Wall Street's 7% AI Reset: Why Cramer Sees Salesforce and Walmart as the Safer Bet

AI infrastructure pullbacks have turned a market-broadening story into a debate over timing

Jim Cramer says Wall Street is rotating out of some of the market's biggest AI infrastructure winners and into companies with growth drivers outside the data center buildout. He was blunt about how that shift can be read: "You can call it a broadening," he said, "or you can call it fleeing".

The visible change is in AI-linked memory and buildout names. After peaking in June, stocks tied to AI server shortages have tumbled as investors bet those shortages - and the extraordinary profits they created - won't last forever. Cramer's view is that investors have lived through this pattern before: "The stocks just anticipate the ending".

That leaves the market at an awkward junction. Capital is moving toward companies with growth drivers outside the data center buildout, but it is still unclear whether investors are simply diversifying into newer winners or cutting losses in the most crowded AI trade too early.

Salesforce and Walmart look easier to judge than the latest AI winners

Cramer highlighted Costco, Walmart, ServiceNow, Salesforce, and Johnson & Johnson as stocks benefiting from that reshuffling. Salesforce and Walmart stand out here because the cases for them feel more tangible than the high-flying AI narrative.

Salesforce: product use and buybacks are doing some of the talking

Salesforce may not be the flashiest AI trade, but its core tools still matter in day-to-day business operations. Cramer said Marc Benioff has made progress building an agentic AI product that might be the best in the category and pointed to Slack as a widely used system. That matters because the software is already embedded in workflows.

The stock also got hit for reasons that were more about timing than collapse. Cramer said he liked the quarter, but noted that Salesforce's growth isn't supposed to reaccelerate until the second half of the year. After the company reported strong numbers with a softer near-term forecast, some shareholders sold.

Another part of the case is capital management. Cramer said Salesforce is buying back massive amounts of stock, 25 billion already, another 25 billion to go and called that Real sign of conviction, speaks louder than words to me.

Walmart: weak sentiment can make a durable business more interesting

Walmart is a different kind of setup, but it still fits the same idea: buy businesses whose demand is easier to see. Cramer recently explained what Wall Street analysts are getting wrong about Walmart. The comment section leaned on the usual Walmart skepticism, but that noise can be a clue. When everyone hates a store people still shop at, the stock sometimes gets more from sentiment than from fundamentals.

How to think about the rotation without treating TV calls as gospel

This is best treated as a practical market framework, not as personality-driven investing.

Salesforce and Walmart belong in one lane; AI memory buildout names are in another

The group rotating into companies with growth drivers outside the data center buildout includes Costco, Walmart, ServiceNow, Salesforce, and Johnson & Johnson. The appeal is straightforward: demand is easier to observe in software subscriptions, enterprise workflows, and grocery and general-merchandise traffic than it is in a story driven by future spending and shortage economics.

By contrast, the AI memory and buildout group is now a proof-first trade. After shortages gave those companies unprecedented pricing power, investors want evidence that those economics can hold. If they do, the selloff may look like a reset. If not, the crowd has simply moved on.

What to watch from here

• Watch whether AI memory stocks stabilize and whether pricing power looks durable again. If it does, the bearish rotation thesis weakens quickly.
• Watch whether Salesforce can turn product usage and buybacks into the reacceleration expected in the second half of the year.
• Watch whether Walmart keeps looking like a sentiment problem rather than a demand problem.

Both the broadening case and the fleeing case can have merit in the short run. The practical approach is to favor businesses where demand is visible enough to defend the story even if the market re-rates again.