The AI Build‑Out Isn’t Slowing — and NVIDIA Still Sits at the Center of Many Major CapEx Plans

The folks floating around the idea of an “AI bubble” certainly seem smart. While many of them may possess next-level foresight, I wouldn’t look to bet against the revolutionary technology, especially as we begin to get a glimpse of the monetization opportunities.

If you’re not blown away by Anthropic’s Claude Mythos enough to ditch the broad AI bubble belief, perhaps what comes next might be convincing enough to stay invested in the AI trade, as the benefits broaden out across more sectors of the economy. Notably, the big banks look like AI winners that stand to reap real cost savings.

Though a correction or even a bad crash (especially for the parabolic movers) cannot be ruled out, given AI can advance by leaps and bounds, even if share prices on AI stocks don’t, I do think that dismissing all of tech as an AI bubble is to miss the major shifts going on underneath the hood.

Nvidia is still winning big and its latest upswing might have room

As the AI buildout accelerates, the semis continue to be a great place to be. And Nvidia (NASDAQ:NVDA | NVDA Price Prediction) continues to stand out as the “pick and shovels” play to stick with, especially as the firm gets into PC chips.

It’s a sign that there’s a real opportunity in edge AI. And as Nvidia expands its footprint further across the stack, Nvidia stands out as a “safe” play in AI, at least relatively speaking, given its position as an AI enabler that’s more than capable of serving up new surprises and (circular) deals across AI.

In short, Jensen Huang and company don’t seem content with winning in the hardware stack of AI; they probably want to win that entire five-layer cake. And I’d argue they’ve made the moves to bring it home.

As the hyperscalers keep scaling their CapEx, it looks like Nvidia is just going to keep on selling. It’s more of the same, and for the AI bubble callers, a difficult choice needs to be made, especially if investors become conditioned to accept the fact that collective CapEx, especially among the Mag Seven, is going up.

As the massive CapEx goes towards the chokepoints of the AI boom, questions linger as to how to value the many names that have already gained by triple-digit percentage points. Is this AI revolution going to keep rewarding momentum chasing? Or is this going to be like prior bubbles, where booms have led to painful busts?

Nvidia might just be a cheap stock, after all

If the monetization phase falls flat, there’s no question that a name like Nvidia could prove vulnerable, as hyperscalers scale back, AI buildouts stall, and the revolution enters a wintry period. But, at the same time, something like Anthropic’s Claude Code or Claude Mythos shows the unfathomable potential behind models at the frontier.

Uncovering thousands of bugs? Automating an even greater chunk of a firm’s coding? Getting a step closer towards recursive self-improvement? 

These are profound applications that are easy to dismiss. And while the technology might not be perfect, as pointed out by skeptics who will shine the light on its flaws, it’s really hard to tell what lies just around the horizon.

Could massive improvements around the corner lead to an event horizon where profound, disruptive change happens, making it impossible to go back?

I have no idea, but the case for Nvidia not being a bubble, I think, could rise with every game-changing innovation that lands. Even if custom silicon rivals move in, the GPU king is already branching out into other corners of the AI boom that could prove just as lucrative.

Given the odds of an AI bubble and the potential demand that could materialize if it’s not, I still view a name like Nvidia as one that possesses asymmetric upside. And that’s probably why the name’s remained a hedge fund favorite.