The metaverse was, for about two years, the most heavily promoted idea in technology. Then the money and the attention moved to artificial intelligence, and the word largely disappeared from view. It is worth understanding what it actually meant, partly because the underlying ideas did not go anywhere.
Where the word comes from
The term was coined in Neal Stephenson’s 1992 novel Snow Crash, describing a shared virtual world people inhabit through avatars. It sat in science fiction for three decades before technology companies adopted it as a business plan.
That origin matters, because the fictional version was a dystopia. The people in it were escaping a collapsed physical world, which is not usually mentioned in corporate presentations.
What was actually being proposed
Definitions varied by whoever was selling one, but most shared four claims:
- Persistent. The world continues whether or not you are logged in, unlike a game session that ends.
- Shared. Many people occupy the same space simultaneously and can interact.
- Embodied. You are represented by an avatar, often through virtual or augmented reality rather than a flat screen.
- Interoperable. Items and identity move between different virtual spaces owned by different companies.
The first three already existed and had for years. Persistent shared worlds with avatars date back to the 1990s, and games like Second Life, World of Warcraft, Minecraft, Roblox and Fortnite are all recognisably that.
Interoperability was the whole idea, and it never happened
The genuinely new claim was the fourth: that you could buy something in one company’s world and carry it into another’s.
This is where the concept collapsed, and the reason is commercial rather than technical. Platforms make money by keeping users and their purchases inside. A company that lets you take your possessions to a competitor has given away its main advantage.
There were also real technical obstacles. An item built for one game’s art style, physics and balance is meaningless in another. But those were solvable; the incentive problem was not.
What went wrong
The hardware asked too much. Headsets were expensive, heavy, isolating and uncomfortable for long sessions. A technology requiring you to strap a screen to your face fights a permanent battle against convenience.
There was no compelling reason to be there. The flagship demonstrations were meetings and shopping, both of which most people found worse than the existing versions. A video call is easier than a virtual meeting room, and a website is easier than a virtual shop.
It got entangled with speculation. Virtual land sales and blockchain assets attached the concept to a speculative bubble, and much of the reputational damage came from that association rather than from the technology.
Then AI arrived and absorbed the investment, the engineering attention and the narrative almost entirely.
What survived
Quite a lot, under different names.
Virtual reality is a healthy niche in gaming, fitness and simulation. Training in VR is genuinely valuable where the real thing is dangerous or expensive: surgery, aviation, industrial maintenance.
Augmented reality, overlaying information on the real world rather than replacing it, has proved more practical than full immersion and is where most current headset development is aimed.
Persistent social worlds are thriving, they are simply called games. Roblox and Fortnite have hosted concerts and events with audiences larger than physical venues could hold. That is the metaverse as originally described; it just never needed the label.
The useful lesson
The metaverse is a good case study in a recurring pattern: a real technology, a genuine long-term direction, and a marketing campaign that arrived a decade early and promised a product nobody had built.
The pieces still exist and are still improving. What failed was the claim that they added up to a single destination everyone would move into, and the assumption that competing companies would cooperate to build it.
Join the discussion