AI Is Growing Fast. But What's Really Driving That Growth?

A few weeks ago, I wrote about Nvidia's investments in companies like xAI and the SPV structures helping fund AI infrastructure.

The feedback was interesting. Most people weren't surprised that Nvidia was investing in the ecosystem. What caught their attention was something else.

If the companies building AI are also helping finance its growth, what does that mean for the numbers we're all watching?

After reading Groundbrkr's article, The Second Derivative, it reinforced something that's been coming up more often in conversations with clients. As finance leaders, we spend a lot of time measuring growth. Maybe we should spend a little more time understanding what's driving it.

Growth tells us what happened. Capital explains why.

Nobody questions that AI demand is real. The largest cloud providers are investing at levels we've never seen before. Nvidia continues to report remarkable results. Capital keeps flowing into AI companies at an incredible pace. Those are facts.

The more interesting question is what's sitting underneath those numbers.

How much demand is coming from customers?

How much is being accelerated by abundant capital?

Those are very different drivers, and they lead to very different conversations.

Revenue tells us what happened. Following the money helps explain why it happened.

Following the capital

One reason I wrote about Nvidia's SPV investments is that they highlight something many people overlook. The companies building AI are no longer just selling into the ecosystem. In some cases, they're also helping fund it.

That creates a powerful cycle.

• Investors fund AI companies.
• AI companies invest heavily in infrastructure.
• Infrastructure providers invest back into AI companies and adjacent businesses.
• More capital enters the ecosystem.
• Investment accelerates again.

There's nothing inherently wrong with this. In fact, many transformational industries have relied on creative financing to grow. The question isn't whether this model should exist.

The question is whether customer demand eventually becomes strong enough to sustain it without the same level of capital support.

A client conversation that stuck with me. Not long after publishing the Nvidia article, we were speaking with a client about an AI initiative they were considering. The discussion wasn't about whether AI could create value. Everyone in the room agreed it probably could. The conversation quickly shifted to something much more familiar.

How much capital would this require?

When would they see a measurable return?

Would this improve the economics of the business, or were they feeling pressure to invest because everyone else seemed to be moving?

Those are the conversations finance leaders should be having. The challenge wasn't evaluating AI.

The challenge was separating genuine business demand from market excitement.

That's exactly what brought me back to both the Nvidia SPV discussion and Groundbrkr's article.

Revenue isn't the whole story

Another metric getting a lot of attention is Remaining Performance Obligations, or RPO.

CJ Gustafson has written extensively about this, and I think his perspective is an important one. RPO tells us what customers have committed to buy. It does not tell us whether those investments will generate attractive returns. That's an important distinction.

As accountants and finance professionals, we've always known that booked revenue and economic value aren't the same thing. That hasn't changed because we're talking about AI.

A signed contract tells you someone committed to spend the money. It doesn't tell you whether spending it was the right decision.

The questions we're asking

When we're advising clients, we spend very little time debating whether AI is real. It clearly is.

Instead, we're asking questions like:

• Is this growth being funded by customers or investors?
• Is cash flow keeping pace with reported performance?
• Will these economics still make sense if capital becomes more expensive?
• Are we investing because the expected return is compelling, or because we don't want to be left behind?

Those questions apply whether you're evaluating an AI investment, acquiring another business, or launching a new product. The framework is the same.

What We're Watching

We don't know exactly how the AI investment cycle will play out. Neither does anyone else.

What we do know is that the strongest businesses eventually separate themselves from market excitement through solid fundamentals. Growing revenue, healthy margins, strong cash flow, and disciplined capital allocation. 

Those are the metrics that endure long after the headlines move on.

That's why articles like The Second Derivative resonate with us. They encourage finance leaders to look beyond the surface and ask better questions about what the numbers actually represent.

For us, that's the real takeaway. AI will almost certainly reshape how businesses operate.

The companies that create lasting value won't necessarily be the ones spending the most or growing the fastest. They'll be the ones that turn investment into durable economic returns.

As advisors, that's what we'll continue to watch, and it's where we'll continue to help our clients focus.