Meta reports fourth-quarter earnings Wednesday, and Wall Street's got one question: will the company's billion-dollar AI gamble actually pay off? After spending much of 2025 overhauling its AI division - including a massive $14.3 billion investment in Scale AI to poach founder Alexandr Wang - investors are hungry for proof that CEO Mark Zuckerberg's bet on artificial intelligence is more than just expensive promises. With analysts expecting revenue of $58.59 billion and earnings of $8.23 per share, the real story isn't in the numbers but in what comes next.
Meta is about to show its hand. When the social media giant reports fourth-quarter results Wednesday after market close, investors won't just be scanning for revenue beats or user growth metrics. They'll be hunting for concrete evidence that Mark Zuckerberg's massive AI overhaul - the one that's been burning through billions - is actually starting to work.
The numbers Wall Street's watching are straightforward enough: analysts polled by LSEG expect earnings of $8.23 per share on revenue of $58.59 billion. Online advertising sales, still Meta's core business, should come in around $56.98 billion according to StreetAccount estimates. Daily active users across Meta's family of apps are projected to hit 3.58 billion.
But those familiar metrics mask what's really happening inside Meta. The company spent much of 2025 essentially ripping apart and rebuilding its entire AI operation. The centerpiece? A $14.3 billion investment in Scale AI that brought founder Alexandr Wang and his team into Meta's orbit. Wang now runs TBD, Meta's top-tier AI unit tasked with developing the kind of powerful models that can compete with OpenAI and Google.
The TBD unit exists because Meta needed a reset. When Llama 4 launched last spring to a lukewarm reception from developers, it became clear the company's AI strategy wasn't working. Now Meta's testing a new frontier model internally - code-named Avocado - that's supposed to leapfrog Llama entirely. CNBC reported the company plans to release Avocado during the first half of 2026.
"Being able to make a significantly larger investment here is very likely to be a profitable thing over some period," Zuckerberg told analysts back in October, according to CNBC's earnings coverage. That "some period" qualifier is doing a lot of work, though. Investors have gotten increasingly nervous about Meta's AI spending spree, especially as capital expenditures balloon.
For the fourth quarter alone, StreetAccount projects capital expenditures of $21.97 billion - much of it tied to the data center buildout that's become table stakes in the AI arms race. Meta committed Tuesday to paying Corning up to $6 billion through 2030 for fiber-optic cable to wire up those AI data centers. Alphabet and OpenAI are racing to build similar infrastructure, turning data center capacity into the new competitive bottleneck.
While Meta pours resources into AI, it's quietly pulling back from another expensive bet: the metaverse. Earlier this month, Meta laid off more than 1,000 Reality Labs employees working on VR initiatives. The cuts hit internal game studios particularly hard and sent a chill through the VR developer community. CNBC reported the moves sparked fears of a "VR winter" given Meta's outsized influence on the industry.
Meta tech chief Andrew Bosworth tried to calm those fears last week, insisting the company isn't abandoning VR entirely. But the resource shift is unmistakable - Reality Labs is getting starved while AI gets fed. Analysts expect Reality Labs to post a fourth-quarter operating loss of $5.67 billion on just $940.8 million in sales. Since late 2020, that division has racked up over $70 billion in total operating losses.
The strategic pivot makes sense on paper. Ray-Ban Meta smart glasses - which blend AI capabilities with wearable tech - have generated more buzz than any VR headset Meta's shipped. The glasses represent a more practical path to ambient computing than strapping on a headset to attend virtual meetings nobody asked for.
What investors really want to hear Wednesday night is how Meta's AI investments translate into actual business results. Can the company's AI tools drive higher ad rates? Will Avocado attract the developer ecosystem that Llama 4 failed to capture? How quickly can Meta turn billions in AI spending into revenue?
Zuckerberg's been here before. He bet big on mobile when Facebook's business was still desktop-centric, endured quarters of investor skepticism, and ultimately proved the pivot was necessary. The AI transformation feels similar - a painful, expensive transition that might be essential but requires a leap of faith from shareholders.
The difference this time is the competition. OpenAI has consumer mindshare. Google has search distribution and decades of AI research. Microsoft has enterprise relationships and OpenAI's technology. Meta's starting from behind in a race where the finish line keeps moving.
Fourth-quarter results will offer clues about whether Meta's catching up or falling further back. Zuckerberg's track record buys him some patience, but not infinite amounts. At some point, the AI investments need to show returns beyond vague promises about future profitability.
Meta's fourth-quarter earnings will reveal whether Zuckerberg's expensive AI pivot is visionary or reckless. The company's $14.3 billion bet on Scale AI's team, the upcoming Avocado model launch, and the aggressive data center buildout all point to a company trying to buy its way into the AI elite. But with Reality Labs bleeding billions while getting starved of resources and capital expenditures hitting nearly $22 billion in a single quarter, Meta needs to show progress fast. Investors have tolerated the spending so far based on Zuckerberg's track record - but patience in tech is measured in quarters, not years. Wednesday's results won't settle the debate, but they'll show whether Meta's AI strategy is gaining traction or just gaining costs.
