Why Elon Musk’s Twitter Won’t Be What Fans or Critics Expect

There are signs that the technically minded Mr. Musk is thinking about structural changes to Twitter if his $44 billion acquisition goes through that could have significant and sometimes contradictory impacts. Among them: his declaration that Twitter should make the algorithm that determines what users see open source and therefore more transparent. These changes could reach deep into Twitter’s fundamental infrastructure as well as its policies of self-governance. If they work, they could increase Twitter’s reach, or, if not, diminish it.

In shaping the future of Twitter, Mr. Musk has a range of possibilities to consider. One alternative is being incubated by Twitter itself: a project called Bluesky. It is an independent corporation that Twitter launched in 2019 and has been funding with the goal of developing standards and protocols for social-media services to communicate with one another—and, ultimately, of separating the display of the contents of social media from the data they contain, giving users more flexibility about what they see and with whom they interact.

Farther afield, there is the open-source Twitter alternative Mastodon, first released in 2016, that allows anyone to create their own social network; the weblike standard ActivityPub; and various attempts to create new user-controlled and user-owned social networks using blockchain technology, such as DeSo.

Among the most ambitious possibilities that people outside Twitter have proposed: Twitter could become an open communication protocol like email or the Web—or a proprietary but near-universal protocol like the networks of

Visa

and

Mastercard.

There’s more below on how that could work, but think of this as a future in which Twitter is more a service than a platform. Its contents would be distributed through a variety of other apps, with revenue coming from Twitter selling access to its central repository of data and tweets, both to other companies and, through subscriptions, to individuals.

That version of Twitter could be, depending on who is using it and how it’s accessed, both calmer than it is today and more frenetic; more full of members of one’s political party or less; more rife with harassment or nearly devoid of it.

None of this is happening in a vacuum. Across the tech industry, a movement to change social media has been fomenting for some time. Arguably, it began with critiques by former Big Tech insiders and continued with calls for increased regulation of social media from politicians on both sides of the aisle, especially in the wake of revelations from

Facebook

whistleblower Frances Haugen. In Europe, it’s taken the form of actual regulations—which a top European regulator pointedly reminded Mr. Musk this week he would be required to follow.

Many visions for Twitter’s possible future share certain features. One is that they separate or aim to separate the data that is at the heart of Twitter—all those tweets, as well as the social graph comprising who follows whom—from the apps and services that display that data.

Twitter itself used to work like this. A decade or more ago, before it had a real business model—advertising—Twitter was a repository of tweets and other data that other services could access and display. App stores were filled with Twitter client apps from other companies, like Twitterific and TweetDeck (which Twitter later acquired). Their only connection to Twitter was accessing its databases through an API. That version of Twitter faded when it limited outside developers’ access to Twitter’s pools of data, to control its platform and make sure users saw the ads that quickly became its primary source of revenue.

In a future somewhat resembling that past, Twitter could become the underlying database in which users’ data and content are stored. Other companies and app developers would pay to access this data, giving Twitter a business model divorced from advertising, as independent tech analyst Ben Thompson outlined recently in his Stratechery newsletter. He posited that Twitter could be split in two, with the existing application—what people see in the Twitter app or at twitter.com—a separate entity that would license Twitter’s content and compete against other client apps that did the same, each with differing user interfaces and content policies.

Another potential trait of future Twitter is open-source algorithms. In theory, this lets experts peer into the code that determines what users see, and what is highlighted or suppressed—something all major social-media platforms have so far declined to do. Mr. Musk said in an interview at a TED conference this month that the goal would be to assure users that “there’s no sort of behind-the-scenes manipulation, either algorithmically or manually.”

For a sense of how an open-source Twitter with a separation between content and algorithms might work, it’s useful to look at Mastodon, a free open-source software for Twitter knockoffs that resembles what Twitter co-founder and former CEO

Jack Dorsey

has repeatedly proposed for Twitter.

Hugo Gameiro, a developer based in the small, picturesque Portuguese city of Leiria, started one of the first services built with Mastodon five years ago, called Mastohost, that makes it relatively easy for anyone to create their own, Twitter-like social network using the underlying open-source software that powers the system.

The results have been positive overall, he says, and his service now hosts nearly 800 independent, Mastodon-powered social networks, each a kind of mini-Twitter, ranging in size from a handful of users to as many as 20,000. Mr. Gameiro charges a monthly fee for his hosting services, similar to the business model of countless other managed hosting services on the internet, from WordPress.com to some of

Amazon’s

cloud services.

While Mastodon is tiny compared to Twitter, which claims 217 million so-called monetizable daily users and posted $5.1 billion in revenue last year, its users’ experience with content policies is also instructive for Mr. Musk, should he decide to go through with his promise to allow any expression on the platform that isn’t illegal.

That same stance on acceptable speech was “in my terms of service on day one,” says Mr. Gameiro. Then one of his networks, started in Italy, turned out to be devoting itself to deliberately spreading conspiracy theories and obvious misinformation. Because Mastodon is open source, and administrators of social networks built with it can take all their users and data and move them to a new host, he asked them to leave and they did so.

He has been approached by others who wanted to start networks devoted to discussing matters or sharing material that, while not technically illegal, he wasn’t comfortable hosting. All of these communities can still be built with Mastodon, whether or not he hosts them, because they can always find someone else who will, or set up a server to host it themselves, adds Mr. Gameiro.

At Twitter’s scale, it’s impossible to moderate all content with humans, however, and using AI to filter our social-media feeds is a challenging and imperfect solution. The fact that objectionable material is bound to appear on Twitter, even if not in an individual’s feed, could make Twitter incompatible with advertising as a business model. This could be untenable, since, even if Mr. Musk seeks to reduce Twitter’s reliance on ads, it will need advertising revenue for some time if it’s to pay the interest on the debt Mr. Musk is using to finance his purchase of the company.

The other important lesson Mr. Musk could take from Mastodon is that a transparent, user-controlled alternative to Twitter has existed for years and yet it remains tiny compared to Twitter, which is itself tiny compared to Facebook, Instagram, TikTok and other social-media behemoths, none of which are very transparent or in the least user-controlled.

The history of technology is often about good ideas being tested again and again until the right form of an invention matches the right historical moment. It’s possible that a Musk-led Twitter is just what the world needs to demonstrate that there are technical solutions to messy human problems like the tensions between free speech, profitability and transparency. It’s also possible that the same kind of thinking that yields fast electric cars and affordable rockets will be inadequate to the task of transforming Twitter.

For more WSJ Technology analysis, reviews, advice and headlines, sign up for our weekly newsletter.

Write to Christopher Mims at [email protected]