Note: this is an essay in progress. I would appreciate feedback from readers on these ideas and arguments.
Maybe we don't all need to quit Facebook and Instagram absolutely. At the beginning of a second Trump Presidency, with the CEOs of the major companies all seeming to tip their hats to the new "King" -- and with a newly bro-ed up Mark Zuckerberg posting his enthusiasm about working with the new administration -- it does seem a little weird to just be going about one's business on these platforms.
What's been happening on Meta is yet another reminder that we don't own our social media engagements; by and large, they own us. And it's not just Meta; Google and Amazon also seem to be shifting gears (getting rid of DEI targets, among other things). Perhaps now might be a good moment for many of us to rethink our dependence on engagement-driving algorithmic media more generally.
Algorithmic media are essentially digital platforms that use algorithms to maximize engagement and time investment in the platform. That would include most social media people use -- TikTok, Instagram, Facebook, X -- but also most streaming video channels like YouTube, Netflix, and Hulu. Manipulative algorithmic media want you to stay on their platforms as long as possible, as your eyeballs guarantee ad revenue and subscriptions.
What's not engagement-driving algorithmic media? Well, BlueSky, for one.
Bluesky, which is currently not owned by impulsive, right-leaning billionaires, is pretty polarizing. There are certainly enthusiastic supporters -- and a fair amount of premature jubilation on the site itself as it's grown over the past year. But there are also skeptics ("didn't we do this with Mastodon a couple of years ago?" -- or as one of my students recently suggested "isn't that site just a lot of millenials patting each other on the back?").
And wait, doesn't BlueSky actually have its own algorithms? To be clear, BlueSky does have algorithmic feeds (the default feed, "Discover," has an algorithmic component). But it also gives you the choice to opt out -- with feeds like "Following" and "Mutuals" that are exactly what they sound like, and that give you the folks you follow in reverse chronological order (i.e., exactly what you signed up to see).
Beyond that, BlueSky allows users to curate their own algorithms -- giving users what CEO Jay Graber calls "algorithmic choice."
So despite the fact that it does have algorithmic elements, I think of BlueSky as a site where we can challenge our passive experience of them, and where their use can be much more transparent. We're not subject to a "master algorithm" that none of us can see or choose.
*
The skeptics argue, reasonably, that the attempt to compete with engagement-driving algorithmic media will fail, and that one big reason will be that without going out of its way to juice engagement, BlueSky is simply not stimulating enough. But we're looking for an alternative, not necessarily a permanent replacement. Also, perhaps a little bit boring might good for us?
Non-algorithmic feeds on BlueSky can feel a bit boring precisely because the only things you typically see if you are on the 'Following' feed are things that your friends actually posted. Sometimes they will have exciting news and brilliant comments. But other times, they are just out living their lives & there may not be something shiny and new every time you log on. That 's healthy.
Engagement-driven algorithmic media like Instagram and TikTok are incredibly good at sucking you in and wasting your time and attention; there's always more to look at, more content to scroll. I have friends and family who readily cop to wasting hours and hours looking at these on their phones, often somewhat apologetically joking about "brain rot." I don't think it's hopeless, in part because most people see what's happening, and simply don't have the energy to find something else to do. I tend to see algorithmic media platforms as bad habits, not addictions. If we can quit them -- or at the very least, pause them from time to time -- we'll all be better off.
Here's why I think we should learn to pause our engagement-driving algorithmic media in a little more depth:
1. Engagement-driven Algorithmic Media has been a driver of large-scale disinformation and far-right content that is undermining democratic norms worldwide.
We know this. This is pretty widespread and commonplace, especially after the past eight years, and maybe I don't need to say very much to support it. Anti-vax conspiracies, election misinformation, Q-Anon, rampant WhatsApp fake news and decontextualized videos (which are especially problematic in India), and all manner of racist and xenophobic fear-mongering percolate through algorithmic social media. Twitter had certain safeguards that aimed to clamp down on this, but Elon Musk wanted to end them to increase his own clout (though he pretended it was about "free speech"), so in 2022 he bought the platform. The others appear to be following suit, with Meta recently announcing they would be ending their in-house fact-checking system (replacing it with crowd-sourced "Community Notes").
I sometimes hear people protest that the version of algorithmic social media they spend their time on is not the right-wing "manosphere." And yes, Elon Musk and Andrew Tate are not the only people on there; there are feminist spaces, Black spaces, queer spaces. Pro-Palestinian content on TikTok was probably a factor in leading the US Congress to try and ban the platform in 2024. But by and large, these are exceptions rather than the rule -- niches rather than the mainstream. (The algorithm on Meta platforms has, users complain, tended to suppress pro-Palestinian content.)
