Meta drops its fourth-quarter earnings Wednesday, and Wall Street's laser-focused on one question: Is the company's radical AI overhaul worth the staggering price tag? After spending $14.3 billion to acquire Scale AI's leadership and another $6 billion commitment for data center fiber optics, CEO Mark Zuckerberg faces skeptical investors who want proof that this massive pivot will deliver returns. The social media giant completely restructured its AI division in 2025 following Llama 4's lukewarm reception, betting everything on a new frontier model code-named Avocado that's set to launch in the coming months.
Meta just put all its chips on AI, and Wednesday's earnings call will show if Mark Zuckerberg's billion-dollar bet is already paying dividends or if investors should brace for a costly waiting game.
The numbers tell part of the story. Analysts polled by LSEG expect revenue of $58.59 billion and earnings per share of $8.23 for the fourth quarter. But those figures are table stakes compared to what Wall Street really wants to hear: concrete evidence that Meta's dramatic AI restructuring justifies the eye-watering capital expenditures that have sent some investors running for the exits.
Meta spent 2025 completely overhauling its artificial intelligence operations after Llama 4 landed with a thud last spring. Developers gave the model a tepid reception, forcing Zuckerberg to confront an uncomfortable reality - Meta was losing ground to OpenAI and Google in the AI arms race that's reshaping Big Tech.
The company's response was characteristically aggressive. Meta invested $14.3 billion in Scale AI primarily to acquire the startup's founder, Alexandr Wang, and his top talent. Wang now runs Meta's elite TBD unit, the internal group tasked with developing frontier AI models that can compete with GPT-5 and Google's latest offerings.
Inside TBD, engineers are racing to complete Avocado, Meta's next-generation model and Llama successor that's scheduled to launch sometime in the first half of 2026. CNBC reported in December that internal testing has been ongoing, though details about Avocado's capabilities remain tightly guarded.
But building cutting-edge AI models requires massive infrastructure, and Meta's been throwing money at data centers like there's no tomorrow. Capital expenditures for the quarter are projected to hit $21.97 billion according to StreetAccount, driven almost entirely by the company's data center expansion. On Tuesday, Meta announced it's committing up to $6 billion to Corning through 2030 for fiber-optic cable to connect its AI infrastructure.
The spending has made some investors nervous, but Zuckerberg's been unapologetic. "Being able to make a significantly larger investment here is very likely to be a profitable thing over some period," he told analysts in October, characterizing the massive outlays as necessary given AI's breakneck development pace.
Meta's advertising business - still the company's financial engine - is expected to generate $56.98 billion in fourth-quarter sales according to StreetAccount estimates. Daily active people across Meta's family of apps should come in around 3.58 billion. Those are healthy numbers, but the real test is whether Zuckerberg can articulate how AI investments will boost that core business in 2026 and beyond.
Then there's the elephant in the room: Reality Labs. Meta's metaverse division just laid off more than 1,000 VR employees earlier this month, including entire internal studios working on virtual reality experiences. The cuts underscore how dramatically Meta's priorities have shifted toward AI and related hardware like Ray-Ban Meta smart glasses.
Analysts estimate Reality Labs will post a $5.67 billion operating loss this quarter on just $940.8 million in sales. The division has now racked up over $70 billion in total operating losses since late 2020, when Zuckerberg first started pouring resources into his metaverse vision.
Meta's tech chief Andrew Bosworth insisted last week that the company isn't abandoning VR, but the layoffs have sparked what CNBC characterized as fears of a "VR winter" among developers who depend on Meta's outsized influence in that industry.
The tension between Meta's AI ambitions and its metaverse legacy will likely dominate Wednesday's analyst call. Investors want Zuckerberg to pick a lane, or at least explain how both initiatives fit into a coherent strategy that justifies the combined capital drain.
What happens next depends largely on Avocado's performance when it launches. If Meta's new model can compete with the best from OpenAI and Google, the company's spending spree will look prescient. If it underwhelms like Llama 4, Zuckerberg will face much harder questions about whether throwing money at the problem is actually working.
Wednesday's earnings call is less about Meta's fourth-quarter numbers and more about whether Zuckerberg can convince Wall Street that his company's radical transformation will pay off. With $14.3 billion committed to Scale AI talent, nearly $22 billion in quarterly capex, and a new frontier model launching within months, Meta's placed the biggest bet in its history on artificial intelligence. Investors will be listening for concrete timelines on Avocado's launch, specific use cases showing AI's impact on advertising revenue, and honest assessments of how Reality Labs fits into this new AI-first future. The market's patience with Meta's spending isn't infinite, and Zuckerberg needs to show results soon.
