Two questions for Twitter's future

How long can this last?

At midnight Pacific Time on Wednesday, an email arrived in Twitter employees’ inboxes, ostensibly from Elon Musk, reading, in part, “going forward, to build a breakthrough Twitter 2.0 and succeed in an increasingly competitive world, we will need to be extremely hardcore.” The email concluded with a link to a Google Forms document, with the instructions: “If you are sure you want to be part of the new Twitter, please click yes on the link below.”

If you are even minimally experienced at internet, you might assume that an email from your “boss,” written in somewhat strained English, concluding with an instruction to click “yes” on a Google form to keep your job, was a targeted phishing email -- a devious but ultimately recognizable attempt by unknown hackers to gain access to your corporate account. But this was a legitimate email, legitimately from Twitter’s majority owner and CEO, and any resemblance to an incursion attempt is due to the fact that, as David Roth put it, referring to a completely different incident in Musk’s short tenure, “the official voice of Twitter is now ‘Estonian phishing email.’”

When the deadline Musk set in his email passed on Thursday afternoon, early reports suggested that something like three quarters of Twitter’s remaining employees -- that is, those remaining after Musk laid off half of the workforce last week -- had clicked “no” on the Google Form, and elected to take three months severance pay rather than continue, hardcore-ly, at Musk’s Twitter. This was not just cosseted programmers, designers, product managers: Apparently the company’s entire payroll and tax team quit. (Musk had already fired nearly the entire HR and communications departments, which goes a long way toward explaining the Google Forms email.)

Amidst the fallout of the mass resignation, the tone of the reporting on Twitter alternated between “screwball comedy” --

Zoë Schiffer @ZoeSchiffer

Elon Musk and his team have a list of Twitter employees who signed the offer indicating they want to stay at Twitter 2.0. Those names are in a spreadsheet. Now they're trying to figure out who's NOT on the sheet, who is on parental leave, who's actively resigning in Slack...

-- to “chaotic coup d’etat in process” --

Zoë Schiffer @ZoeSchiffer

NEW: Twitter just alerted employees that effective immediately, all office buildings are temporarily closed and badge access is suspended. No details given as to why.

-- to “frantic dispatches from nuclear reactor meltdown”:

Nitasha Tiku @nitashatiku

!!! “I know of six critical systems (like ‘serving tweets’ levels of critical) which no longer have any engineers," "There is no longer even a skeleton crew manning the system. It will continue to coast until it runs into something, and then it will stop.”

washingtonpost.comHundreds said to have opted to leave Twitter over Musk ultimatumThe number of likely departures prompted Musk to ease his return-to-office edict and managers to meet to decide which engineers to ask back.

But of course this combination of unfolding emergency collapse, cutthroat power struggle, and enjoyable incompetence is precisely the kind of thing Twitter feels as though it was designed for, and the power-user core that writes the vast majority of tweets was delighted to follow along. What is Twitter, after all, if not a platform for rubbernecking and gawking? Many people, not least of them Musk, like to analogize Twitter to a digital “town square”; my feeling is that Twitter is probably better understood as a digital “sidewalk outside a crime scene,” its users pressing against the police caution tape, watching the bodies get carted out of the building.

This delight in catastrophe can often lead users of Twitter to overrate the likelihood of catastrophe. On Thursday night the rubbernecking glee seemed to spill over into a funny collective sense that Twitter’s demise was imminent on an hourly scale. Anxious about losing friends and contacts, people shared their Mastodon handles, Substack newsletters, email addresses. (A cynic might suggest that people were less anxious about losing “friends” than they were about losing the social/professional equity they’d obtained from the site. Not me, though.) A “Twitter Space” hosted by the deranged criminal mastermind and Buzzfeed journalist Katie Notopoulos served as a kind of wake, during which people from various countries and corners of Twitter talked about what the end of the service might mean for them. It ballooned to a 20,000 concurrent listeners, including VCs, YouTube creators, furries the Twitter-addicted model Chrissy Tiegen, and @masturbatelog, an account that tweets every time its owner masturbates.

Katie Notopoulos @katienotopoulos

I can’t believe almost 200k people listened (at one point like 20k concurrently) to the Space @broderick and I hosted last night about twitter. Thanks to everyone who listened or spoke, please know it meant Marc Andreessen had to listen to the furry cum pizza guy

I don’t blame anyone for cultivating a sense of cataclysm, if that’s a fun way to spend a Thursday night. But [insufferable nerd voice] Twitter isn’t going to suddenly disappear, never to return, today, tomorrow, or anytime soon. Some bad things could happen, and functionality is likely to decline. But I still am pretty sure that Twitter will survive more or less the same, at least in the medium term. What’s less clear to me is exactly how that survival plays out. I’ve already done too much Twitter prognosticating, but let me lay out what I think are the two key questions for what Twitter feels and acts like over the next 12-24 months:

