I spent last Tuesday morning staring at a chart that made me uncomfortable. Not because of the red candles — we're all used to those. No, it was a market cap figure for a company that makes air conditioners. Vertiv hit $100 billion. A hundred billion dollars for cooling systems and power distribution units. And my first thought wasn't about the stock. It was about what this means for the story we've been telling ourselves about decentralization.
Here's the thing about the AI gold rush that nobody in crypto wants to admit: the real money isn't being made by the miners or the protocol developers. It's being made by the people selling the shovels. And right now, the biggest shovel seller in the AI infrastructure game is a company that, until recently, was best known for making sure your office server room didn't overheat.
I've been thinking about this since I audited my first Ethereum whitepaper back in 2017. Back then, we believed that code was law and that decentralization would flatten every hierarchy. We were young. We were optimistic. And we completely missed the fact that physical infrastructure — the stuff that actually runs the machines — would become the most concentrated, centralized bottleneck of all.
Let me walk you through what Vertiv's rise actually tells us about the future of AI, crypto, and the uncomfortable marriage between the two.
The Context: What Vertiv Actually Does
Vertiv isn't a tech company in the Silicon Valley sense. It's an industrial company that makes the physical systems that keep data centers alive: power distribution units, uninterruptible power supplies, thermal management systems, and — crucially for the AI era — liquid cooling solutions for high-density GPU clusters.
When NVIDIA's H100 draws 700 watts and the B200 pushes even higher, you can't just rely on fans and air conditioning anymore. You need precision liquid cooling. You need power systems that can handle the load without tripping. You need the kind of infrastructure expertise that takes decades to build.
Vertiv had that expertise. They just needed the market to catch up to their technology. And with the AI boom, it did.
The company's market cap crossing $100 billion isn't just a story about one company's success. It's a signal that the market is finally pricing in the physical reality of AI: that every model training run, every inference request, every autonomous agent action requires real-world electricity, real-world cooling, and real-world physical infrastructure.
The Core: What This Means for the Decentralization Narrative
Here's where I need to be honest with you, because this is the part that keeps me up at night.
We've spent years telling people that blockchain would democratize access to financial services. That decentralization would break the monopoly of centralized institutions. That your keys, your kingdom.
But the AI infrastructure buildout is creating a new kind of centralization that makes the old banking system look positively distributed.
Think about it. Who controls the physical infrastructure that AI depends on? A handful of companies: Vertiv, Schneider Electric, ABB, and a few others. Who controls the cloud platforms that host most AI workloads? AWS, Azure, Google Cloud. Who designs the chips? NVIDIA, AMD, and a few others.
The actual decentralization of AI isn't happening at the infrastructure layer. It's happening at the application layer — and that's a problem.
I saw this pattern before, during the ICO boom. In 2017, I audited over 40 Ethereum whitepapers and smart contracts for a boutique consultancy. I identified critical governance flaws in three major projects, including a $50 million Ponzi scheme disguised as a decentralized exchange. The pattern was always the same: beautiful rhetoric about decentralization, but the actual control sat with a small group of founders and early investors.
The infrastructure layer is even worse. You can't fork a data center. You can't create a DAO that votes on cooling system design. The physical world doesn't work that way.
The Contrarian Angle: Maybe Centralization Is the Price of Progress
Now, let me play devil's advocate with myself, because that's what I do when I'm trying to figure out whether I'm being paranoid or prescient.
Maybe the centralization of AI infrastructure is actually fine. Maybe it's even necessary.
Consider this: the reason Vertiv can build reliable, high-performance cooling systems is because they've had decades of experience and massive R&D budgets. A decentralized network of small cooling providers couldn't deliver the same reliability at scale. The same way you wouldn't want your hospital's life support system to be run by a DAO, maybe you don't want your AI training cluster to depend on a distributed network of hobbyist cooling enthusiasts.
There's a reason we have professional standards, certifications, and regulatory oversight for critical infrastructure. It's not because we love bureaucracy. It's because when things fail, people die. Or in the case of AI infrastructure, when things fail, billions of dollars of compute time goes up in smoke.
I've seen this tension play out in my own work. When I launched OpenLedger Academy in 2020, I was convinced that DeFi would democratize finance. I recorded 50 video tutorials and hosted three live Twitter Spaces, attracting 10,000 unique sign-ups within six months. I believed in the vision.
Then the FTX collapse happened in 2022, and I watched $50 billion evaporate because a centralized exchange had too much control. I pivoted my platform to focus on regulatory literacy and long-term holding strategies. I published a 10-part series on "Surviving the Winter" that reached 50,000 readers.
