Is xAI a neocloud now?

You know how sometimes you order a pizza and realize the real money isn’t in the pizza, it’s in owning the ovens and renting them out? That might be exactly what’s happening with Elon Musk’s AI company xAI. While everyone’s focused on their chatbot Grok and whether it can compete with ChatGPT, the quieter story is that xAI appears to be building out massive data centers and potentially renting that computing power to other companies. The AI model is almost like the storefront. The data center is the actual business.

This kind of company is called a “neocloud” - basically a newer, smaller version of Amazon Web Services or Google Cloud, but focused specifically on renting out the specialized computer chips that AI systems need to run. Think of it like a gym that started by offering fitness classes but realized the real profit was in renting out the equipment to other gyms. Building and renting computing infrastructure is a steadier, more predictable business than constantly racing to have the smartest AI model. It’s less flashy, but the margins can be very solid.

So why should you care? Because understanding where the real money flows in AI helps you make smarter decisions, whether you’re investing, building something, or just trying to save on your own AI costs.

Here’s how this can actually work for you:

First, keep an eye on neocloud pricing as competition heats up. When more companies fight to rent out AI computing power, prices for smaller customers drop. If you’re a small business currently paying for AI tools, check back every few months on alternatives. More competition almost always means better deals trickle down.

Second, if you have any savings you’re willing to put into individual stocks, the picks-and-shovels approach to AI investing suggests looking at infrastructure companies rather than the flashiest AI apps. The companies building the data centers and cooling systems and power infrastructure tend to have more durable businesses than whoever has the hottest chatbot this month.

Third, if you’re a freelancer or consultant, learning even the basics of how cloud computing and AI infrastructure work makes you genuinely more valuable. You don’t need to become an engineer. Understanding enough to help a small business choose between AI service providers is a real skill gap right now, and people will pay for that guidance.

The company building the tools often makes more money than the company using them, so watch where the infrastructure is going.