1. How long can the complex system that is “Twitter” continue running on its own?

Based on what I’ve read and heard, the big threat to Twitter in the short to medium term isn’t simply the sheer number of people that have left, but that the whole layer of engineers and programmers who have “deep” knowledge of the technological infrastructure of the company is gone -- and, worse, that they left without leaving much documentation behind. This thread, from a former SRE (site reliability engineer), outlines many of the scenarios Twitter might fact in the next few months, and why the lack of institutional knowledge will make those scenarios much worse. I’ll quote a few:

1) Random hard drive fills up. You have no idea how common it is for a single hosed box to cause cascading failures across systems, even well-engineered fault-tolerant ones with active maintenance. Where's the box? What's filling it up? Who will figure that out? […]

4) Bad code push takes the site down *in a way that also fucks up the ability to push new code*. This is the absolute nightmare scenario for teams like mine. When something like this happens, it's all hands on deck. Without deep systems understanding, you might never get it back.

5) Mystery SEV. Suddenly, the site goes dark. The dashboard is red. Everything seems fucked. There's no indication why. You need to call in the big guns. Teams with names that end in Foundation. Who are they? How do you call them? […]

7) Someone, say, entirely hypothetically, @wongmjane, finds a critical security vulnerability in your prod iOS app. You need to fast-track a fix, *stat*. You have a team of experts who know how to navigate Apple's Kafkaesque bureaucracy for app updates, right? I sure hope you do.

How long can Twitter keep running normally before it runs into one or more of these problems? What if there was, say, a wildly, globally popular sporting event that a huge population of people wanted to watch and talk about at once?

The thing to be clear about here is that all of these dire scenarios represent fixable problems; it’s just that attrition and loss of institutional knowledge makes fixing them significantly more difficult. Brief hiccups in site functionality suddenly last hours, or even days; at the same time, little things like public-facing statistics or video embedding or tweet translation start breaking 10 percent of the time instead of one percent of the time. This MIT Technology Review interview with a former Twitter engineer outlines the process of decay:

He presents a dystopian future where issues pile up as the backlog of maintenance tasks and fixes grows longer and longer. “Things will be broken. Things will be broken more often. Things will be broken for longer periods of time. Things will be broken in more severe ways,” he says. “Everything will compound until, eventually, it’s not usable.”

Twitter’s collapse into an unusable wreck is some time off, the engineer says, but the telltale signs of process rot are already there. It starts with the small things: “Bugs in whatever part of whatever client they’re using; whatever service in the back end they’re trying to use. They’ll be small annoyances to start, but as the back-end fixes are being delayed, things will accumulate until people will eventually just give up.”

Without a real alternative at Twitter scale, it seems most likely that people will accept the loss of functionality as the price of being able to tweet “come to Brazil” at Dua Lipa. And presumably, Musk and the remaining employees will be staffing up with new engineers and contractors, learning the systems as they go. Dan Luu, a former Twitter employee and programming writer, doesn’t think that the loss of employees is an “existential threat,” but it makes Musk’s job significantly more difficult:

FWIW, unlike a lot of the more dire predictions, I think this is all recoverable and isn't an existential threat to the business (unlike the situation with advertisers, which could be very serious if Elon doesn't want to fund Twitter himself), but the way things have been handled so far make things much more difficult for seemingly no benefit.

There was a tedious but straightforward path to making Twitter wildly profitable within a year or two. Everything that people with institutional knowledge would've done can still be done and I think there's a decent chance that Elon will do it (I'd give him higher odds than either of the past two CEOs), but having new people ramp up and understand thousands of services, hardware operations, etc., is making the problem pointlessly difficult.

Which leads us into question two:

2. How much money is Musk willing to burn on the project?

Again, my guess is that Twitter is going to become less functional, but not so much less functional that the core of power users who have braided the site into their daily lives walks away. There’s no alternative place where you can meet new people, learn interesting news, get really mad at a guy you’ve never met and never will who thinks it’s white privilege to cook for for your neighbors, be asked for feet pics by a municipal election candidate, etc., and an 85-percent-functional Twitter will be regarded as “good enough” for the people who love that kind of thing. (What are you going to do on your phone when you have five free minutes, read a book??) Put another way, the rest of Twitter will go like the whole verification episode did: it’ll be stupid and annoying for a while, then chaotic, then pretty funny, and then, at the end of the process, basically the same as before, but slightly worse.

Or, that’s how it will go if Musk works hard at rehiring a workforce and repairing the site infrastructure. The problem is that arresting and reversing the decline in functionality is going to take a lot of money and a lot of time. One of the ironies of the Musk takeover is that Twitter -- a mismanaged company with great tech and a captive, influential user base -- was probably a pretty good candidate for a competent leveraged buyout operation. Musk, egged on by his friends and their weird psychodrama about “blue checks,” has now eliminated any possibility that his takeover will be smooth and inexpensive. He’s on the hook for the billion-plus dollars it will take to return Twitter to the state it was when he bought it -- let alone to improve it and make it more efficient. How long is he willing to keep trying before declaring bankruptcy and moving the company off his books?