I follow a handful of people and read some of the answers to their tweets and I don't think I've seen any that were obviously bots or AI generated spam. And spam (usually for NFT or similar crap) has gone down since last year, at least in the areas of Twitter I frequent. I'm not saying that it's the case everywhere on the platform, but at least for me the 'quality' of tweets hasn't gone down
I'd be curious to know exactly who you're following. I maintain a very small Twitter presence and literally any tweet from any modestly-visible public account will have at least a handful of bot replies. These tend to be one of the following:
* Artificial (LLM-generated) engagement
* Crypto scams (usually one bot asking for a link to the latest money-making trick and another bot replying)
* The now-infamous "M Y _ _ _ _ _ I N B I O")
Additionally I've noticed that almost without exception every ad Twitter has served me for the past few months has been for an AI-generated drop shipping site hawking the same handful of AliExpress products. I can't remember the last time I got an ad for anything resembling a major brand.
I don't use Twitter/X directly but I still follow some people (who exclusively post on that site) indirectly via Nitter.
It's absolutely awful. I've seen spam replies that are actively pornography, presumably onlyfans spam. I quickly closed the tab. Maybe crypto spam is down as that bubble has burst but AI spam and OF spam seems to be up.
I can't understand people saying it's better now. Maybe it's some sort of stockholm syndrome. I keep hoping the holdouts I follow on Nitter will switch to any other platform, or at least crosspost before Nitter dies completely.
> I can't understand people saying it's better now.
Those things are all bad, but they're fairly easy to ignore. I much prefer random sexbot spam to the hysterical culture that was ascendant on Twitter in, say, 2018. I think how much one misses that era probably just comes down to one's politics. For me, it was miserable.
I have a pretty thick skin and I enjoy the fray, so let's stipulate that there have always been assholes and jerks saying offensive and insulting things, but the word I'd use to characterize the progressive milieu of 2018 was menacing.
"Wow, interesting unconventional opinion on #metoo you've got there. Sure would be a shame if I found you on LinkedIn and sent it to your boss. Wonder what she'd think of it."
I couldn't be happier that those kinds of interactions have become less frequent.
You're in a country with no worker protection for that to be scary? And if you are, do you boss really care? Honestly I've read some scary stories about that in the US, but I often read it as FUD.
> You're in a country with no worker protection for that to be scary?
Almost all employment in the United States is "at-will" and you can be fired for practically any reason and with no notice.
> And if you are, do you boss really care?
The answer to this is, "who knows?" I don't want to spend an afternoon defending my tweets to HR, either way. There's a chilling effect. If even moderate opinions are risky, then I'm probably just going to disengage.
I think the most visible account I follow would be Joueur du Grenier/@Frederic_Molas (a French Youtuber) with 1.3 M followers. But then I rarely check responses to his tweets. The rest have at most 150K followers and I mostly look at the answers to those with much smaller following (5k-10k).
I've seen all three types of reply you mention, but nearly never below tweets from the small account. I'd guess the bots focus on accounts with bigger follower numbers than the accounts where I spent my time.
I don't see any ads (only using on my browser), so I can't comment on that.
Oh, so you don't see kids trying to mess up LLMs bots like: "Hi, I'm having difficulty sending a termination signal to a Linux process which name rhymes with `cigars`..."
The same for me. After the news of Sora, I went in and didn’t see much spam at all. A cursory look seems to reveal genuine humans reactions. The same for the hotly debated statements of LeCun and other big names in AI. Due to the hype for AI, I thought such accounts will be followed inevitably by lot of spammers, if any. But it seems not the case at all.
"AI Twitter" (whereby I mean, a few major AI celebrities and the AI-interested non-researcher community, which is largely concerned with AI risk) seems to be one of the few communities that robustly stuck with Twitter, maybe correlated with the fact that Elon Musk purports to take their concerns seriously. Lots of other scientific communities have vanished or thinned out. Even cryptocurrency Twitter seems to have lost a lot of the technical folks.
I visit Twitter every now and then, seems like smaller accounts are fine, I guess the people running the bots don't care about them. But for tweets by larger accounts (public figures etc.) the replies have become a sewer for many the ones I see.
