This year, investing 1 trillion yuan in AI, Meta going all-in, Microsoft being pressed to the ground and rubbed.

Meta shares surged over 10% after hours, while Microsoft fell 8%.

Author: Facing AI

On the same evening, two earnings reports, two different sentiments.

Once Meta’s earnings were released, the stock price soared throughout after-hours trading.

Almost simultaneously, Microsoft was declining.

The numbers on the books aren’t that far apart; what truly widens the gap is the market’s attitude towards the “future.”

01 Mark Zuckerberg really has everyone hooked

Meta’s latest earnings report is truly impressive.

For Q4 FY2025, Meta:

· Revenue of $59.893 billion, up 24% year-over-year, significantly exceeding Wall Street expectations.

· Net profit of $22.768 billion, up 6% year-over-year.

· Diluted earnings per share of $8.88, up 11% year-over-year.

The majority of revenue still comes from advertising, with Q4 ad revenue at $58.137 billion, accounting for 97% of total revenue. Other business revenues are smaller but grew 54% year-over-year.

From an operational perspective, the daily active users (DAU) across Meta’s family of apps averaged 3.58 billion, up 7% year-over-year.

In Q4, ad impressions across the app family increased by 18% year-over-year, and the average price per ad increased by 6% year-over-year; for 2025, these two metrics are projected to grow by 12% and 9% respectively for the full year.

Meta has always attributed positive changes in its ad business to AI, as AI helps boost both ad volume and prices.

For the full year of 2025, Meta’s total revenue reached $200.966 billion, a 22% increase from $164.501 billion in 2024; net profit was $60.458 billion, down 3% from $62.360 billion in 2024.

However, since the AI industry took off, the market reaction to the earnings reports of the big tech giants has never been solely based on performance; investors are more concerned with: what’s next?

Over the past two years, the persistent skepticism surrounding Meta has been its exaggerated spending. Zuckerberg is all-in for AI.

If you only look at the numbers, nothing has changed.

Meta has further increased its expenditure forecast, with capital spending in 2026 expected to be between $115 billion and $135 billion.

In the past six months, Meta has actively restructured its AI business architecture, establishing a super-intelligent laboratory.

During the earnings call, Zuckerberg stated that the company plans to release its latest AI models in the coming months.

“We will showcase our current rapid development momentum,” he said, adding that Meta hopes to “expand into frontier areas” through its AI R&D efforts.

The scene Zuckerberg presented was: everything is ready, and we are about to deliver big.

The market chose to believe—or rather, to bet—that Meta can really turn things around this time.

During the earnings call, Meta’s stock price rose over 10% in after-hours trading.

02 Is Microsoft no longer aggressive?

Contrasting sharply with Meta is Microsoft.

All eyes on Microsoft are focused on its Azure cloud computing business, which is experiencing strong demand from enterprises developing and deploying AI services.

In the first fiscal quarter ending in September, Microsoft said demand for Azure services “significantly” exceeded its capacity. The company expects revenue growth in this segment to accelerate in the second quarter.

In fact, the latest second quarter results show Azure cloud revenue grew by 38%, slightly slowing from the previous quarter. Microsoft’s overall revenue growth also slowed, from 18% in the previous quarter to 17%.

Investors are also closely watching the growth signals of Microsoft’s Copilot brand products, which are the main channel through which Microsoft sells AI software tools to office workers.

With Anthropic releasing a new AI tool, Claude Cowork, earlier this month to positive reviews, shareholders are increasingly worried that Microsoft’s related businesses might be “snatched away.”

In after-hours trading, Microsoft’s stock price once fell over 8%.

Microsoft was among the earliest to “place bets” successfully in this AI wave, making a high-profile entry that stirred the industry. It heavily invested in OpenAI and is closely tied to it. In July last year, Microsoft’s market cap briefly surpassed $4 trillion.

But early entry also comes with its frustrations—early entry means early validation. While Meta was still experimenting with different paths, investors were already eager, and Microsoft’s confident, steady, and comprehensive AI investments are yet to show a proportional contribution to revenue.

When this contribution is delayed or insufficient, market patience is also being tested.

Regarding the slowdown in Azure growth, Microsoft defended itself during the earnings call.

CFO Amy Hood said, “If I had allocated all the GPUs launched in Q1 and Q2 to Azure, our KPI (growth rate) would have already exceeded 40%.”

Hood explained that the issue isn’t that Azure cloud services are selling poorly; quite the opposite, demand is too strong, and supply is insufficient. Microsoft’s computing power must serve not only Azure but also AI products like Microsoft 365 Copilot and GitHub Copilot, making allocation challenging.

She even revealed that a large portion of Microsoft’s current expenditure is on GPUs/CPUs, highlighting how tight the current computing resources are.

Additionally, Microsoft CEO Nadella directly rebutted rumors during the earnings call, which speculated that Microsoft’s AI tools’ usage had declined due to competition.

Nadella disclosed that Microsoft 365 Copilot’s daily active users increased tenfold, with paid users growing 160% year-over-year, reaching 15 million.

03 The future, the future, still the future

Although ChatGPT is about to turn four this year, major tech companies generally emphasize that we are still in the “early stages” of AI industry development.

Regarding the future, the giants are broadly optimistic.

Zuckerberg has explicitly regarded AI smart glasses as the next-generation core computing device, comparing this turning point to the moment when smartphones replaced feature phones.

Next, Meta’s key leap is to gradually reconstruct its ad-centric business model into a new revenue system centered around “personal superintelligence.”

Microsoft’s vision of the future, however, is more “engineering” and “system”-oriented.

In Nadella’s view, AI is not just a single blockbuster but an entire set of capabilities embedded into operating systems, office software, development tools, and cloud infrastructure upgrades. Copilot doesn’t need to prove how much money it can make all at once; as long as it can continuously increase the stickiness and ARPU of Microsoft 365, GitHub, Azure, and other products, commercialization will naturally follow.

The problem is, the market’s patience for the “future” is not evenly distributed.

Meta is still in a stage where “burning money is acceptable”: its advertising foundation is solid, cash flow is abundant, and AI is more like a gamble on the next entry point.

Microsoft, on the other hand, is already in a position where “it must deliver results”—it is the earliest to bet, the deepest in investment, and the most complete in narrative. Naturally, it is also the first to be asked to provide measurable returns.

Thus, the same phrase “we are still in the early stages” means very different things for these two companies.

Ultimately, the question isn’t whether AI is good or not, but whose timeline it favors.

Meta still has room for storytelling, while Microsoft has already reached the point where it must turn its story into numbers.