The Metaverse in 2026: What Happened to the $46 Billion Bet on Virtual Worlds
Remember when Meta changed its name from Facebook and poured $46 billion into building the metaverse? Three years later, the company quietly rebranded its metaverse division, laid off thousands of Reality Labs employees, and pivoted its messaging from “the metaverse is the future” to “AI is the future.” The grandiose vision of a persistent, interconnected virtual world where billions of people live, work, and play has not materialised. What has materialised is a collection of useful but unglamorous VR and AR applications that serve specific use cases well — enterprise training, gaming, fitness, and spatial computing — without becoming the general-purpose platform that tech companies promised.
What’s actually working
The most successful VR/AR applications in 2026 aren’t trying to be “the metaverse.” They’re solving specific problems. Meta’s Quest 3S at $299 has become a genuinely good gaming and fitness device — Beat Saber remains the killer app, and Supernatural has a loyal fitness user base. Apple’s Vision Pro carved a niche as a $3,499 productivity device for spatial computing — replacing multi-monitor setups with virtual displays floating in physical space for a small but enthusiastic user base.
Enterprise VR is where the real money is. Walmart trains over 1 million employees annually using VR simulations that replicate Black Friday crowd management and other scenarios. Surgeons at Johns Hopkins use VR for pre-operative planning, visualising patient anatomy in 3D before the first incision. Boeing uses AR headsets on its 737 production line, overlaying assembly instructions directly onto aircraft components and reducing wiring harness installation time by 25%.
The VR/AR market in 2026
| Platform | Price | Best use case | Scale |
|---|---|---|---|
| Meta Quest 3S | $299 | Gaming, fitness, social VR | ~30M monthly active |
| Apple Vision Pro | $3,499 | Spatial computing, productivity | ~2M units sold |
| PlayStation VR2 | $549 | Console gaming, exclusives | ~5M units sold |
| Microsoft HoloLens 2 | $3,500 | Enterprise AR, manufacturing, military | Enterprise contracts only |
| Roblox (all devices) | Free | User-created games, social, events | ~80M daily active |
Why the original vision didn’t work
The metaverse hype of 2021–2022 failed for several overlapping reasons. Hardware wasn’t ready — VR headsets were too heavy, too expensive, and too isolating for extended daily use. There was no compelling reason for most people to spend hours in a virtual world when the physical world and flat-screen apps served their needs. Social dynamics didn’t translate — Horizon Worlds and Decentraland attracted small enthusiastic communities but never achieved the network effects needed for mainstream adoption. And the AI boom stole the spotlight and investment dollars, redirecting corporate attention from spatial computing to language models.
The lesson isn’t that VR and AR are dead. It’s that immersive technology follows the same adoption curve as every other transformative technology: useful applications come first, then infrastructure, then platform effects. The smartphone took 15 years from the first iPhone to ubiquity. VR/AR is probably on a similar timeline.
What comes next
The next inflection point will likely come from lightweight AR glasses — devices that look like normal eyeglasses but overlay digital information on the physical world. Meta is reportedly spending heavily on “Orion,” AR glasses designed for everyday wear. Samsung and Google are co-developing an XR platform. Apple is rumoured to be working on a lighter, cheaper Vision device for 2027. If any of these deliver a comfortable, all-day-wearable device with useful AR features at a consumer price, it could be the iPhone moment that VR headsets never achieved. Until then, the “metaverse” is best understood not as a destination but as a set of technologies — VR, AR, spatial computing, 3D engines — that are individually useful and collectively building toward something bigger. Whether that something arrives in three years or ten remains genuinely open.
