Meta just introduced an idea investors haven't seriously considered during the AI boom: that compute scarcity might not last forever. The company signaled it could make excess AI capacity available to outside customers, and Wall Street reacted like it had heard bad news.
Chip stocks got hit hard. Micron dropped over 10%. AMD fell nearly 7%. Even Nvidia, which has felt untouchable for two years, slipped. The trade has been heavily supported by the assumption that compute demand would keep overwhelming supply. Meta just gave that assumption its first real test, and the market flinched.
Here's what makes this confusing. Nobody is actually spending less. AI infrastructure spending is still accelerating, and the biggest technology companies are still committing enormous amounts of capital to it. Alphabet just raised its own guidance again and told investors 2027 will be bigger still. So the money keeps flowing in one direction while the stock reaction points the other way.
The problem was never whether companies are willing to spend. It's whether the returns arrive before investors run out of patience. That's exactly what investors are starting to differentiate between.
You can actually watch that patience get tested in real time if you compare two companies doing the exact same thing. Alphabet's stock jumped 7% after its last earnings call, because cloud revenue came in ahead of what analysts expected. Meta's fell 6% the same season, for spending more. The difference wasn't simply spending more or spending less. It was evidence of monetization, some measurable sign that the investment is converting into results.
That shift isn't unique to those two companies. It's showing up across the sector. JPMorgan noted that correlation between the big AI spenders has actually hit new lows this year. Investors have stopped treating "AI stock" as one trade. They're grading each company on its own proof of work.
Spending was the story for two years.
Conversion is the story now.
Investors used to reward "we're building the future" on faith alone. Now the question is, "Where does the future show up in the numbers?" That's a much higher bar, and it's why the market can look nervous and confident about AI at the same time. It's grading two different things at once.
The technology may be progressing exactly as expected. But markets and businesses operate on very different timelines. Investors measure quarters. Businesses measure years.
The question investors are asking now is no longer how much companies are willing to spend. It's how quickly they can turn that spending into something measurable.