An article by author Konstantin Tkachuk opens the second installment of his series on DePIN with an eye-catching thesis: DePIN 1.0 is dead; what remains is Infrastructure-as-a-Service. This thesis is half right, and the correct half is the part worth discussing.
The model he outlines is backed by data. AWS accounts for 18% of Amazon’s revenue but contributes 57% of its operating profit. In telecommunications, the three largest cell tower owners in the United States grew their revenue from $4.5 billion (in 2010) to $19.4 billion (in 2024) on the back of long-term leases, while mobile carriers drove one another into price wars, causing ARPU to tumble from $47.5 to $33.4 over the same period. This is the law Clayton Christensen formulated back in 2003: when one layer of the stack commoditizes, attractive profits shift to an adjacent layer. Tkachuk’s insight lies in applying this principle directly to crypto.
The warning signs across crypto are already visible. Kalshi users trade egg prices while relying on crypto deposit rails underneath via Zero Hash (USDC, BTC, SOL). Hyperliquid users experience the exchange first, with the underlying chain operating in the background. PayPal users see PYUSD rather than LayerZero. Stablecoins have exceeded $310 billion in supply and $10.2 trillion in annual transaction volume—and virtually no one pays attention to which blockchain they settle on. Crypto is turning into a feature, not the product on sale.
Meanwhile, DePIN is experiencing a rout. Storj—one of the few projects with genuine revenue—filed for Chapter 11 in July, even while describing its “underlying business as sound and right-sized.” Helium Mobile was sold off to Noble Mobile without ever turning a profit. This summer alone, at least three pioneering DePIN companies restructured, sold, or uncoupled their operational businesses from their token units.
The most alarming metric lies in the lack of transparency. Messari tallied roughly $72 million in on-chain revenue across the entire DePIN sector in 2025, whereas Aethir alone self-reported $127.8 million because enterprise contracts settle off-chain where dashboards cannot see them. An industry born out of on-chain transparency cannot even audit itself.
The core irony centers on the enterprise procurement “wall.” Large enterprise clients demand SLAs, SOC 2, MSAs, 24/7 support, and fiat invoicing—no one inquires about tokens or underlying chains. Tkachuk summarizes it neatly in a single line: they offer tokens and dashboards, while clients demand an MSA and a support hotline. The intermediary crypto sought to displace has returned, stepping in as a quality guarantor between the network and the client.
Therefore, “DePIN is dead” warrants a more accurate interpretation. What has died is the model of issuing tokens to mobilize hardware and subsidizing supply indefinitely without a closed demand loop. What remains is a hybrid model: tokens still solve the cold-start problem in a way no Web2 company can, but a legal entity must sit on top to market and deliver the service. The real, unresolved questions are not whether DePIN is dead, but two things: as value migrates to the service layer, how much can tokens still capture? And with GPU pricing reaching parity with centralized neo-clouds, does “decentralization” remain a defensible moat, or is it merely a narrative for fundraising?
Stablecoins won by vanishing into the product. If DePIN aspires to become crypto’s second killer feature, it has only one path forward: deliver reliable, accountable services instead of selling narratives.