RIP metaverse 🪦

In October 2021, Facebook changed its name to Meta (META). This was a company building the next version of the internet. The new moniker would bring together the firm’s apps and technologies, “to bring the metaverse to life.” The best way to predict the future is to create it. 

The metaverse was pitched as a place where people would work, shop, hang out, play, attend events, build businesses and own digital goods. In the new utopia, you could just do things. 

Look on my worlds, ye mighty

Snoop Dogg bought a house there. Microsoft described its $68.7bn Activision Blizzard takeover as “building blocks for the metaverse”. JP Morgan opened a lounge, describing the metaverse opportunity as “limitless”. Gucci sold digital bags.

But the non-virtual reality of the metaverse was headsets and motion sickness, legless avatars and NFTs. Importantly, there wasn’t a great deal to actually do there. Perhaps because of all this, Meta is now scaling back its ambitions, after spending a mega fortune on the meta project. Besides, there’s a new cool kid in town, which you may have heard of: AI. 

Capital idea

Not even half-a-decade ago, money was pouring into the metaverse. In 2022, McKinsey said $120bn had been invested in the first five months of that year, more than double 2021. The consultancy expected the metaverse could generate up to $5tn(!) in value by 2030.   

Finance was told to build, build, build! Banks should offer adviser meetings in virtual worlds, train staff for simulated bank robberies in VR, and even think seriously about lending against virtual property. 

The metahype hit humanity at just the right time. This was the era of cheap money. The pandemic had normalised wasting inordinate amounts of time in digital worlds. Gaming, in particular, had shown that people would spend real time and real money in virtual spaces. And remote work had left companies wondering whether the office would ever return. It was now clearer than ever people no longer had to be in close proximity to one another in order to work and play. 

Reality bites

The contemporary reality isn’t quite so optimistic. In 2025, Meta’s VR/AR arm, Reality Labs, posted $2.2bn in revenue and an operating loss of $19.2bn. In 2024, it made $2.1bn in revenue and lost $17.7bn.

In March this year, Meta said Horizon Worlds and Events would soon disappear from its Quest store. Then there was a note indicating these existing worlds would stay in VR. Was the platform being scaled back rather than scrapped?

Meta has already laid off more than 1,000 Reality Labs employees this year. And closed three VR studios. And discontinued its work-focused metaverse platform, and paused content for its VR fitness app Supernatural. The writing is on the virtual wall. 

Hype swerve

Not every metaverse-adjacent company has hit the brakes. In 2025, Roblox (RBLX), an immersive gaming and creation platform aimed at youngsters, saw revenue up 36% year-on-year to $4.9bn, bookings up 55% to $6.8bn and average daily active users at 97.8m. Digital worlds work when they are built around communities and a product people actually use. 

But the firm faces many of the same problems as other virtual worlds, including the online world in which we all now live: how to keep children safe, and moderating content at a vast scale.

Gaming is not entirely immune from metashock, though. Rec Room, a social gaming platform similar to Roblox, said it is shutting down 1 June 2026, despite having reached 150m users and once being valued at $3.5bn. 

Move over, metaverse

Tech manias often collapse when promise gets ahead of product. There were real ideas inside the metaverse. Many of them may still come to bear. But the metaverse was a lot. Creating a whole new category is costly. Platforms are built over many years, something Meta knows all about. But someone had to build the metaverse, and our man Mark stepped up.

AI is the inevitable future now. It is the story investors want to fund. Meta said the majority of its $160bn-odd expense growth this year would be driven by infrastructure costs, and the second largest contributor would be employee compensation tied to technical talent, “particularly AI”. Meta also said it expected 2026 capital expenditure to land between $115bn and $135bn, driven by investment to support its superintelligence labs and core business.

Meta is still carrying large Reality Labs losses. It is still selling Quest devices. But the centre of gravity is now firmly with AI and the race to build what Mark Zuckerberg calls personal superintelligence. 

The metaverse may turn out to be a very expensive stepping stone toward more wearable, persistent computing, just away from the consumer. Microsoft’s industrial metaverse describes a mix of digital twins, IoT, mixed reality and cloud data used in industrial settings. In defence, AR has been folded into helmet displays, battlefield visualisation and training systems.

Ready player done

Headsets remain heavier and pricier than phones or laptops. Importantly, nobody has figured out how to make users look good with the devices strapped on. Apple’s (AAPL) $3,500 Vision Pro is receiving fresh visionOS 26 updates. There’s even a newer model available. But the device is still sold as a premium product.

The smartphone, for all its faults, folds neatly into everyday life. A headset asks users to tolerate a level of friction most consumer tech has spent two decades trying to eliminate. However impressive the demo, that is a much harder sell.

The figures speak for themselves. Apple shipped only about 45,000 Vision Pro units in 2025, down from roughly 390,000 in 2024. Meta did better, shipping 1.7m Quest units in the first three quarters of 2025.

Big ick

There may be a broader reckoning here, too. Half-a-decade ago, these Goliathan platforms were less exposed and vulnerable than they are today. A social media wake-up call has been brewing for a while but, post-Haidt, regulators and public officials are now looking very closely at these products, especially with regards to younger users. Do their algos and engagement tools cause foreseeable harms? Spoiler alert: they do. 

The metaverse is at odds with a more mindful approach to tech. There’s a growing feeling that we, adults and children alike, spend too much time looking at screens, doomscrolling and bickering. The idea we would situate ourselves in virtual space for much of our waking lives feels unseemly in the current climate.   

Things are certainly changing. A landmark case in Los Angeles saw a jury find Meta and YouTube (GOOGL) liable for designing addictive products that harmed a young user, awarding $6m in damages and assigning 70% of the responsibility to Meta and 30% to YouTube. In New Mexico, a separate jury found that Meta knowingly harmed children’s mental health and safety and violated state law, awarding $375m in penalties. Social media bans for under-16s will no doubt become more common. 

Story selling

The term metaverse was coined by sci-fi author Neal Stephenson in his 1992 novel Snow Crash, to describe a shared digital world where people appear as avatars. In the novel, the metaverse is a place to escape the tedium of reality, rather than the all-purpose app later envisioned by Big Tech.

The metaverse story was, even at its peak, met with a fair degree of scepticism. There was never really a killer use case for it. Strip away the pitch and it was often hard to say what it was really for. AI may be the capital story currently being spun, but its utility and versatility are unlike anything the metaverse promised. 

That the metaverse began in fiction feels fitting: all stories, however grand, eventually come to an end.