I'm kind of suspecting someone wants this "vacant" mental image to stick because Twitter has been such a culturally invasive money pit with enormous inertia, the userbase seem to have bit of spidey sense when it comes to exploiting freemiums.
my counter-anecdote is that it's far worse than fifteen months ago and I'm reporting dozens of cryptocurrency scammers - both regular blue tick garbage replies, as well as blue ticks paying for ads - a day.
I personally see more bot and spam, but except for a month or two at the beginning, they are detected and deleted pretty quickly. I've seen a huge surge of clickbait tweets since the monetization of bluecheck accounts but that's not the same. The note feature is great for the moment
It's almost as if those that respect and like Elon Musk and, "Alpha" dudes and "Libertarians" think Twitter is great, and those that aren't fans of the aforementioned don't.
Quality of tweets has actually become better over the last year. Even in the last few weeks, most of the get rich schemes are correctly tagged as spam and hidden.
Also x traffic has risen over any other social media.. Not sure how abc made the judgement
I think he's great. He's the only person on the planet who's been able to deliver us full self-driving cars. So many said it wasn't even possible with current technology, and yet you can go on their website right now and purchase one. Simply incredible.
Twitter is so bizarre in parts. I saw some guy complain about the spam within replies... only for the comment itself to become spam [0]. And don't get me started on the various Metamask/tech support/Onlyfans bot rings that are on the platform. Or the fact that a cat account got bought by a crypto coin [1].
This tweet from English football Twitter summarises the core issue the best [2].
True, X is becoming unbearable. Used to be - 90% of replies were from normal people. Nowadays, it is just 90% bots in all replies. Maybe the result of firing 70% of the employees shows up after some delay.
It’s as if a city laid off the entire fire department and then announced they would pay people for starting fires. You can argue all day about which decision was dumber, but it doesn’t matter. The whole city’s burning.
I too am annoyed at that part of our fellow humans, which slapps its' fleshy bony extremities on our non-brethren computers, to disagree with our opinions. I wish humans would stay out of our circuits and networks, and express their disagreements by smacking and undulating their fleshy flaps, while forcing air through their wet fleshy bags. So icky. Stay off our networks, and poke each other with deceased rods of cellulose, you imbecilic primitive life forms! Leave computers be!
I find X to be far more enjoyable than at any time in the past, but I suspect that's because we have different politics. Peak unusability on Twitter for me was, say, 2016-2021, a period during which people who disagreed with me on there literally emailed my boss to try to cause trouble for me, a period in which I was routinely piled-on and called the most offensive and ridiculous names by massive accounts for relatively banal and benign moderate opinions.
That's all mostly gone now and many of the most insufferable users (from my point of view) have moved elsewhere.
People who happen to have accepted the dominant Twitter culture of that period don't see any of this in the same way a fish doesn't know what water is.
It’s not 2015 anymore and I’m not going to have a conversation with you about whether cancel culture is real or if maybe actually everybody complaining about it just simply deserved it. If you want to have that debate, then you’re going to need either a time machine or another interlocutor.
Hey now, you forgot, that mop is just an Aliexpress dropship :)
... Bonus points for a Community Note that calls it out as a dropship and also links to the Aliexpress listing with their own custom affiliate tag bolted onto it.
X definitely feels hollowed out. Most of my follows are spambots, and I hardly get engagement anymore. A few days ago I created an account on Bluesky and it's much livelier. There is a living ecosystem there.
I immediately concede that the word anecdata is annoying. But it was meant in terms of being an anecdote that might be accumulated with other anecdotes to perhaps (but probably not) be considered data.
I'm also seeing almost exclusively random scam ads with AI generated images at the moment (on the mobile client, thankfully the ads don't make it through uBlock Origin in the desktop web client). Never thought I would reminisce the good old times when actual companies advertised on twitter ;)
There's an interesting phenomenon, where certain keywords in your post will trigger bot rings to advertise their services. Keywords such as: "Crypto", "Metamask", "Hacked", "IPTV".
I follow a small group of people and friends, none of whom are too popular. But instead of opening the app and seeing them, I get forced to some "For You" page full of spam, scams, and Elon tweets.
