We think of the world in abstractions: you press a button and a light turns on. It's a necessary shortcut, not a bulletproof one. Joel Spolsky famously wrote about how programmers experience abstraction leakage, that moment when you push against the limits and bugs of a high-level programming abstraction and suddenly have to know about and deal with all the messy details underneath. There's always, always, some messier lower-level detail, from an esoteric networking configuration that makes your Facebook business possible to handling cosmic rays flipping memory bits.
A system doesn't need to be very complex to have abstraction leakage — sometimes your water pipes break and your open-faucet-and-water-is-there abstraction leaks in a very literal sense — but the more layers you add to it the more chances there are for leakage and the more work you have to put to prevent and handle them.
Tesla offered last week a clear example of this. "Use a key to open a car door" has been a pretty basic abstraction for decades now. "Use a magic key to open a car door from far away" might sound like an interesting but straightforward upgrade of this abstraction, but under it there are many sophisticated moving parts, and when one of them fails you suddenly find yourself facing a server error message when you try to open your car door.
The Metaverse promises an escape from all of this, leaving behind the leaky messes of reality and social networks and corporate processes for a neat, new, better world. This promise is an obvious lie.
"The Metaverse" is not a place, not an universe, is not The Matrix. It's, or rather will be or would be or is the underside of the marketing promise of, an unholy combination of Fortnite and Facebook and Zoom, the warmed-over corpse of AOL and Geocities and every other corporate dream of taking the fabulous, creative, lightly-connected wilderness of the Internet, dressing it up in a movie archetype or three, and replicating there all the scarcities and limitations (and associated opportunities for profit and power) of physical space.
I'm not sure which ones of the proposed Metaverses will be built, but I'm sure none of them will become the Internet-eater they are betting on, and I do know whatever is built will be very leaky. Not just in the technical sense of, say, network configuration issues somewhere in the US breaking down an office building "in the Metaverse" used by people living in Madrid, but also for things like fraud, IP abuse, political manipulation. and every other "meatverse" problem that will unavoidably leak upwards to a Metaverse just as it does to a social network or an e-commerce company. Adding a VR interface and a power fantasy of every activity having to go through the rules (and toll booths) of your company on top the basically broken idea of the giant corporate walled garden will not make it any less broken.
Can we build a good Metaverse? To be a Metaverse it must be deeply interoperable: everything and everybody must be able in principle to interact with everything and everybody. To be good, everybody must be able to set up their own corner of it, not beholden to a company, an interface, or a goal, simply through committing to a minimum of technical rules, no of which forces you to buy from or depend on a particular provider, or even any at all. You must be able to create whatever you want, however you can, from a huge billion-dollar investment to something you did in half an hour on a lark. It must be able to support businesses, but built from the start for free peer-to-peer interactions. A place of invention with low barriers of entry and no ceilings to growth, where billions can get information and one person doesn't have to get permission from anybody to build an encyclopedia about something only three people will ever care about.
We can build a good Metaverse. We did build a good Metaverse. We called it the Internet. It's leaky, but so is everything, and, unlike the big walled gardens and the nightmarish utopia of an all encompassing metaversal one, we can grow and improve our little corners on our own.