Ever since Philipp Lorenz-Spreen stopped posting on X (formerly Twitter) last year and switched to Bluesky, he has noticed how differently he responds to the two platforms. “Whenever I go to X now I see outrageous shit, and I actually get sucked into it,” says Lorenz-Spreen, who studies how social media and digitalization impact democracy. “Bluesky is slower and more boring—but it is a good boring, you know.”
Lorenz-Spreen, who is at the Dresden University of Technology, is one of hundreds of thousands of researchers who have moved to Bluesky recently. Journals, societies, universities, and research institutes have followed suit, quickly making the platform the de facto meeting place for academics—many of whom seem to enjoy this “good boring.” Now, researchers are starting to study—and debate—how the unique features of this new digital home might affect discussions of research and workplace culture, and efforts to engage with the public. They’re also asking whether Bluesky’s current culture can possibly last.
Although Bluesky looks a lot like X—and started as an internal project in that company—there are important differences. For one, Bluesky’s default timeline simply shows all the posts from the accounts a user follows in chronological order, with the most recent first. That’s a major contrast to the recommendation algorithms that most platforms now rely on, which select content from across the platform to keep users engaged.
The lack of such an algorithm may be why the platform feels less gripping to Lorenz-Spreen and others. But the chronological feed allows smaller accounts to gain more attention than they would if swamped by algorithmically chosen content. “One thing that struck me is the visibility of people with smaller followings and early-career researchers,” says Joe Bak-Coleman, a researcher affiliated with the University of Konstanz. That is particularly true in the feeds Bluesky lets users build themselves and that can include all posts related to a certain research field, for instance. “I’ll see a Ph.D. student or a postdoc posting right alongside a tenured faculty member who has got a book and a giant lab and a lot of funding,” Bak-Coleman says.
The lack of an algorithm could also reduce some of the negative effects of social media, says Jay Van Bavel, a psychologist at New York University. “For instance, it won’t incentivize posts that drive conflict/comments, which were prioritized by the algorithm on X.” That could have a tremendous effect over time, says Mark Carrigan, a digital sociologist at the University of Manchester. “One of the things that’s so destructive about an algorithmic timeline is that the incentive is not to win the esteem and approval of your existing network. It’s to get a reaction,” he says. He hopes moving away from this type of discourse could contribute to improving academic culture.
Bluesky is also significantly smaller, with about 26 million users overall, compared with X’s hundreds of millions of monthly users. Twitter’s size allowed scientists to interact with others such as politicians, journalists, or people working at nongovernmental organizations, says Carl Bergstrom, a researcher at the University of Washington—many of whom are not on Bluesky. “It’s not the same here, not even close.” That could change as Bluesky grows, but it is unlikely to ever become as huge as Twitter, says University of Colorado Boulder information researcher Casey Fiesler. “There’s not going to be a new Twitter. There are going to be a lot of
different things.”
But to Kevin Munger, a computational social scientist at the European University Institute, the sequestered nature of Bluesky could end up being its biggest benefit. Interactions with the broader public on X set up a popularity contest that drove academics toward what social media users valued, he says. “It actually sells short academic expertise.” Bluesky, on the other hand, “is much closer to a specialized place for academics to talk to other academics.”
Bluesky’s design may even foster a different network structure in which individual academics have less power to shape the discourse, resulting in a more democratic platform. In a study of more than 200 million interactions on Twitter from early 2020, a team led by Manlio de Domenico, a physicist and network scientist at the University of Padua, found that information flow about COVID-19 was segregated into different groups on Twitter, each of them highly centralized and driven by just a few accounts. That is driven in part by the fact that X gives larger accounts an advantage, in turn helping them grow even larger.
“Bluesky appeared different,” de Domenico says. At the end of October 2024, he started to gather data on Bluesky’s users through its application programming interface—a tool that makes its data publicly available for free. (Twitter used to offer a similar one.) “The current migration is breaking the typical rich-get-richer effect” he wrote in mid-November when posting his finding—on Bluesky, of course. A debate quickly developed, with other users pointing out that the finding could simply be a result of the network’s rapid growth. That is possible, de Domenico says. But as he has continued to monitor the platform in the months since, “the original finding is still there.”
The data also suggest Bluesky users with a given number of followers tend to follow more accounts than X accounts of a similar size, de Domenico says. That, too, could mean Bluesky is less of a broadcast platform for large accounts. But only time will tell whether that is indeed a lasting feature.
Users may be in the honeymoon phase now, but the experience is likely to get worse over time, Carrigan says—not just because Bluesky is likely to grow further, but because there will be pressure to make money eventually. Ideas including selling subscriptions or services, such as certain domain names, have been discussed. But the company is likely to eventually switch to advertising, he says, creating the same incentive that operates on other platforms: to keep people online as long as possible. “Just because Bluesky is at the earliest phase, before enshittification has started, doesn’t mean it won’t be enshittified.”
