittensor rewards work. A promising idea can struggle if the team cannot translate its intended outcome into tasks that miners have an incentive to perform well. Effective subnet design connects the competition directly to improving customers’ value, giving technical development a clear commercial purpose.

New contributors have more ways to navigate those mechanics as AI agents make documentation easier to explore and explain. People with expertise in areas such as video compression, prediction markets or model training can use these tools to understand how their knowledge fits into the network.

Mog does not believe Bittensor needs to become a household name to succeed. Like TCP/IP, it could underpin widely used products while remaining unfamiliar to their customers. GPU marketplaces, inference platforms and specialized applications would earn recognition in their own markets, with the network’s progress visible in the businesses and workflows that come to depend on them.