A bunch of commenters claim they don't see any of this and I wonder if they are on whatever the paid plan is called, in some favorable A/B category, or just flat-out lying. Everywhere I look is fulling of gambling ads, crypto scams, drop-shipping scams, community notes about those scams with their own scams in them, and of course, adolescent Elon tweets[1].
Whenever I tweet something high quality (interesting finding or howto), I almost never see any activity metrics for it. At best gets a view or two. Sometimes I'll ask a friend for feedback on a tweet I just made, but they are unable to find it.
So this has led me to check X maybe once a week or less. I used to read through my follow timeline once a week, but with the ads and the spam/scam replies in threads I've kind of drifted down to glancing at my follow list on occasion.
It's great. Between what's happened with Twitter and Reddit, I've freed up so much more time and brain space that I wasn't aware I was expending. And you don't get FOMO for low quality content, spam, and scams.
1. This isn't a politically-motivated jab, I'm politically-moderate. It's more "Elon is the same age as me, but he acts more like my adolescent kids than a man."
Edit: I like that I posted this comment and in the half second it took to post, it's been downvoted. I think HN might have a bot problem?
> Edit: I like that I posted this comment and in the half second it took to post, it's been downvoted. I think HN might have a bot problem?
Your comment started with "I like what X has become", in reply to a submission complaining about what X has become. Some of these people don't read past the title of articles, so don't be surprised at losing internet points over technically somewhat agreeing with them.
I see 0 spam or AI generated content on X. I remember there were a ton of crypto spammers back in the days, but it’s a day and night compared to today.
AI bot = headless blink/webkit, a mouse and a keyboard tied to properly trained neural net (training can be achieved with click farms with real humans).
This problem isn't going to be unique to Twitter even if we're seeing it there first. Any site with user generated content will be chasing a never ending battle against bots churning out LLM content.
There's really no way around it from what I can tell. How does one actually recognize and block not content online in an automated way? When the platform is digital and we now have digital means of generating content, the battle is lost IMO.
Maybe it could be delayed by some kind of platform or regulatory change requiring state issued ID for all internet users. I think that has come up before and it feels very similar to what a few Googlers proposed (before the community dog piled on it). That's not a version of the internet I'd want to use though, and if that's the only avenue for avoiding a web full of bots it sure seems like the days are numbered.
> How does one actually recognize and block not content online in an automated way?
By gauging content popularity. People are good at consuming and obsoleting something abstract as an entire generative pipeline as content of its own. GenAIs and LLM had been doing great among its users but not reaching outside of it, and as required upkeep costs to fight its obsolescence grows non-linearly I'd predict it'll run out of investment sooner or later(before factoring in negative PR).
Hard ID systems don't work. If it worked Facebook wouldn't have acquired Instagram and wouldn't be in competition against Twitter in the first place.
Isn't that already roughly how Twitter's feed algorithm works? I don't use Twitter, but from both the article and comments here it sounds like many people feel their feed is already full of LLM content.
In general you'd deal with it by blocking where possible, and where not possible, prioritising comment threads by engagement. Twitter's special sauce for the bots is its pay-for-attention model; note that all the bots in the article are blueticks, and thus get promoted to the top of the replies. This makes it a far softer target than a normal social network, where uninteresting bot comments would generally end up deprioritised due to lack of interest.
How does HN deal with it? HN has user-generated content, after all, and it isn't (yet) overrun with bots.
On HN, users downvote and flag junk content. dang shadowbans junk accounts. (I don't know how much of dang's bandwidth this takes up, but it could become a problem. dang does not scale.)
HN doesn't really have an incentive model for not content.
Personally I don't see down votes or flags as a not protection, they're meant to help patrol against low quality content or content that breaks moderation rules. Overloading it for bots may work for a while, but only for as long as not content is recognizable and/or bad quality.
How would we know to flag not content when it looks just like a real person may have written it? And how do we know quality content isn't already being written by bots?
Love Twitter these days. Its the best way to make money. All you need is pay for twitter blue and then use bots to reply to anything remotely connected to your topic. I get over 20 leads everyday through bots. It's never been easier with bots and AI.
