Alex Gunz revisits the ongoing Twitter saga following its takeover by Elon Musk.
Back in December I wrote a blog predicting the outcome of Twitter’s takeover by Elon Musk. My argument was that it would go down one of two paths (and only those two paths), but that it was impossible to know which one. Has this Schrödinger’s bird theory held up? And does Twitter’s recent drama mean it needs revisiting?
My original reasoning was that big social networks like Twitter benefit from a ‘network lock-in’ effect. Here’s how it works - the thing that makes networks attractive (even addictive) is not so much their engineering, but rather the boundary-spanning connections they make between a lot of people who are otherwise hard to find.
Twitter’s real value was that multiple elites from around the world (academic, political, journalists, sports, media, etc) all chose to communicate there, along with everyone else, and this created a uniquely rich pool for interaction and information flows.
Why does that function as a lock? Because even if many people want to leave, they also want to be where everyone else is, and everyone else is still there. It’s a classic collective action problem: each individual is afraid to leave in case nobody follows so hardly anybody ends up leaving, and the crowd remains large.
Those few who do move, then, are sat on their own, thinking that maybe they ought to go back. Smaller tighter communities might be able to coordinate their way past this problem, but loose and diverse ones will struggle to.
My prediction was that one of two things would happen. Either network lock-in would hold a sufficient number of users together and Twitter would keep its status as the ‘public square’. Or else defections would reach some critical mass, the lock-in would break down, and Twitter’s audience would deflate like a football in the hands of a scissor-wielding six-year-old.
So, was I right? Mostly yes, but partly no. The part I was right about was that Twitter did largely manage to tread the first path. There was some curiosity around alternatives like Mastodon, but these never hit critical mass, and Twitter largely kept a lock on its place as the public square.
The part I was wrong about was my downplaying of the importance of social media engineers. Two of Twitter’s engineering moves have had enough impact on their audiences that I must revise my proposition.
The first move was Twitter taking away its verified status from legitimate public interest figures and giving it, instead, to anyone with a few bucks and an axe to grind. Elon Musk did this to try to monetise writers’ desire for more ‘clout’ (i.e., audience size), but what he was misunderstanding was that Twitter’s clout algorithm is not there for the writers’ benefit, but for readers.
It was built to keep people hooked by feeding them a steady stream of voices that they wanted to hear from, saying (for better or worse) the sorts of things they wanted to hear. Replace this with content from people who are paid to be put in front of you, and readers very quickly have much less reason to spend their free time obsessively scrolling through their Twitter feeds (again, for better or worse).
This also makes it harder to find relevant information and analysis, which takes away one of Twitter’s main unique advantages. Some anecdata that this may indeed be happening comes in the form of the laments that are starting to be heard from users (and especially journalists) that Twitter is no longer nearly as good as it once was at providing real-time insight when dramatic things occur in far-flung parts of the world.
The second move was Musk’s decision to fire so many engineers, which seems to be starting to cause problems now too. Network capacity issues recently caused Twitter to restrict the number of tweets that users could see, leading to much anger and displeasure. This arguably wouldn’t have been necessary if they’d had more engineering resources still in place.
I’m forced to concede, then, that engineering may matter, though perhaps only in one direction. Network lock-in might protect you against your competitors’ superior engineers, but perhaps not necessarily against any inadequacies in your own.
Consumers are deeply reluctant to switch to another social network when they can’t bring along the people they want to talk to, but you may nevertheless push them out if you don’t maintain some minimum viable floor in your service level.
Predicting the future
What next, then, for Twitter? I predict that its future depends on two things. First, whether its floor on usability stays low. Second, whether an alternative emerges with a floor that a sufficiently large number of people view as sufficiently more acceptable (perhaps BlueSky or Instagram Threads?).
If both of those things do happen, then a slow-bleed of Twitter users could hit critical mass and turn into a flood. If that inflection point does arrive then we might all find ourselves reading a lot of media stories about the surprising speed with which Twitter’s public-square status collapsed.
Schrödinger’s bird does, then, seem to be back in its box. Only time will tell if tweets still emerge when it is opened.