After Donald Trump was banned from Twitter in 2021, Donald Trump Jr. made a public appeal to Elon Musk for help. “Wanted to come up with something to deal with some of this nonsense and the censorship that’s going on right now, obviously only targeted one way,” he said in a video that was posted to Instagram. “Why doesn’t Elon Musk create a social-media platform?” (The video was titled “Here’s How Elon Musk Could Save Free Speech.”)
This was—I think we can say it—prescient.
A little more than a year later, Musk was promising not an entirely new site, but a hostile takeover of a familiar one. And he explicitly presented this action as a corrective to right-wing grievances about “shadowbanning” and censorship. He promised to use his new platform to combat the “woke mind virus” sweeping the nation and said he wanted to save free speech. (His supposed devotion to unfettered expression, it’s worth noting, sometimes comes second to his personal feuds.)
So here we are. Liberal activists used to be the ones suggesting that the social network could be used to organize in defiance of the state; now technology accelerationists are the ones saying this. “Elon acquired Twitter, fired the wokes, and removed DC’s central point of control over social media,” the tech-world iconoclast Balaji Srinivasan wrote in November. At a conference he led in Amsterdam the month before, he talked about how the tech world could build a “parallel establishment” with its own schools, financial systems, and media. Co-opting existing organizations could work, too.
Previously, the “alt-tech” ecosystem was a bit of a sideshow. It encompassed moderation-averse social-media sites that popped up in the Trump era and resembled popular services such as Twitter and Facebook; their creators typically resented that their views had been deplatformed elsewhere. Parler and especially Gab (which is run by a spiky Christian nationalist) were never going to be used by very many normal people—apart from their political content, they were junky-looking and covered in spam.
But now, alt-tech is emerging from within, Alien-style. Twitter’s decade of tinkering with content moderation in response to public pressure—adding line items to its policies, expanding its partnerships with civil-society organizations—is over. Now we have X, a rickety, reactionary platform with a skeleton crew behind it. Substack, which got its start by offering mainstream journalists lucrative profit-sharing arrangements, has embraced a Muskian set of free-speech principles: As Jonathan Katz reported for The Atlantic last month, the company’s leadership is unwilling to remove avowed Nazis from its platform. (In a statement published last week, Hamish McKenzie, one of Substack’s co-founders, said, “We don’t like Nazis either,” but he and his fellow executives are “committed to upholding and protecting freedom of expression, even when it hurts.”) The trajectory of both resembles that of Rumble, which started out as a YouTube alternative offering different monetization options for creators, then pulled itself far to the political fringes and has been very successful.
These transformations are more about culture than actual product changes. Musk has tinkered plenty with the features of Twitter/X in the past year, though he’s also talked about changing far more than he actually has. More notable, he’s brought back the accounts of conspiracy theorists, racists, and anti-Semites, and he got rid of Twitter’s policy against the use of a trans person’s deadname as a form of harassment. In a recent Rumble video, the white supremacists Richard Spencer and Nick Fuentes praised Musk’s management of the platform, saying that the “window has shifted noticeably on issues like white identity” during his tenure. And in support of anecdotal claims that hate speech rose after Musk’s acquisition of Twitter, a team of researchers has shown that this was actually the case. They observed a large spike right after the acquisition, and even after that spike had somewhat abated, hate speech still remained higher than pre-acquisition levels, “hinting at a new baseline level of hate speech post-Musk.”
Social-media platforms are kind of like parties: People’s perception of them matters almost as much as the reality. You can see changing attitudes about whom Twitter or X is for in recent polling from the Pew Research Center. Two years ago, 60 percent of Republican or Republican-leaning Twitter users thought the site was having a negative impact on American democracy. In 2023, the number was just 21 percent. And the percentage of those users who thought Twitter was having a positive impact jumped from 17 to 43 percent. Conversely, Democrats and Democratic-leaning users were more likely this year than they were two years ago to say that Twitter was having a negative impact on American democracy, and less likely to say it was having a positive one. Pew also found a partisan divide regarding abuse and harassment on the platform, with 65 percent of Democratic users saying these are major problems and just 29 percent of Republican users agreeing. The gap between the two positions has quadrupled in the past two years.
When I spoke with Keith Burghardt, a computer scientist at the University of Southern California who worked on the hate-speech study, he emphasized that the research doesn’t address the specific cause of the increase. It could be that reductions in staff or the disbanding of Twitter’s Trust and Safety Council had a significant effect. Or that may have stemmed from other changes in moderation policy or enforcement that aren’t visible to the public. It could be that Elon Musk’s public statements made people rush to the site to see what they could get away with saying. “In fact, it’s important to mention that we found hate speech increased a little bit even before Elon Musk bought Twitter,” Burghardt told me. “Perhaps because of users anticipating a perceived drop in moderation.”
This is not an end point but a funky state of limbo. Instagram’s Threads signed up 100 million people in its first week, but activity dropped after the big debut and growth appears to have slowed. Platform migration is complicated, and early research has found that many people who are unhappy on X have not left the site entirely. Instead, they tend to make secondary accounts on alternative sites. They show “wavering commitment” to staying on X, while still being more active there than they are on alternatives like Bluesky or Threads. Pew data published in May showed that the majority of “highly active tweeters” were still Democrats and Democratic leaners. However, these people were posting less frequently than they had been before, and this data was collected before Musk’s recent public display of anti-Semitism.
“Does the Musk Twitter Takeover Matter?” Deana Rohlinger, a sociology professor at Florida State University, asked in a February analysis of the site’s supposed mutation. The question was rhetorical; when I spoke with her recently, she said the answer was definitely yes. “Despite its flaws,” Twitter was something of a common space in its prime, she said. New microblogging sites may want to serve that same purpose, but she isn’t sure that it will be possible. The hostility between these two entrenched, polarized online factions may be so much that they just don’t want to share a space anymore. “Perhaps it’s a reflection of our broader political environment and media environment,” she said. “I don’t know that you can re-create what Twitter did, because things have changed entirely too much. It’s not 2006 anymore.” The Twitter diaspora, she thought, was cursed to just drift.
As alt-tech has taken over the mainstream, the old mainstream has found itself in a funny position. Five years after the #DeleteFacebook campaign, many are cheering Mark Zuckerberg—a literal Elon Musk sparring partner—as a hero in the platform wars. Our only response to the current state of the web seems to be a sigh of resignation: Sure, let’s just do everything on Instagram.