It wasn't always like that. In the 2010s, we were briefly excited about the subversive potential of hashtag activism. Both #blacklivesmatter and #MeToo really exploded as hashtags on Twitter in the mid-2010s, before essentially becoming redundant. The interesting thing about the hashtag moment is that it entailed users creating metadata to find other users -- without the intercession of an algorithm.
As certain topics and keywords seemed primed for virality and as search engines and suggestion algorithms improved, you didn't need hashtags anymore to find what you were looking for -- the platforms were only too willing to guess what you wanted to serve it to you.
If we're waiting for algorithmic media to make our progressive ideas go viral, we'll be waiting a very long time. Maybe we should just give up on the idea of social justice through virality entirely?
2. Engagement-driven Algorithmic Media is Bad For Your Mental Health
If you post something on a non-algorithmic feed on BlueSky, there's a high chance that your followers who are frequent users of the site are going to see it. They may or may not engage with what you say, of course, but there's no longer a concern that the algorithm will penalize you in some secret way some invisible infraction. (Oh no, you linked to another site. Downrank, your post will be buried. Or: oh no, you said the word "Palestine" -- instant downrank.)
Some of us find workarounds to try and game the "master algorithm." ("P**stine") But why? Why do we put so much of our time and effort into living in a system we didn't build and can't control, and that isn't really responsive to our interests?
If a post that would be buried on X on Bluesky, you'll probably get a much more modest response from your followers to any given comment or idea, and it's highly unlikely that anything you'll say will ever go viral there. But that's ok -- it's probably healthier for us to have our thoughts lead to modest engagement with a small number of people, rather than aspirationally broadcasting to thousands via algorithmic intercession.
This constant second-guessing of the algorithm in the pursuit of virality and clout is bad for us. It encourages dishonest posting, self-censorship, and an embarrassing mix of narcissism and self-doubt. It essentially reminds you, every day, that you're there to service the platform's content -- not to support your own interest in self-expression or communication with a network of real-life and digital-only friends.
Forget virality and forget clout. 10 likes for your thoughtful comment is good; 5 likes is plenty. Have your say, then go about your day. Put your phone away; go off and smell the grass and read a book. Don't waste your life thinking about your mentions. (Sorry if that sounds too directive; to be clear, this is a note to self as much as it is instructions to you, dear reader.)
3. Algorithmic Streaming Platforms Make Us Lazy -- And Are Bad for Artists.
The dynamics on streaming video platforms may be a little different -- as of right now, we're not worried about right-wing capture of the platforms or the master algorithms driving them. That said, the engagement-driving algorithms on these platforms are still really problematic.
Spotify has huge power to promote some artists while ignoring others via its playlists. Their algorithms sometimes do weird things, like promoting a particular Pavement song no one had ever paid attention to before because it appeared to be the pinnacle of "Pavement-ness," making it have a massive resurgence both within Spotify and especially on TikTok. (And leading to Pavement's first Gold record.)
In recent months I've started to refuse to listen to Spotify's playlists generated for me, moving to instead just select Albums created by the artists themselves.
And remember that Spotify pays its artists pennies for millions of spins. So I have also been investing in vinyl for albums that are really important to me -- to support the artist, of course, but also to support my sense that I want to have full control over my taste and what I listen to, at least at home. I'm not canceling Spotify, but I'm trying to limit my dependence on it & give myself alternatives.
The same lesson could apply to Netflix and other streaming video. In my view, we're giving these platforms too much of our mental space, giving them too much control over our media engagement.
Let's go back to picking the movies we want to see and seeking them out -- and if necessary, paying to rent them the way we used to do in the old days. Let's try and waste less of our time watching second-rate movies and shows that are utterly and completely algorithmic. Many in-house Netflix films are written and produced with certain algorithmic keywords in mind ("strong female lead"; "gritty"; exciting action), and then served up to us to hit the exact notes we expect them to hit.
As a side note: everyone has been talking (sometimes with chagrin) about a future where our streaming media will actually be written, acted, and produced entirely by AI. I would argue that the prevalence of algorithmically engineered content in our streaming platforms means we're already well on our way towards that. (We may not even notice it when it does go full AI.)
A mediocre Netflix action movie or streaming series is fine, once in a while. But as I've been saying throughout, we'll all be better off if we learn to hit pause on the master algorithm from time to time, so take in real works of art.