Competition for Twitter is good. Threads, thus far, is not.

On July 6th, Facebook (note 1) launched Threads, their new messaging microblogging system built on top of Instagram… in other words, their Twitter killer. Within 24 hours, more than 70 million people – myself included – had signed up for the service, making it the fastest growing social network in history.

But let’s be clear: Facebook had a serious head start. More than 3 billion users around the globe use one of Facebook’s core products – Facebook, Instagram, or WhatsApp – and the company can advertise the new service to any of those users. If you had an Instagram account – as more than 1.5 billion people do – joining Threads just required down downloading an app to your mobile phone, no registration necessary.

(note 2)

Facebook had some giant scaling advantages that startups like Blue Sky and Mastodon lack. Facebook’s servers are built to handle hundreds of millions of simultaneous users. So, an adoption curve that would have destroyed most startups came off with a hitch. More important, perhaps, Instagram already had support for a private social graph, the ability for people to follow some users and not others, which is the tricky part of building a social network.Threads appears to be a reskinned instance of Instagram, which means that we’re not dealing so much with the launch of a new social network, but the launch of a clone: it looks like Twitter, but it’s an Instagram clone under the hood.

There’s a few things to like about Threads. First, at the moment, it has no ads, as opposed to Twitter, which seems more cluttered by promotions each day. Perhaps this is a temporary state while the network is in its early phases, or perhaps it’s a loss leader strategy like Uber ran, offering rides below costs to build up userbase and destroy local taxi industries. Facebook might be willing to lose money by running Threads ad free for quite some time if it knocked Twitter out of the market. And Elon has made it clear that he’s seeking revenue through the most desperate means possible. So let’s approach the ad-free status with a note of caution – it’s likely temporary, possibly a hardball business tactic and is being introduced by a man who’s name is synonymous with surveillance capitalism.

Second, Threads promises to support ActivityPub, the protocol behind open social networks like Mastodon. I am comfortable saying that this would be a good thing. Even if Threads became an enormous Mastodon node with a billion users, it would be a benefit for an ecosystem where adoption has been lower than we might have hoped, in part because of low usability. If Facebook’s good at one thing, it’s usability. And perhaps Threads joining ActivityPub would be a good kick in the pants for the rest of the space.

However, Threads is not yet on ActivityPub, nor have they said when they will be on ActivityPub. Making Instagram’s social graph, with all its accompanying privacy settings, work on ActivityPub is non-trivial, as the techies like to say. It’s certainly possible that the network will remain backended on Instagram for the foreseeable future so long as Zuckerberg and crew are able to claim credit for engaging with the world of open standards.

Third, and perhaps most important, Threads is not Twitter. That’s a serious plus these days, as Twitter is run by a crazy person who seems determined both to turn the site into a far-right partisan network, and to make the tool entirely unusable, perhaps at the same time. A Twitter with more predictability and fewer Nazis would be a very good thing. Many of us, myself included, miss the existence of a “big room” social space where voices could jockey for attention and social movements could take hold. By tipping Twitter in a decidedly partisan direction, Elon has made Twitter a lot less useful and a lot less welcoming.

But before we toast the new non-Twitter, let’s pause for a moment and remember Mark Zuckerberg doesn’t have a a terrific track record in running responsible social networks. In particular, Facebook has been an extremely dangerous and divisive tool, especially outside of the United States. Facebook’s own reports confirm that the network helped the government of Myanmar in a genocidal campaign against its own people.

Facebook may now recognize that it is ill-equipped to fight propaganda, mis-and-disinformation, and incitement to violence in languages that most of its staff do not speak, and where it has not invested heavily in local trust and safety teams. Threads is not yet available in the EU, where it may not comply with EU privacy rules. Perhaps it would have been wiser to open only in a few English-speaking countries, where it is best positioned to moderate content. But even if Threads were only in the US, it is likely to face serious moderation challenges We are heading towards a 2024 US election likely to be riddled with disinformation as it features both Donald Trump and Robert F. Kennedy Jr. Facebook, which fears both being seen as a disseminator of disinformation and criticism from the right for “censoring” speech may have just taken on a moderation challenge it will live to regret.

Still, isn’t a non-Elon, less crazy, relatively Nazi-free social network worth celebrating?

No. No, it’s not. This is the wrong way to build the future of social media.

The lesson we should have taken from Elon’s destruction of Twitter is that online public spaces are too important to hand over to capricious billionaires. They are too important to be supported solely through targeted advertising. We need new models for how these spaces are moderated, governed, and supported. (My lab is exploring one of these models, and we acknowledge and celebrate others as well.)

Mastodon, which offers a blend of smaller, locally governed online spaces and the ability to host “big room” conversations across multiple nodes, has been exploring a model in which local control and local fiscal support replaces a single set of content standards and an advertising model.
My lab at UMass is experimenting with small, community-based social networks and highlighting other topic-specific projects from subreddits, to special purpose networks like An Archive of One’s Own.

