Micron and BigBear Stocks Diverge After Results. What It Says About the AI Trade.

Large cloud-computing companies are rushing to build data centers to support artificial-intelligence technology. (Courtesy Dell Technologies)

The artificial-intelligence trade, once an elemental force that lifted stock prices almost indiscriminately, appears to be transforming itself.

Software stocks are no longer nearly certain to rise— BigBear.ai is the latest one to stumble after earnings. But Micron Technology is a reminder that the hardware side of the AI boom is still robust.

BigBear followed the example of its peer C3.ai by posting disappointing revenue figures that sent the shares plunging, a reminder that picking winning AI software stocks is difficult. For every SoundHound celebrating adoption of its technology, there are other companies facing a difficult sales environment.

Investors upbeat about software might counter that BigBear and C3.ai are small and volatile stocks from companies that aren’t leaders in delivering innovative AI products. Earnings season showed plenty of companies in the sector delivering impressive results: Palantir Technologies , Datadog, and Duolingo are just a few of the bigger names saying the integration of AI was leading people to turn to their products.

Still, Duolingo is an example of the worries hitting the sector. While the stock gained nearly 30% after the language-learning company reported its earnings last week, the share price is now lower than it was before that. That appears to be partly due to ChatGPT developer OpenAI choosing to show off its newly launched GPT-5 AI model’s abilities by having it create a language-learning tool based on a short prompt.

“AI is making it clear that almost anyone from an able-bodied start-up, to a big cloud…can create an application so great—that it can compete quickly and potently,” wrote Melius Research analyst Ben Reitzes in a research note. “The advantages of an asset-light model, sustained by ‘predictable’ subscriptions no longer look unique in an AI world where coding and agents are complete commodities and launching a disruptive application takes very little time with very few people.”

By contrast, hardware stocks exposed to the AI infrastructure buildout are largely going from strength to strength. Micron is the latest example. The memory-chip maker raised its financial guidance, reporting booming demand for high-bandwidth memory to be incorporated in Nvidia’s AI processors.

It seems unlikely there will be much letup in demand for hardware. OpenAI CEO Sam Altman said in a social-media post on Tuesday that the company intends to roughly double its computing capacity over the next five months.

“Large-cap tech earnings…have confirmed strong cloud revenue growth and yet another rise in AI capital expenditure. We now forecast global AI capex will rise to $500bn in 2026,” wrote Mark Haefele, chief investment officer at UBS Global Wealth Management, in a research note on Tuesday. The firm’s previous call was for $480 billion.

Write to Adam Clark at adam.clark@barrons.com