Two earnings reports, four days apart, tell you almost everything about how Wall Street is scoring the AI trade right now. Shopify beat estimates and its stock jumped 18% in a single session. AMD also beat estimates — a record quarter, revenue up 50% — and its stock fell 7% the same week. Same broad theme in both reports. Opposite market reaction. That’s not noise. It’s a pattern, and it’s worth understanding before you look at any AI-adjacent stock in your own portfolio.

01 — The two reportsWhat actually happened

Shopify’s second-quarter numbers were strong on their own — revenue up 34% to $3.58 billion, gross merchandise volume climbing 30% to $115.6 billion, a sixth straight quarter above 30% growth. But the number that moved the stock wasn’t the top line. Management’s presentation leaned hard on what it called “agentic commerce”: AI tools, including its Sidekick assistant, that the company said were driving measurably more orders and faster customer acquisition, with next-quarter guidance raised above what analysts expected. In plain terms, Shopify showed investors a direct line from AI spending to more revenue, not just more cost.

AMD’s quarter, on paper, was arguably just as good. Revenue hit $11.5 billion, up roughly 50%, with earnings per share of $1.66 beating the $1.60 consensus. But on the same earnings call, SpaceX’s Elon Musk told analysts that SpaceX would build its AI computing infrastructure exclusively on Nvidia going forward — a reversal from his earlier comments that SpaceX and Tesla would keep buying from both companies. AMD’s own numbers barely mattered after that. The stock dropped as much as 10% intraday before settling down around 7%, while Nvidia, the beneficiary of the exclusivity, rose about 4%.

02 — The patternMonetization vs. spending promises

Zoom out and this isn’t really a Shopify-versus-AMD story. It’s the clearest example yet of a distinction Wall Street has been drawing all earnings season: companies that can show AI turning into revenue today are getting rewarded, sometimes generously. Companies whose AI story is still mostly about capital expenditure — building data centers, buying chips, promising future capacity — are getting scrutinized much harder, even when the underlying numbers are good.

Sorting this earnings season’s AI stories

Worth being precise here: “spending” isn’t automatically a red flag, and “monetization” isn’t automatically a green light. AMD’s underlying business is still growing fast, and infrastructure buildouts like SpaceX’s are a normal, even necessary, phase before monetization shows up. The point isn’t that one category is right and the other wrong — it’s that the market’s patience for the “trust us, it’ll pay off eventually” story is visibly shorter than it was a year ago, and that’s changing which earnings reports move stocks and by how much.

03 — What this means for your portfolioBefore you chase either headline

• 1
  Don’t extrapolate one earnings reaction into a ruleA single quarter’s stock move tells you what the market cared about that day, not a durable law about AI monetization versus AI spending.
• 2
  Separate the fundamentals from the headlineAMD’s actual results were strong. If you own it, the quarter itself isn’t the reason it fell — the lost customer narrative is. Those call for different levels of concern.
• 3
  Watch this week’s earnings through the same lensSuper Micro Computer and Cardinal Health report Tuesday, Applied Materials Thursday. Ask the same question of each: is the AI story about revenue already showing up, or capacity being promised?
• 4
  Check concentration, not just convictionIf your portfolio leans toward AI infrastructure names specifically, this week is a reminder that a single customer comment can move a stock more than the quarter’s actual numbers.

04 — Using AI tools to screen for thisAsk your platform the sorting question directly

This is a genuinely useful screen to run through an AI research or portfolio tool, if yours supports natural-language queries against your holdings or a watchlist. Rather than asking something vague like “is AI a bubble,” ask it to actually sort:

A tool that can pull recent earnings language and actually categorize your specific holdings that way is doing something a keyword stock screener can’t. One that comes back with a generic explainer on AI investing instead is, again, a useful signal about how much is really happening under the hood.

The honest takeThis divide will probably keep sharpening

If there’s a single takeaway from this week, it’s that “AI exposure” stopped being enough on its own to move a stock in one direction. The market increasingly wants to know which side of the ledger a company’s AI story sits on — money coming in, or money going out with a promise attached. That’s a healthier dynamic than blanket enthusiasm, even though it makes individual earnings reports more volatile and harder to predict in advance. Expect more sessions like this one before the market settles into a stable view of who the real AI winners are.