🔰 Tired millennials batting away recommendation algorithms
Sorry but ‘For You’ has never really been ‘for me’
Hello screen experts, let’s dive straight in shall we:
New social media just dropped; and it’s named ‘Threads’ which is a popular thing people like to make on Twitter, it’s main rival — confusing!
hip, cool, and completely different from everything else out therefull of boomer-friendly memes and celebrities.
Are we getting bored of traditional social media or is it getting bored of trying to impress us?
Threads reveals that recommendation algorithms were doing so little work when compared with the delivery of Actually Good Content
Just so you know, I will not be spending this issue dunking on Threads — that would be too easy and I’m an intellectual, actually. But looking at it for a few minutes and then deleting it (sorry I just really didn’t like it) kind of signalled a few things for me in terms of where we are with this whole ‘social media’ R&D project that we can never seem to escape from.
There is a general sentiment emanating from commentators that traditional newsfeed-style social media is over, because people’s tastes are changing. I wonder if that’s what’s actually happening, or if it’s more that the owners of these platforms have slowly (and now rapidly) eroded them into something we no longer recognise. It’s not that social media users are bored of the format — they are just tired of being dicked around. What’s happening on Bluesky right now perfectly exemplifies this.
Bluesky started giving access to people on the waitlist back in February or March. I have been on the platform for about two weeks. It’s still in beta. I am already seeing people fed up of how things are going, and asking followers to find them on Mastodon instead. This screenshot sums up the frustration pretty nicely:
Seeing as they’re a ‘Twitter clone’, Bluesky have not yet (and may never) install a trust and safety team. Instead they just seem to have two software engineers and a band of volunteers making consequential decisions that they are not equipped to make. This has been especially tough given how hopeful Bluesky’s initial userbase was for the platform. As soon as I logged in, it was obvious that Blueksy was a safe haven for shit posters, trans folk, and sex workers. These are people who left Twitter to rediscover their community in a place that was safe, fun, and free of Elon Musk’s unearthly stench. When I got there, the vibe was very much harmless nonsense, nudes, and the informal understanding that hateful right wing people would not be engaged with, and blocked on site.
The fact that you can choose so easily whether or not you want to see adult content gives Bluesky a content moderation edge that other platforms don’t have; compare this to Instagram, who decided to scrub any nudity from their feed entirely, or Twitter who’ve now gone too far the other way, allowing any old spammy unwanted filth to populate the time line, and you can see why Bluesky has been an attractive offer so far.
Users who came to the platform earlier on in the beta wanted it to work so badly that they took active steps to make it safer and better for themselves. There’s a user called Aveta who joined in April, and used every single invite code she got to bring over 500 black people onto the platform. She also recently posted about a content moderation panel held by Bluesky in which they only invited one black person — who is an expert on fediverse content moderation — and continually de-centred their voice from the conversation, saying that they needed to let “other experienced devs” speak.
This gruesomely unpromising start obviously laid the groundwork for millions of blind-spots: up until literally today (I am writing this on Friday the 14th of July) there has been no mechanism in place to prevent users from creating handles that contain slurs. I’m pretty sure building in this safety feature was not even on the agenda; the only reason it’s actually happening is because the complaints were piling up. There have been accounts with the N-word in in their usernames that have taken weeks for Blueksy to remove. I don’t know the first thing about building a social media platform, but I do know that preventing harmful and offensive usernames is probably something you want to add right at the beginning — not in a scramble during beta testing while everyone yells at you.
The atmosphere on Bluesky now is an uncomfortable mixture of ‘I really thought it was gonna be different this time’ and ‘oh right yeah… this place is run by a bunch of libertarian crypto bros, how could I forget’. There is genuine fear over what’s going to happen when the platform is no longer in private beta. Judging from the wafer-thin trust and safety regime, the only reason why Bluesky is not already infused with hate groups is because the people who got themselves on the waitlist early were the people who were actively running away from those hate groups. As such, invite codes have only been used on like-minded people, and the block button has been used liberally on right-wingers and trolls. So what we’re seeing here is a lot like what happens on Reddit, where communities self-govern content moderation efforts to their tastes. But this has happened in spite of what Bluesky leadership have been doing — once the platform opens up to everyone, it will probably be impossible for communities to adequately protect themselves from hate, unless they are actually given some controls (e.g. hiding or deleting replies to their posts, shared block lists, etc).
