Last week, I saw a headline that almost made me spit out my morning coffee: “3M and Microsoft are independently building AI data center infrastructure.” Not because it’s surprising—we all know the hyperscalers are pouring billions into compute. But 3M? The company that makes Post-it Notes and Scotch tape is now a core player in the AI backbone. That’s not a diversification move. It’s a signal that the demand for compute has shattered traditional industry boundaries. And for anyone watching Web3, it’s the loudest whisper that decentralized physical infrastructure networks (DePIN) are no longer a theoretical play—they are the only logical next step.
Let me be clear: this is not about whether Microsoft’s Azure will dominate. It’s about the fact that the infrastructure itself is becoming a bottleneck so severe that even industrial material giants are pivoting. The article I read from Crypto Briefing—yes, a crypto outlet covering industrial materials—gave almost no technical details. No chip specs, no PUE ratios, no timeline. But that emptiness is its own kind of data. It told me that the market is so hungry for compute narratives that even a vague announcement moves the needle. And that’s exactly where blockchain’s role becomes non-negotiable.
Code is law, but people are truth.
I’ve been in this space since 2017—back when I launched CapeHorizon, a DAO for funding Cape Town’s creative arts. I coded the smart contracts myself, raised $120K in ETH, and watched it collapse under Ethereum’s gas fees during the November congestion. That failure taught me something that most AI infrastructure builders still ignore: centralization creates fragility. Microsoft and 3M building giant server farms might satisfy peak demand for now, but they introduce single points of failure—both technical and geopolitical. A power outage in Virginia or a trade war over chip exports could wipe out training runs worth millions. Decentralized compute, on the other hand, distributes risk across a global mesh of nodes. It’s not just idealism; it’s resilience engineering.
Embrace the volatility, find the signal.
The signal here is DePIN. Projects like Akash, Render, and io.net are already tokenizing GPU compute, allowing anyone to rent out idle hardware. But the real insight from the 3M news is that the compute demand is expanding beyond GPUs into the physical layer: cooling systems, connectors, power management. That’s where blockchain can do more than tokenize—it can verify. Imagine a smart contract that audits the energy efficiency of a cooling unit in real time, rewarding operators for hitting PUE targets. Or a liquid-staking derivative tied to hardware utilization rates. I saw a glimpse of this during my 2020 DeFi liquidity trap— when I churned through three yield farms simultaneously and accidentally discovered the composability risks of leveraged strategies. The same complexity applies to compute markets. Without transparent, on-chain proof of work (literally), we’ll repeat the same trust failures.
But here’s the contrarian angle: most DePIN projects today are vaporware. They raise hype, sell tokens, and deliver a fraction of promised capacity. The real opportunity might not be in the token layer at all, but in the physical infrastructure itself—just like 3M is doing. The commoditization of compute hardware means the biggest winners could be the material suppliers and logistics providers, not the blockchain protocols. My 2021 NFT experiment, AfricanCode, taught me that buzz without operational discipline leads to stagnation after the mint. DePIN faces the same risk: a thousand GPU-sharing tokens will compete for attention, but only the ones with actual hardware contracts and verifiable uptime will survive.
Vibes > Algorithms – but only when the vibes are backed by auditable data.
So what does this mean for the bear market? Survival matters more than gains. I spent the 2022 crash studying ZK-rollups while my portfolio dropped 70%. That curiosity led me to a key insight: zero-knowledge proofs can be applied to compute verification. Proof of Replication, Proof of Space, even zk-SNARKs for training runs—allowing a user to trust that a remote GPU executed the correct model without exposing the data. This is the bridge between AI and Web3 that no one is talking about. Microsoft and 3M are building walls; blockchain can build windows.
Build in public, live in truth.
My 2026 project TruthChain—focused on authenticating AI-generated content via on-chain proofs—validated this direction. We onboarded 10,000 users who wanted to verify source material. The same principle applies to compute: you need to know that the machine you’re renting is actually running your workload, not hallucinating results. Centralized data centers offer Service Level Agreements (SLAs) backed by legal contracts. Decentralized networks need cryptographic SLAs. And that’s exactly what the next wave of Layer2s and ZK-co-processors can provide.
Final takeaway: The 3M-Microsoft news isn’t about two companies building sheds. It’s a signal that compute has become as essential as energy. And just like energy grids moved from monopolies to microgrids with solar and batteries, compute must move from hyperscaler lock-in to a decentralized mesh. The tokenized compute market could surpass DeFi in total value locked by 2028. But only if we stop hyping and start building the infrastructure—both physical and cryptographic. Embrace the volatility, find the signal. The signal is DePIN. Now let’s verify it.