Pitfalls ahead for the metaverse bandwagon

Amidst all the hype regarding the rolling out of the metaverse, there are some nitty gritty challenges that really need to be dealt with first.

At risk of leaping aboard an already-overloaded bandwagon, recent headlines suggest that something called a metaverse is being talked up using the sort of vague but ambitious terminology that stern-faced misanthropes might interpret as a low-effort pitch for venture capital if it wasn’t for Meta’s preexisting and almost infinitely-deep pockets.

Many of us bear the metaphorical scars of previous interactions with this sort of push towards an anticipated next big idea, many of which are characterised mostly by the reluctance of anyone to characterise them in any detail. After all, defining what it is might seem to risk defining what it isn’t, and who’d want to screw up series B by making that sort of mistake.

Assessing the practicalities

If that’s a sufficient tsunami of cynicism to have convinced the scowling engineers in the audience that we’re going to take a very cautious approach to this, let’s look at the practicalities. The idea of a sort of grand unified virtual world is an interesting one, and it’s the practicalities, or at least our reluctance to define them, which are fundamental here. This is, despite snickering comparisons, not much like stereo 3D: that was a well-defined technological solution looking for a problem, while anything like a metaverse is an idea in desperate search of a well-specified technological expression. 

If there is any widely-held assumption about the metaverse at all, it’s that the mere mention of the word might conjure up visions of people organising Covid-safe playdates for the developmentally isolated children of the pandemic in comically outsized VR headsets. A purist, though, might claim virtual reality is neither a necessary nor a sufficient condition for a metaverse. Given that, in extremis, it’s difficult to distinguish between a metaverse, which is a large, shared environment in which physically distant people can interact and share experiences, and the current experience of the internet, which is a large, shared environment in which physically distant people can interact and share experiences.

It’s also been proposed that we shouldn’t really be referring to a metaverse, but the metaverse, though that beckons a waste of potential; why have only one Ferrari in one garage when both are free to reproduce. At least some approaches also refer to a shared environment that has, possibly, in some respects, a one-to-one mapping over the real world. That might not entice anyone who’s recently returned from work and squeezed, sideways, breath held, into a modern big-city apartment. There’s not much joy to be had from a limitless virtual world that carefully duplicates a space so cramped that the closets are shared between adjacent rooms in the same way as the real ones (yes, seriously; recent UK accommodation blocks are barely a step up from hot bunking). 

Underpinning the metaverse

Without labouring the point any further, it seems safest to assume that these are all things that anything called a metaverse might do under some circumstances, depending on what everyone actually wants. Ideas like universality (so that anyone can access it and do things in it) and interactivity (more than the current internet, probably) seem important, as does the crucial idea that people will come to attach similar importance to things and events in a metaverse as they do to events in the real world. While all of that is very valid, though, it’s not a sufficiently solid plan that someone could sit down and start bashing out source code.

And that is the gap this debate needs to bridge. To boil this right down to the fundamentals, programming languages do not express vaguely-defined ideas very well, no matter how portentous anyone thinks those ideas are. They expresses mathematical certainty. It’s the job of a software engineer to bridge the gap between the desires of the client and the capabilities of the development environment, and that’s a full-time job even when the client has been very specific. For a metaverse to happen, the unfashionable, unpalatable, un-venture-capitalisable reality is that someone’s going to have to get specific, and persuade a lot of other people to agree.

We might wonder exactly what sort of organisation is in a position to do that. As Meta has hinted, with its publication of Facebook: The Game, the inevitable self-interest of a profitmaking organisation is almost antithetical to the principles of universality and access which are prerequisites of a worthwhile metaverse. The obvious alternative is that the software used to implement a metaverse must be open source, which is equally antithetical to universality and access in that it’s likely to have a user experience about as appetising as a sewage sandwich. That’s even trickier if we realise that the best advantage of a metaspace is only realised when people can build things in it which are useful to others.

Flattening the learning curve

One of those barriers is the arrogant and unspoken assumption of much of the open source community that everyone is, or should be, a computer scientist. That assumption has been firmly rejected by reality on more or less a continuous basis since the 1970s. Consider the world wide web which runs on top of the internet in the same way a metaspace might. Creating websites from scratch requires a practically occult combination of HTML, CSS, JavaScript, server-side languages such as PHP, server configuration files, and more. All of those languages, with the possible exception of Javascript, are laughably badly designed for what they’re often used to do. 

The result is that anyone can mount any website on any subject if, and only if, that person can negotiate an absolute Mobius loop of a learning curve. That’s why sites like the much-promoted Squarespace exist. That’s a website-generating website, which is not the meaning of the term “meta” anyone has in mind in this context. Similar problems (and half-baked solutions) must not stymie the creativity of a metaverse’s inhabitants.

So, metaverses must somehow avoid both the economic motives of commercial involvement and the abject user-hostility of most non-commercial development while finding a way to define whatever it is they’re really to be, without putting off anyone who wants it to be something else, or hyperventilating true believers who want it to simultaneously be everything.

The likelihood of anyone ever being able to step back, sigh with satisfaction, and declare the metaverse complete is very remote. Something useful may come out of all this. More probably, several useful things might, some of which might be vaguely recognisable as implementations of what we now call a metaverse. Similarly, Tim Berners-Lee probably didn’t see most of the modern internet coming and that’s fine. But it’d be nice if, knowing that, we could sidestep some of the more obvious and spike-infested pitfalls that many of us can clearly see approaching through the fog of poorly-articulated technological ambition.

Just like Meta didn’t.