In this episode of The Sunny Ray Show, host Sunny Ray welcomes back investor and venture builder Paul Anthony Claxton for a wide ranging conversation on the infrastructure powering the AI boom. Claxton explains why data centers have become one of the hottest asset classes on earth, arguing that AI companies like OpenAI and Anthropic underestimated how much power and infrastructure consumer demand would require. He describes a shift from centralized mega data centers toward micro and edge facilities built with local communities as stakeholders, backed by federal lobbying support from his team. The two dig into capital deployment discipline, the access disparity that could determine who really controls AI, and how quantum computing will demand even more distributed power. Claxton also warns founders about bad actors and disqualifying events hiding inside cap tables, and how to spot manufactured traction during fundraising. The conversation closes on a provocative note, with Claxton predicting capitalism itself is approaching a breaking point, a supernova driven by AI's relentless expansion and humanity's push toward space.
Paul Anthony Claxton returns to unpack why data centers, power access, and quantum computing will decide who really controls AI's future.
Why are data centers suddenly the hottest asset class on earth?
AI can't operate without data centers, and I don't think OpenAI or Anthropic predicted how much infrastructure the consumer boom would require from the power grid. Right now we have centralized infrastructure, but the market is shifting toward edge and micro data centers in local communities. Communities need to be stakeholders, since these facilities demand huge power. We're not just deploying capital, we're working federal channels and lobbying to build data centers that benefit everyday citizens, not just the AI economy.
Who actually controls the power behind tomorrow's tech?
We do, the people do. It comes down to access. Right now most people can afford something like ChatGPT, but as systems get more sophisticated and expensive, companies will need direct access to compute power and GPUs, and that access gap will widen. We still control it today, but as we become more converged with AI, I think we'll lose our grasp on that control unless we address the affordability and access disparity.
What do most investors get wrong about data center deals?
Most VCs aren't really focused on data centers, we're still caught up in the shift from SaaS to AI wrapper companies. Data centers are complex, with tons of line items across capex, equity, and debt financing, and most capital isn't patient enough for how long they take to build and generate revenue. I look at them differently, as hard assets and collateral that support our startups long term, and it's a small but important part of our thesis.
Where do bad actors hide in this gold rush, and what counts as a disqualifying event?
Bad actors sit right inside our own companies. Anyone raising under 506b or 506c, especially owners with twenty percent or more voting shares, needs to fill out a bad actor questionnaire, and investors and employees need proper KYC and diligence too. I see founders who don't know who's giving them two million dollars. A disqualifying event could be a criminal conviction, outstanding litigation, or even a stake in a competitor, and it can sabotage your whole round months later.
How does quantum computing change the data center game?
Quantum computing requires far more power than what's currently needed, and to get toward general intelligence and true edge computing at scale, we'll need distributed power through micro data centers. The American grid is divided into regions that borrow power from each other, but AI requires so much power that you can't keep shifting it back and forth at scale, the system breaks. We have to deliver power at micro scale, not macro scale.
Is capitalism really headed for a supernova?
Capitalism was always meant to expand, it's built on human evolution, and you can only get so big before it breaks. That's why we're pushing toward Mars, we need room to commercialize and expand. I think AI's evolution will be the trigger, and we're going to see the greatest economic fallout in maybe a thousand years. No one really owns their money in the banking system, so honestly the safest place right now might be fixed annuities.
What is manufactured traction, and how do you spot it?
Manufactured traction is when founders tell investors what they want to hear, like claiming ten million in pipeline off an LOI that doesn't really mean anything. Or they list big name advisors on the cap table who aren't actually doing anything for the company. I'd rather hear all the bad stuff upfront, because we invest in people first and their ability to pivot and execute under pressure, the market opportunity comes second.
Building something daring? Sunny talks to founders like this every day. Fifteen minutes to see if your story belongs on the stage.
Claim your pre-interview