In a choppy, bifurcated market, the biggest AI winners may not be the storytellers but the companies building the backbone
Back in the California Gold Rush, most men who chased the glitter came home with blisters, debts, and a handful of “I almost struck it rich” stories. On the other hand, the men who sold them picks, shovels, wagons, and whiskey did far better for themselves.
Today’s investors are making the same old mistake as the glitter chasers.
Only now, they are chasing the glitter of artificial intelligence — the demos, the headlines, the glamorous software stories — while missing the businesses supplying the machinery that makes the whole boom possible.
In a market that has turned flat, violent, and unforgiving, that mistake could prove costly.
The evidence is already there. The broad market has stopped behaving like the easy bull run of the past decade.
War headlines, oil shocks, private-market stress, and growing fears around AI-driven disruption have created a far more selective environment.
Meanwhile, some of the strongest signals in tech are not coming from the flashiest AI apps, but from the companies building the semiconductors, networking gear, memory systems, and data-center capacity behind the scenes.
Recent updates from names like Broadcom (AVGO) and Marvell (MRVL) suggest demand in that layer remains robust.
That is the part many investors still do not fully appreciate: in a messy market, the winners are rarely the most obvious crowd favorites. Often, they are the suppliers.
And that raises the more important question: if 2026 is shaping up to be a choppy, bifurcated market instead of a broad melt-up, should investors stop chasing AI’s storytellers… and start buying its backbone?
In this week’s episode of Being Exponential With Luke Lango, we will look at why AI infrastructure stocks may be the most durable opportunity in today’s market, why software could remain under pressure, and why the next phase of the AI trade may reward discipline far more than excitement:
The Argument for AI Infrastructure Stocks
One of the more compelling ideas from the latest Being Exponential podcast is that we are no longer in the kind of market investors grew used to over the last decade.
This is not a clean, broad rally where nearly everything rises together. Nor is it a straightforward bear market where the safest move is simply to duck and cover.
Instead, it looks more like a flat, selective, highly volatile tape… the kind of market that punishes lazy dip-buying and rewards precision.
And in that environment, AI infrastructure stocks may have the strongest hand.
The core argument is simple.
Even with war fears rattling markets, even with oil spikes creating macro stress, and even with broader concerns building around private equity, labor weakness, and software disruption, the AI buildout itself still appears very real.
That matters.
Because when investors strip away the noise, capital flows back toward the parts of the market where demand is visible, earnings are tangible, and spending remains urgent.
That is exactly the case for AI infrastructure.
These are the companies tied to the physical and technical backbone of the AI economy: semiconductors, memory, networking, data-center buildout, and other hardware-heavy enablers.
Unlike the more speculative corners of AI, these businesses are not selling a distant dream. Many are selling the actual equipment required to keep the boom going.
Our podcast points to recent strength from companies like Broadcom and Marvell, whose data-center businesses have continued to post powerful growth and improving outlooks.
That is not a trivial signal. If the companies closest to the infrastructure layer are beating expectations and lifting long-term targets, it suggests enterprise demand has not broken.
That also helps explain the growing divide between AI infrastructure and software.
The SaaSmageddon Is Here
Earlier in the year, hardware and infrastructure names were leading the market. Then wartime volatility briefly reversed the trade, with software catching a defensive bounce while semiconductors and industrial names slipped.
But as fears of a larger economic shock began to ease, the old pattern reportedly returned: hardware started leading again, while software rolled back over.
Why? Because software faces a more uncomfortable question. AI may not just help software companies… it may disrupt parts of the software business model itself.
If that pressure continues, investors may keep favoring the companies building AI’s foundation over the companies trying to defend old margins against it.
That distinction could define this market.
The big takeaway is not that investors should blindly buy every chip stock in sight. No. It seems 2026 may require a different playbook.
In a flat market, leadership narrows.
Stock selection matters more.
Timing matters more.
The winners can still do extremely well, even if the major indices go nowhere.
That is why AI infrastructure stands out. It may be one of the few areas where fundamentals, momentum, and long-term narrative still line up.
The Bottom Line
Of course, the risk has not disappeared. The podcast makes that clear. If a deeper economic shock emerges — especially from something like a private equity or credit event — even the AI infrastructure trade could get hit.
After all, the AI spending boom is ultimately funded by real economic activity. If consumers and businesses pull back hard, the money feeding that buildout slows down too.
But that appears to be a later risk, not the dominant one right now.
For now, the market seems to be sending a more immediate message: this is no longer the age of buying everything. It is the age of buying what is still working.
And if that is true, the modern equivalent of picks and shovels may not be a romantic, old metaphor at all. It may be the most practical investment lesson of the year.
To see the bigger picture — including the risks around war, private equity, gold, and why this market may stay choppy for months — the full podcast is well worth your time.