And here's what I learned: the market doesn't care about ideology. It cares about reliability. It cares about trust. And sometimes, centralized systems are just more reliable.
The Real Risk: What Happens When the Shovels Get Too Expensive
But here's the thing that really worries me about Vertiv's $100 billion valuation. It's not that they're successful. It's that their success is creating a feedback loop that could make AI infrastructure even more concentrated.
When Vertiv's stock price goes up, they can raise more capital. When they raise more capital, they can invest more in R&D. When they invest more in R&D, they build better products. When they build better products, they win more contracts. When they win more contracts, their stock price goes up.
It's a virtuous cycle for Vertiv. But it's a vicious cycle for anyone trying to compete with them.
I saw this same dynamic play out in the crypto mining industry. In the early days, anyone with a GPU could mine Ethereum. Then ASICs came along, and the industrial miners took over. Now, mining is dominated by a few large players with access to cheap electricity and scale. The decentralization that Satoshi envisioned became a centralized industrial operation.
The same thing is happening with AI infrastructure. The "democratization of AI" is becoming the "industrialization of AI," and the people who control the physical layer will control the entire stack.
I'm not saying this to be doom-and-gloom. I'm saying it because we need to be clear-eyed about what's actually happening.
The Opportunity: Where Decentralization Still Matters
Now, let me offer some hope, because I'm not a pessimist. I'm a realist with a vision.
There are still areas where decentralization can and should play a role in the AI ecosystem. I've been working on this since 2024, when I launched TruthLayer, a platform that verifies AI-generated content using blockchain timestamps. We secured $1 million in seed funding by demonstrating how blockchain could combat deepfakes.
The key insight was this: while the infrastructure layer is becoming more centralized, the verification and trust layer can still be decentralized. Blockchain can provide immutable records of what was created, when it was created, and by whom. This is critical in an age where AI-generated content is becoming indistinguishable from human-created content.
Think about it this way. The physical infrastructure of the internet is highly centralized — a few companies control the undersea cables, the data centers, and the backbone networks. But the content layer is decentralized — anyone can create a website, a blog, or a social media account. The same pattern can apply to AI.
The infrastructure will be centralized. The applications can be decentralized. The verification layer must be decentralized.
This is where crypto has a real role to play. Not in trying to compete with Vertiv on cooling systems, but in building the trust infrastructure that AI desperately needs.
The Takeaway: What This Means for You
So what should you take away from Vertiv's $100 billion milestone?
First, understand that the AI boom is real, and it's creating massive value in unexpected places. The "picks and shovels" companies are often better investments than the miners themselves. This is true in crypto, and it's true in AI.
Second, be skeptical of narratives that claim decentralization will solve everything. The physical world has constraints that code can't overcome. Sometimes, centralization is the price of progress.
Third, look for the intersection of AI and blockchain where decentralization actually adds value. Verification, provenance, identity, and governance are all areas where distributed systems can make a real difference.
I've been in this industry for 28 years, and I've seen countless cycles. I've watched projects rise and fall. I've seen fortunes made and lost. Through it all, I've maintained one belief: the technology is real, but the narratives are often exaggerated.
Vertiv's rise to $100 billion is a reminder that the AI revolution is happening whether we're ready for it or not. The question isn't whether we'll have AI infrastructure. The question is who will control it, and whether the values of decentralization can survive the physical realities of scale.
I don't have all the answers. But I know that the conversation we need to be having isn't about market caps or stock prices. It's about what kind of digital world we're building, and whether it will be a world of digital manors where a few lords control the infrastructure, or a world where the infrastructure serves everyone equally.
Democracy isn't a transaction where every voice holds weight. It's a system that requires constant vigilance and participation. The same is true for decentralization. It's not a feature you can install. It's a practice you have to maintain.
And right now, we're not maintaining it very well.
The question I'm asking myself — and the question I want you to consider — is this: if the physical infrastructure of AI is inevitably centralized, what are we going to do to ensure that the benefits of AI are still distributed? How do we build a system where the value created by AI doesn't just flow to the people who own the data centers, but to everyone who contributes to the ecosystem?
That's the challenge of our generation. And it's a challenge that blockchain technology — if we use it wisely — can help us meet.
But only if we stop pretending that decentralization is a magic wand and start treating it as a design principle that requires intentional effort.
I'm going to keep building. I'm going to keep teaching. And I'm going to keep asking the hard questions. I hope you'll join me.
Because the future isn't something that happens to us. It's something we build. And right now, we're building it with air conditioners and cooling systems. The question is whether we can build something better on top of that foundation.