This would be a wonderful moment to see a flowering of alternatives to Twitter. But it would be great if those alternatives used a different business model than Twitter’s model, and if they weren’t controlled by a capricious billionaire.

Here are some things not to like about Threads. At the moment, it’s mobile only. That’s a little frustrating for pathetic social media geeks like me, who are cutting and pasting updates to Twitter, Mastodon, BlueSky and now Threads, flipping through our tabs to connect with different audiences on different networks. (This is a problem Gobo hopes to solve, if only Reddit and Twitter would stop screwing with their APIs.) But mobile-only is a real problem for people who use screen readers or other assistive technologies. And mobile-only means it’s very difficult for helper apps like Block Party – which gave Twitter users greatly improved blocking tools, until Elon made their API keys unaffordable – to work with Threads.

Threads without a web interface also makes it much harder for scholars to analyze and study. When we research the spread of mis/disinformation on social media platforms, we use APIs to retrieve large volumes of posts when possible, and “screen scraping” – parsing the HTML a website sends to a web browser – when we don’t have API access. Facebook is legendary for making their HTML hard to scrape, for taking legal action against people who try to scrape data, and for offering only limited information through their APIs. Offering a tool that’s mobile only makes researching even harder, and suggests that transparency isn’t going to be a top priority for the product team.

Second, at present, Threads does not offer a reverse chronological feed. What you get instead is an algorithmically driven feed that combines Instagram influencers trying out the new platform with the people you are actually following. This suggests that Threads will go the way Twitter has been trying to go, realizing that the power of putting content in front of users is one of the main privileges of running these platforms.

Personally, I think any social network should be required both to give you the option of a reverse-cron feed and to allow you to pipe that reverse-cron feed into your own client where you can search and combine social media along your own rules – that’s the point behind the Gobo project. Facebook historically has been one of the worst networks to work with on projects like this – they’re highly resistant to third-party tools interacting with their data. And they’ve got a point: because Facebook works on a private protocol – i.e. your expectation that your post might only be shared with your friends – they are able to argue that these cannot be studied by researchers in the same way that tweets can, and that third parties will have to work hard to comply with Facebook’s privacy standards in building a third party clients. I noted with some dismay that when I signed up for the service yesterday, I had a choice between posting publicly or posting just to my friends. In other words, Facebook is going to extend its privacy graph into this new space.

So we’ve got a new social network that looks a lot like Twitter run by the company that largely invented surveillant ad targeting in social media. It’s got a content algorithm that is currently inescapable. It’s mobile only, making it very hard to modify, use with a helper app or study as a researcher. And it’s run by a company that has a lousy track record of protecting vulnerable users, preventing misinformation and abuse, and listening to community needs.

So no, I’m not celebrating. I don’t want Facebook to build a better Twitter and kill off a possible wave of experimentation in this space. We should have different tools with different models for different communities. But Facebook has rarely played well with others.

If you’re working on Threads and want to prove me wrong, here are some steps you could take.

– Open an API to threads that allows me to build a third party client – I would love for Gobo to support Threads alongside Reddit, Twitter and Mastodon.

– Clearly document how Threads expects researchers and third parties to deal with private posts – will we be allowed to build clients that operate only with public accounts, or with private ones as well?

– Fast track ActivityPub so we can see what adding a huge Facebook node within a Mastodon ecosystem is going to look like.

– Share some information about how Facebook plans to deal with predictable waves of propaganda and mis/disinformation on this new platform, particularly in countries and languages where Facebook does not have a strong presence.

In response to the inevitable comment, “it’s brand new, you can’t expect everything on day one”, I call bullshit. This is a new product from a massive company that should have a better understanding of the problems and limitations of social media than any other business in the world. If Facebook hasn’t considered moderation questions in detail, or have internal answers about prioritizing ActivityPub, that’s a very serious oversight on their part.

My worry is that Facebook has, in essence, accepted Elon Musk’s cage match invitation. (Like most of the rest of you, I would be rooting for the cage in that match.) What I fear is that Facebook has quickly developed a Twitter clone without considering just how much trouble this will get them into in the 2024 election cycle. I worry that all of us who desperately miss Twitter are embracing this new space without thinking about the social media universe we really want: one where we control what we see and read, one where we have a voice in governance. It’s not enough to have a space that is better than Elon’s, as much as we desperately need that.

(1) I’m referring to the large, publicly traded social media company founded and run by Mark Zuckerberg as Facebook and not as Meta. Facebook attempted to rebrand as Meta because growth of the Facebook product has slowed, because Facebook was associated with privacy violations, poor content moderation, mis and disinformation and numerous other scandals. Much as RJ Reynolds wanted to become Altria, becoming “Meta” was not just about embracing a crappy version of a metaverse, but running from some bad history. Given that Threads raises all the questions about content moderation, disinfo, etc., it’s important that we discuss it as a Facebook product.

(2) Yes, this is a screenshot. Why? Because Elon broke tweet embedding. Yes, I know it works today. But he’s broken it once, and I see no reason he won’t do it again.