Meta CEO Mark Zuckerberg used the company’s Q4 2025 earnings call to set expectations for a gradual, rather than revolutionary, approach to AI development. Rather than promising an immediate breakthrough model, Zuckerberg emphasized Meta’s long-term trajectory in artificial intelligence, stating that “the first models will be good, but more importantly, we’ll show the rapid trajectory that we’re on.”
In June 2025, Meta launched Meta Superintelligence Labs (MSL), headed by former Scale AI CEO Alexandr Wang, signaling a major AI reset for the company. Six months into this initiative, Zuckerberg acknowledged he couldn’t share detailed strategy information but expressed satisfaction with progress. He indicated that Meta would roll out its initial set of models and products over the coming months, with plans to “steadily push the frontier over the course of the year.”
Analyst reactions were mixed. Barclays analyst Ross Sandler noted “underlying confidence and clearly a lot of new things in the works,” while Brian Mulberry from Zacks Investment Management said there was “no real substance” to Zuckerberg’s comments. When JPMorgan analyst Doug Anmuth pressed for more details on MSL’s progress, Zuckerberg declined to elaborate beyond praising the team’s quality.
Meta’s strong advertising business provides financial cushion for its AI ambitions. The company reported 24% revenue growth to $59.9 billion in Q4 2025, with $43.6 billion in free cash flow for the full year despite heavy AI infrastructure spending. AI improvements to ad ranking and display drove 3.5% more clicks on Facebook and over 1% conversion gains on Instagram. Ad impressions jumped 18% year-over-year, while advertiser costs per ad rose 6%.
However, Meta faces challenges. The company previously promised to release the next Llama AI model by end of 2025 but missed that deadline. The previous Llama release in April disappointed developers with weak coding and reasoning capabilities. According to The Wall Street Journal, Meta is developing new models codenamed “Mango” (focused on images and videos) and “Avocado” (specialized for coding). Bloomberg Intelligence’s Mandeep Singh expects Meta to specialize in specific AI areas rather than competing across all fronts simultaneously.
Key Quotes
I expect our first models will be good, but more importantly, we’ll show the rapid trajectory that we’re on, and then I expect us to steadily push the frontier over the course of the year as we continue to release new models.
Mark Zuckerberg set expectations during Meta’s Q4 earnings call, emphasizing that investors should focus on Meta’s improvement trajectory rather than expecting an immediate breakthrough AI model.
I think my answers to a lot of your questions on this particular call may be somewhat unfulfilling because we’re in this interesting period where we’ve been rebuilding our AI effort. And we’re six months into that, and I’m happy with how it’s going.
Zuckerberg acknowledged the lack of concrete details about Meta Superintelligence Labs’ progress, asking for patience as the company is only six months into its AI reset under Alexandr Wang’s leadership.
We want to see real bottom-line profits driven by AI, and it seems that Meta is still far from that reality.
Brian Mulberry, analyst at Zacks Investment Management, expressed skepticism about Meta’s AI strategy, highlighting investor concerns about when AI investments will translate into measurable profit improvements.
This kind of growth rate allows you a lot of cushion in terms of taking your time and getting AI.
Mandeep Singh from Bloomberg Intelligence explained how Meta’s strong 24% revenue growth and robust advertising business provide financial runway to pursue patient, long-term AI development without immediate pressure for returns.
Our Take
Meta’s measured approach reveals a strategic calculation: the company recognizes it’s behind in the AI race but believes its advertising cash cow buys time to catch up methodically. This contrasts sharply with the “move fast and break things” ethos that once defined Facebook. The six-month timeline since launching Meta Superintelligence Labs is remarkably short for expecting major breakthroughs, suggesting Zuckerberg’s caution may be realistic rather than evasive.
The specialized model strategy—Mango for visual content, Avocado for coding—indicates Meta is learning from Llama’s shortcomings by focusing on specific capabilities rather than attempting to match GPT-4 or Claude across all dimensions. However, the missed deadlines and vague timelines risk eroding investor confidence if competitors continue advancing rapidly. Meta’s success will ultimately depend on whether incremental progress can compound quickly enough to close the gap with AI leaders before the market loses patience or the advertising business faces headwinds.
Why This Matters
Meta’s cautious AI strategy represents a significant shift in how major tech companies are positioning themselves in the intensifying AI race. While competitors like OpenAI, Google, and Anthropic have focused on headline-grabbing model releases, Zuckerberg is betting on incremental progress and sustained momentum over flashy announcements. This approach carries both risks and opportunities.
The establishment of Meta Superintelligence Labs under Alexandr Wang signals Meta’s commitment to catching up in AI capabilities, particularly after its previous Llama model disappointed on critical benchmarks. The company’s massive financial resources—generating over $43 billion in free cash flow while spending heavily on AI—give it unusual staying power in the expensive AI development race.
For the broader AI industry, Meta’s strategy suggests that sustainable AI development may require patient capital and realistic timelines rather than rushed releases. The company’s success in applying AI to its core advertising business demonstrates that practical AI applications can fund more ambitious research goals. However, Meta’s missed deadlines and the lukewarm reception to previous models raise questions about whether gradual progress will satisfy investors and allow Meta to close the gap with AI leaders. The outcome will influence how other tech giants balance short-term market expectations with long-term AI development goals.
Related Stories
- Meta Q4 Earnings: Zuckerberg Bets Big on AI with $135B Capex Plan
- Big Tech’s 2025 AI Plans: Meta, Apple, Tesla, Google Unveil Roadmap
- Meta and Nvidia Billionaires’ Wealth Soars $152B in AI Boom
- CEOs Express Insecurity About AI Strategy and Implementation
- What Zuckerberg Risks by Following Musk’s Lead
Source: https://www.businessinsider.com/meta-q4-2025-earnings-mark-zuckerberg-ai-gradual-growth-2026-1