Why Are My Bluesky Bots So Into Irish

Things haven’t been great on Bluesky lately, where questionable moderation decisions, unpopular changes to Community Guidelines and Terms of Use, and the recent arrival of an official White House and other government accounts are all making it feel not quite like the refuge from Twitter many initially saw it as. For as long as there’s been Bluesky there’s been Bluesky bots, sometimes called “resistance bots” for their bios full of liberal buzzwords and their tendency to repost political content or reply to users’ own political posts in strangely aggressive ways. I don’t get bots on my politics posts very much, but they do come flooding out of the woodwork whenever I post about the Irish language.  

This first happened in late September, when I found out that the Netflix show House of Guinness is the first show on the platform to feature Irish language subtitles. Since I’ve been slowly learning Irish, I thought this was neat, so I posted on Bluesky that I thought it was neat.

This post immediately broke containment, racking up 269 reposts and 2.2k likes, which for an account of my size is a lot of engagement. At first I wondered if there was some huge gaeilgeoirí contingent on Bluesky, but the responses were weird. Many were far more enthusiastic about this than I was, but some of those responses came from accounts with long lists of generic interests or American political slogans in their bios, people I wouldn’t expect to have opinions about the Irish language. Several mistakenly called the language “Gaelic,” which in fairness is not an unusual mistake. A few were weirdly derisive, making fun of the concept of Irish subtitles or spamming weird gobbledygook which just seemed needlessly mean.

After a few days, the buzz around the post died down, and I went on with my life. But then on Saturday morning, I was working my way through a chapter of the book Gaeilge Gan Stró and came across what I thought was a funny, Eugene Ionesco-esque dialogue in which two new coworkers discuss their marital statuses with far more intimacy and passive-aggression than seemed appropriate for an imaginary workplace. As with any funny thing I see in the world, I swiftly posted it to Bluesky.

This post got 99 reposts and 1.6k likes, which again is quite a lot for me, and definitely a lot for a terribly niche topic. Again, the responses ran a weird gamut: more than my House of Guinness post this time seemed real, and it was actually cool to talk about Irish with some folks and explain my very nascent grasp on its grammar. But the post was also appreciated by an account for “refined platinum,” and more people with a host of resistance interests in their bios, and some accounts claiming to be learning Irish that seemed off to me (though in fairness, you wouldn’t think I’d be learning it from my bio either). There was, again, some ire; a self-described anti-Tesla account was mad at me for learning Irish, and a few were just bafflingly mad about the conversation in the book.

Scrolling through the engagement on both these posts, it’s obvious the likes aren’t coming from me being good at social media; as far as I can tell, the posts reached the bot ecosystem and then just pinballed around it racking up likes. But it’s the responses that gnaw at me. You’ll notice I haven’t linked to any of them, though you can find them in the posts. That’s because of what I think is the most insidious part of the Bluesky bots– I’m never 100 percent sure they’re bots. This is by design, of course, but it’s also indicative of my own struggle with the rampant rise of AI. I’m uncomfortable with the tendency to label anything bad or weird or off as “AI,” and I tend to be overly generous with my benefit of the doubt. Maybe some Appalachian grandma who hates Trump and likes to kayak really does have strong feelings about Irish? Thanks to the popularity of the Irish-speaking band Kneecap and their support for Palestine, there could plausibly be American liberal interest in the language I guess. Even as I write this, I’m imagining some real person seeing me suspect they’re a bot and feeling guilty about how bad they might feel, which of course is exactly how the bot creators hope you’ll feel so that you don’t just report them and end… whatever it is they’re doing.

Because that’s the thing that’s killing me: what do the bots want? A bunch of busty women slid into my DMs with their traditional greeting of “hello,” and those were easy to identify as bots and block. I could imagine the likes are to encourage me to post more on a topic by giving it the illusion of popularity, thus shaping the conversation on the platform. That makes sense for health misinformation or politics or crypto, but far less for learning a niche language. I suppose it’s possible there’s some bot-driven conspiracy to grow the Irish language masterminded from some tech-savvy grandfather’s cottage in the gaeltacht, but now I’m just scripting a TG4 show instead of actually getting to the bottom of things. 

But it’s the angry responses that baffle me most. What are they after? In his 2024 video on Bluesky bots, YouTuber Hank Green said, “If there's someone out here on the internet that's making me mad, there is a high probability that that's not even a real person.. What's the incentive? The only incentive is, like, [to] take up my time, distract me, make me have less faith in my society and fellow people.” It’s possible the bots are here just to help users have a bad time, something real people on social media certainly don’t need any help with, and that they have just enough grasp of the Irish language’s place in Ireland’s political and educational landscape to see it as fertile ground for bad feelings and decide they want me to be unhappy about learning it. But if they just want to sow beef and waste my time, why not seize on my posts about US politics, or video games, or being trans? Why not pick something I’m far more likely to have the kind of strong feelings about that would lead me to engage with them, thus getting them closer to their mysterious ends? Why Irish? What kind of mid-tier endgame do their creators imagine that requires me to be mildly befuddled and generally annoyed?

I haven’t replied to any of the bots or suspected bots (apologies if you are a real person I am ignoring because I’m afraid you’re a bot; see graf six of this post), but I have now written a whole blog on their movements, which is arguably worse than just getting into an argument with them. So much of our lives, both online and in real life, just feels bad these days, the rot economy coming for every online platform and streaming service and tech-mediated interaction. Maybe that’s the bots’ real end: to suck just enough to make me feel like everything sucks, even my least consequential of interests. We’ll see how they feel about this post.