Now, on the other side of the grand social media experiment with have Threads, which represents kind of ‘new Facebook’ without politics. If you look at it without following anyone — which is what I did because I was not planning on actually hanging around — it just serves you a horrific vast slurry of polite bullshit. It’s literally a fully-sanitised mantra-sharing love fest made of inoffensive celebrities and memes about being over 30.
The confusing thing about Threads is, the leadership want it to be a place where big important conversations happen, but they also want to leave politics out of it. So, I dunno, people will just end up talking about their favourite characters from the Marvel Cinematic Universe I guess?? Threads isn’t actually social media, it’s just a land-grab for users, and a place for waning IG celebrities to maybe feel relevant again. It’s Twitter, but without the exposure to ‘discourse’ and reputational damage.
There’s a lot of chat about Threads ‘winning’ because it got 30 million sign ups in 24 hours and broke 100 million after a week. This can only be considered a ‘win’ if your goal is simply to scale, and not to build a community. Platform creators are wilfully avoidant of the fact that a lot of people really just want to see a chronological feed of posts from other people they follow. Users are saying that even if you do follow a bunch of people on Threads, your timeline will still eventually become populated by what they algorithm thinks you want to see instead.
There is definitely an appetite for having a feed that helps you discover stuff you never would have seen before, but that only really works if the stuff is actually good. An algorithmically determined feed really doesn’t matter if the content is just a bottomless brunch of positive affirmations and birthday cards for footballers. Just look at TikTok. Technologists will endlessly salivate and cower in fear over the unmatched power of TikTok’s recommendation engine — but is it really the algo doing all the work, or is it that the content on TikTok is just actually good? I entered TikTok in the same way I entered Threads: curious to see what it was like, and following 0 people. I deleted Threads immediately because the content was bad. I still have TikTok because it makes me laugh. Threads experienced an unprecedented spike of early adopters because it’s extremely easy to port your Instagram account over, so it doesn’t feel like ‘starting again’. Furthermore, everyone knew that content moderation would be as tight and regimented as Instagram’s. But, I really don’t know if those two things alone are enough to hold people’s interest.
It’s safe to say that when Twitter switched to an algorithmic feed back in 2016, that was the beginning of the end. No one, including it’s creators, really understood what made Twitter culturally significant back then, but it was painfully obvious that a lot of it was to do with how it was a live real-time feed of what people were thinking and feeling in the moment, either about a significant world wide event, or about nothing in particular. It’s now been morphed beyond recognition and no one likes it anymore.
Bluesky users are understandably hungry for something that resembles what Twitter used to be, but better: a safe place to shit-post, share horny photos, and be softly held by a loving and hilarious community. I share in their tired frustration, because for a while there it truly looked like that’s what it was going to be. But the leadership have made it clear that they’re too libertarian to implement any trust and safety mechanisms. The deafening silence coming from them right now suggests that if some are unhappy with how other users are treating them, they should just deal with it themselves — but they’ve provided no tools to do that in a meaningful way beyond the closed beta.
We’ve now entered a strange new reality where Mark Zuckerberg has reconfigured himself as a happy-go-lucky guy who did not destabalise democracy and incite violence in Myanmar, and therefore Threads seems like the ‘safe’ option. Meanwhile, there are hardcore libertarians on Bluesky making pointed comments at those complaining about the racism, saying that ‘this is the reality of decentralisation’. Not only is that not true, but also in Bluesky’s current state it’s not actually decentralised… it’s just a single node on a protocol that everyone is in together. Ryan Broderick hit the nail on the head when he said, “on an internet full of closed platforms, broken embeds, and crumbling indexes, the last reliable way to share anything is a screenshot”. The reason why we’re not all just keeping personal diaries is because we like to share, even if it’s really janky and annoying.
Ultimately, trying to figure out who will ‘win’ at being a microblogging platform is completely senseless. You cannot ‘win’ at social media. No matter what happens, communities will find a way to connect, support each other, yell at each other, and create movements. Nobody wins if one platform is full of boring rubbish, and the other isn’t safe from Nazis.