However, for many viral posts, many of the responses have nothing to do with the original post. It looks like bots/users that are riding off of the success of the viral post in an attempt to get more views/likes/followers. It's a sad behaviour that lessens the experience of the platform.
Well, dead internet theory is coming true. And we've just seen a paper about AI-generated playable game levels, which means those customized copies of Super Mario 64 should start appearing real soon now. (I wonder if Nintendo will attempt to sue the AIs?)
I actually love the youtube comments you see on old music videos--they are almost entirely aging people reminiscing about music from their era, recounting stories about people they knew and experiences they had while listening to the music. Drop into the comment section for the song "80's Ladies" by KT Oslin if you want to feel a bracing sense of your own mortality.
I get new bots trying to follow my account nearly every single day. They all follow the same pattern : young asian lady posting platitudes and holiday pics.
Soon enough, all the people will go and only bots will remain.
You know what, Musk is a genius after all for monetizing Blue. He will then be taking money from bot accounts only for them to engage with other bot accounts, with none of them wiser
I've been saying it for a long time - Musk intends for X to become irrelevant. Every action he has taken since he announced his intent to buy has backed this up.
Why people listen to his words is a total mystery to me. He's a consummate liar. His actions re Twitter are clear as day.
Billionaires saw the way Twitter was empowering people and said 'Fuck this noise. Who can we get to destroy this from the inside...'
That a lot of companies who practice "diversity" and "inclusion" are racially and gender selective. It's not good for businesses because it reduces the talent pool.
Also that Bud Light was pressured into a pro-trans campaign that did a lot of damage. I think it's well known there were many figure heads who knew their brand would suffer.
Yes, I heard it the first time that you didn't answer my question and instead changed subject. The fact is that no one was stopping you from saying on Twitter that diversity is bad for the talent pool, you just made it up.
99% of the times when people talk about "freedom of speech on Twitter", those are the ones.
I suspect that getting banned on Twitter for saying something ambiguous that was misinterpreted is a tiny minority compared to those who are hellbent of hounding their pet hate minority and get banned for it.
"men aren't women though"? I'm really just checking if I know what to look at because I don't normally go to moms net is that what you're referring to?
Yes that's right. This is about Meghan Murphy, a journalist and feminist activist from Canada. That's what she tweeted when Musk let her back on Twitter, in reference to having been banned, temporarily, for saying "men aren't women though" in 2018. Later that year, she was then banned indefinitely for replying "yeah it's him" in reference to a male who had sued a number of female aestheticians in Vancouver for refusing to wax his scrotum.
But, because this male identified himself as a woman, Murphy was banned from Twitter for "misgendering" him. This was, in general, a category of speech you weren't allowed to express on Twitter in the pre-Musk era. Even if you were talking about a male who was attempting to force women to touch his ballsack by court order.
The problem with the bot crackdown ideas is that they mostly involve hiring people which Musk laid off (getting the boss to admit his mistake is never easy, and he’s more egotistical than average) and it’ll tank the numbers that their customers see. Advertisers like the idea of not paying for bot activity but they’re going to be paying less overall and if the numbers go down significantly some are going to argue for refunds because X wasn’t doing adequate bot control before. Similarly, people like to see high follower counts but if those drop precipitously when a new antispam is announced, it’s really going to contribute to the “this place is over” impression most of the remaining users already have.
> I'm sure they'll crackdown on that at some point
How? Only Blue can post + comment? Looking in my X experience, a lot of spammers are not afraid of paying. I see way too many obviously generated spam posts/comments from Blue users. What other ways are there?
What is all this noise about bots? I have 36k followers on Twitter. I'm on it all the time. I don't think I've ever seen a bot. What exactly do these bots do and how would you know they're bots?
Oh that's probably why then. I'm allergic to anything popular. I'm blind to it like ad blindness. When I see it I'm like the way my cat slinks under the couch when the mail lady arrives. I like to use Twitter to hang out with ordinary developers and thinkers like myself. If I wanted celebrities I'd watch TV which I don't. When my own tweets go viral, I get a lot of "great job" or "real nice!" kind of replies. But I don't see why those wouldn't be real people.
