A few DePIN/AI projects have real paying demand; most are narrative tokens. Strip out emissions and ask who actually pays - that test does most of the work.
AI-and-crypto tokens and DePIN - decentralized physical infrastructure networks for compute, storage, wireless, and sensors - are among the hottest narratives in the space. The pitch is using crypto incentives to bootstrap real AI or physical infrastructure. Some projects have genuine usage; most are narrative tokens riding the AI hype with thin fundamentals.
The whole question is whether real customers pay, or whether token emissions are the only "revenue."
DePIN networks use token incentives to bootstrap real infrastructure - compute, storage, wireless coverage, sensor data - and the best have actual paying customers, which is a genuine thesis. AI tokens span everything from compute marketplaces to data and agent networks.
But the AI narrative commands a premium that pulls in a flood of projects with thin fundamentals. Many show "usage" that is really subsidized by token emissions rather than paying demand, and they compete with well-funded centralized incumbents.
| Segment | How it behaves as an asset |
|---|---|
| DePIN / AI with real, paying demand | The thesis tier |
| Promising but subsidy-dependent | High risk; demand may be emissions |
| Pure narrative tokens | Speculation on the story |
| (Competition) | Centralized incumbents pressure the model |
| Point | Why it matters |
|---|---|
| Some have real demand | DePIN with paying customers is a thesis. |
| Most ride the narrative | AI hype pulls in thin-fundamental tokens. |
| Subsidy ≠ demand | Emissions-funded usage can evaporate. |
| Competition is real | Centralized incumbents pressure the model. |
| Emissions dilute | Heavy emissions mask weakness. |
AI and DePIN are the narratives doing the most work in crypto right now, and that is exactly why they demand skepticism. The best DePIN networks - using token incentives to bootstrap real compute, storage, or wireless infrastructure with actual paying customers - have a genuine thesis. The problem is that the narrative premium pulls in a flood of imitators.
The cleanest test I use is to strip out token emissions and ask who is left paying. A lot of impressive "usage" turns out to be activity subsidized by the token itself, which evaporates the moment emissions slow. That is a story funding itself, not a business.
My take: separate paying demand from emissions-funded activity, discount the AI-narrative premium hard, check whether the token captures any revenue, and respect that well-funded centralized incumbents may simply do the same thing cheaper.
The scanner looks past the AI narrative to real, paying demand and token value accrual, and the Vault tracks the networks with genuine usage over time.
A few have real, paying demand and a genuine thesis - especially DePIN networks bootstrapping compute, storage, or wireless infrastructure with actual customers - but most are narrative tokens riding the AI hype with thin fundamentals. The key test is whether real customers pay or whether token emissions are the only "revenue."
DePIN stands for decentralized physical infrastructure networks - projects that use crypto token incentives to bootstrap real-world infrastructure such as compute, storage, wireless coverage, or sensor data. The strongest DePIN networks have genuine paying demand, while weaker ones rely on emissions to simulate usage.
Strip out the token emissions and ask who is left paying - real, paying customers indicate genuine demand, while activity that disappears when emissions slow is subsidized hype. Also check whether the token captures revenue and how it competes with well-funded centralized incumbents.
The AI narrative commands a large premium that inflates valuations beyond fundamentals, drawing in many projects with thin underlying demand. When sentiment shifts or emissions-funded usage proves unsustainable, these tokens can fall sharply, making the category especially volatile.
Not always - a network can generate real usage while its token captures little of the revenue, and heavy emissions can dilute holders. Assessing token value accrual and the emissions schedule is essential, because usage growth does not automatically translate into token value.