Meta’s pivot to AI closes the chapter on the metaverse experiment
What was meant to define the next internet instead became a costly detour.
Nearly half a decade ago, Mark Zuckerberg tried to invent the future. But the future, it turns out, had its own plans. Last week, Zuckerberg was forced to acknowledge that reality when Meta announced it was laying off 10% of its workforce in the metaverse division, about 1,500 employees, and shifting its focus toward AI-powered hardware products. “We’re shifting some of our investment from metaverse to wearables,” the company said.
This was more than an internal reallocation of resources. It amounted to an admission of failure for the biggest gamble in Meta’s history, one so sweeping that it began with a change to the company’s own name. What was meant to position Meta at the center of the next technological revolution instead left it betting on the wrong one, while competitors such as Google, and startups as hungry as Facebook once was, overtook it when the AI revolution arrived. Now Meta is scrambling to close gaps and cut spending on a vision that will not materialize, at least not in the foreseeable future.
In October 2021, Facebook was in the midst of one of the most difficult periods in its history. Libra, the company’s ambitious cryptocurrency project launched in 2019, was in its final stages amid regulatory pressure; its chief architect, David Marcus, would leave the company a month later. TikTok was entering a phase of rapid global growth, posing a serious competitive threat to both Facebook and Instagram. Apple’s privacy changes were beginning to bite into Meta’s advertising business, and the company’s share price was under pressure.
Worse still, a series of revelations based on internal documents leaked by former Facebook employee Frances Haugen laid bare the company’s internal awareness of harms linked to its platforms, from Instagram’s impact on young users to the spread of misinformation around Covid-19 vaccines and even human trafficking. Meta’s public standing sank to a new low. (This October, a new film by Aaron Sorkin, a sequel to The Social Network, is set to be released, focusing on the Haugen revelations, though without Jesse Eisenberg as Zuckerberg.)
Facebook and Zuckerberg needed a reboot. That reboot arrived later that month, when Zuckerberg took the stage at the company’s annual developer conference to unveil a new vision centered on the metaverse, a virtual world that would be integrated across Meta’s products and built around virtual reality hardware such as the Oculus headset, which the company acquired in 2014. “We believe the metaverse is the natural successor to the internet,” Zuckerberg declared. In his vision, many of the activities performed in the physical world would move into immersive, hyper-realistic virtual environments, capable of transmitting subtle facial expressions and enabling complex actions such as virtual surgery.
The bet was so large that Facebook was renamed Meta. But the vision failed to convince skeptics. Calcalist described it at the time as “hollow,” noting that many of the promises were distant, speculative, or potentially impossible. Virtual reality, despite years of bold predictions, had yet to meaningfully transform daily life, and many earlier visions had collapsed as dramatically as they were unveiled.
The years since have only reinforced that skepticism. Meta failed to deliver breakthrough products that captured the public imagination. Virtual meeting spaces populated by rigid, cartoon-like avatars, requiring bulky headsets, proved a poor substitute for Zoom calls. Whatever limited success the metaverse achieved remained largely confined to gaming, and even there the numbers were modest. In 2022, The Wall Street Journal reported that a year after its launch, Meta’s metaverse platforms had only 200,000 monthly active users.
The disappointment is starkly reflected in Meta’s financials. In the third quarter of 2025, revenue from metaverse-related operations, reported under the Reality Labs division, totaled just $470 million, less than 1% of the company’s overall revenue, while operating losses reached $4.432 billion for the quarter. Cumulatively, Reality Labs has generated close to $80 billion in operating losses. These are funds Meta would undoubtedly prefer to have today, as it is forced to invest hundreds of billions of dollars in AI infrastructure.
That is not to say Reality Labs produced nothing of value. The division delivered one of Meta’s most successful hardware products to date: the Ray-Ban Meta AI glasses. Equipped with cameras, speakers, microphones, and connectivity to Meta’s AI models, the glasses allow users to interact with a chatbot, receive real-time contextual information, and capture photos and video hands-free. By February, the glasses had reportedly sold more than 2 million units, and in the second half of 2025 Meta held more than 70% of the still-nascent smart glasses market.
Yet the glasses’ success underscores the metaverse’s failure. They succeeded not because they immersed users in virtual worlds, but because they anchored the defining technological shift of the era, artificial intelligence, in the physical world. While Meta was fixated on virtual reality, companies such as OpenAI and Google were quietly laying the foundations for the real revolution: large language models.
AI was never absent from Meta’s strategy, but under the metaverse vision it was not central. When ChatGPT burst onto the public stage in November 2022 and ignited the AI race, Meta was unprepared. To its credit, the company reset quickly, embracing open-source models and accelerating development, but it remains in catch-up mode. In recent months, Meta has spent billions on acquisitions and aggressive hiring to secure the talent, intellectual property, and expertise needed to close the gap, investments that might have been less urgent had it not poured so much capital into the metaverse.
Last week’s layoffs at Reality Labs were the clearest signal yet that Meta has accepted the verdict. Alongside the job cuts, the company is shutting down VR game studios, freezing or canceling advanced headset development, and redirecting remaining teams toward AI-powered and augmented reality glasses.
Meta is unlikely to formally bury the metaverse vision, or change its name again. Corporations rarely admit mistakes of this scale. But the layoffs and strategic pivot say what executives will not: skeptics were right as early as the fall of 2021. The metaverse was a cinematic fantasy, not an imminent future. And Meta has finally stopped paying for the illusion.
