Elon Musk may want to get rid of his deal with Twitter, but he has some ideas on how to run the bird app, and it involves layoffs, subscriptions, and… a sarcasm button. Musk showed up on Thursday for a video chat with Twitter employees, and the employees immediately leaked the content to reporters — including my roadside colleague Alex Heath and The New York Times’ Mike Isaac, who kept a live blog of the event as it took place. A apparent digression on aliens nevertheless, the results of the meeting were fairly predictable, but enlightening for anyone who has been too obsessed with ominous phrases like “authenticate all people” in recent months.
At Thursday’s meeting, Musk had the energy of a wealthy MMORPG fan buying a studio so he can implement his totally rad spells and weapon designs, while beleaguered game designers worry about the day-to-day activities of their jobs. (In fairness to wealthy gamers, when this ever happened literally, at least the developers didn’t beg their new boss to stop talking them out in public.) Twitter employees repeatedly asked if they could work from home, a promise get from Musk that “exceptional” employees can stay remotely. In less positive developments, Musk echoed hints that Twitter will cut jobs to become profitable. That plan sits alongside tactics like upselling Twitter users on subscriptions and adding TikTok-style algorithmic recommendations, plus the pillars of your average internet business like payment processing.
Technoken on Twitter may be more fun than hanging out with the rest of Musk’s business empire this week. Tesla’s cars are getting more expensive (along with everything else) and workers are being laid off. His lawyers are still seeking a sympathetic court for his years-long tweet-driven battle with the SEC, and they will likely charge Musk for a few more hours to settle the long-running lawsuit of a crypto buyer who accused him of extorting Dogecoin. The FAA is asking SpaceX to make some changes to its Texas launch site, while SpaceX employees are distributing an open letter asking Musk to stop tweeting for God’s sake. SpaceX has reportedly responded by firing at least five of them, a move reminiscent of a retaliation that landed him in legal hot water at Tesla.
At Twitter, Musk still has no responsibilities. He told employees he wants to “steer the product in a certain direction” in the long run, but that he’s “not tied to titles” and doesn’t really care about being CEO. For now, he can just dial in to his crappy hotel Wi-Fi and riff on potential new features like an “irony” label that tells you whether tweets are serious or not. But the more Musk talks about what he would change, the more contradictory his view becomes.
As funny as I find the concept of an ironic button, it’s a classic type of addition to the service: something users hacked together years ago, integrated into the formal interface. (/srs!) But Musk also seems to be just throwing ideas at the wall and sending them back when questioned, with no clear vision beyond “getting a billion users and becoming hugely profitable,” a far cry from his early calls. for unobstructed speech. He’s willing to casually suggest plans that would turn the way Twitter works upside down, but if pushed, he’ll withdraw into positions the company has effectively held for years.
Take the aforementioned “all” people authentication, something Musk promoted as a way to fight spam bots. Verifying that every Twitter user represents a real person would likely be disruptive and compromise anonymity, a feature Twitter tried to keep before Musk. Possibly for that reason, Musk scaled back the idea in Thursday’s meeting, discussing a possible Twitter Blue authentication service where people would pay to prove they’re human and prioritize their supposedly more trustworthy tweets. The thing is, Twitter already gives priority things like answers based on the credibility of the account. And if you’re concerned about free speech, there’s a real trade-off in giving users massive priority based on their ability to pay. So Musk’s proposal will either be a minor tweak to something Twitter is already doing, or seriously jeopardize the speech capabilities of ordinary non-billionaire users.
Musk made an equally well-established distinction between “freedom of speech” and “freedom of reach” on Thursday. “I think people are allowed to say pretty outrageous things that are within the bounds of the law, but then it doesn’t amplify it, it doesn’t get, you know, a huge reach,” he said. “We have to strike a balance between allowing people to say what they want to say, but also making people feel comfortable on Twitter, otherwise they just won’t use it.” The speech/reach division has been a common topic of conversation among platform executives for years, and reducing the visibility of sketchy content is standard practice for Facebook and Twitter themselves. It’s a core part of the vision for Bluesky, the open-source Twitter offshoot that predates Musk, and more tried-and-true decentralized platforms like Mastodon have struggled with the complications of the principle.
It’s also a extremely ironic for Musk to call out because Musk has repeatedly complained about Twitter limiting the reach of content, especially to be content. In April he was speculate about a “shadow ban” that suppresses a tweet that insults Bill Gates, and shadow ban is the purest expression of limiting reach: you can pretty scandalous tweet, but other people don’t have to. Musk has suggested that it would be different if the limits were transparent so that Twitter could solve any problems by making its recommendation algorithms “open source” and letting people investigate them. Like Will Knight at wired explained, this is a red herring. There are real benefits to opening the algorithmic black boxes of social networks, but it almost certainly won’t tell the average person whether their “Bill Gates looks like a pregnant man” tweet should have more favorites organically.
Musk has, for lack of a better term, a dedication to a particular free speech aesthetic. He likes provocative trolls and portrays himself as part of a common sense, straightforward center of American politics, declaring in Thursday’s meeting that he is “the center of the normal distribution of political opinion in the country.” (It’s true that he has his political base covered with both parties, but he also recently tweeted his support for Florida Governor Ron DeSantis — a stark far from centrist Republican — to run for president.) He often describes his support for speaking “within the bounds” of the law,” repeat the phrase at least three times in the Q&A.
However, when faced with the many problems associated with connectivity, Musk sounds like any other risk-averse social network operator. If anything, he seems unusual interested in shaping what is seen on Twitter. Per recode Transcript of the meeting, one of its big goals is for Twitter to offer a more socially aware version of TikTok’s powerful recommendation algorithm, providing users with interesting and informative tweets (I’ve edited the quote slightly for a little more, uh, clarity):
It is important to make Twitter as attractive as possible. And really, that means not showing people content that they would find hateful or offensive, or even frankly content that they would find boring is not good. We don’t even want them to see boring content. Unless – we were talking about TikTok last night. And of course TikTok does a great job of making sure you don’t get bored.
You know, TikTok is interesting, but you also want to be informed about serious matters. And I think, in terms of serious stuff, Twitter could be a lot better at informing people about serious stuff. I do think it’s important that when there are two sides to an issue, it’s important to represent multiple opinions. But you know, and just make sure we’re not some kind of driving story. There will be – give people the opportunity to understand the different sides of problems.
TikTok is also a fascinating case study on the boundary between moderation and invasive censorship. It has almost completely escaped charges of political bias, even during that weird period when Trump wanted to ban it from the country — possibly because the people who shape free speech discourse don’t get together much there. Far from being a “floating story,” however, the algorithm has spawned a bizarrely emerging vocabulary thanks to a soft ban on words like “suicide” and changed the way a generation speaks. Algospeak is everywhere. It’s the kind of system that should prompt deep thought about the power of social networks.
Instead, Musk seems as confident as ever in his power to dictate apolitical and neutral moderation — assuming he will one day wield the banhammer.