Twitter was, for the first few years, a place where all the internet's different cultures could cross paths. It was where all the Venn diagrams of various cultures collided into one amazing town...
Twitter was, for the first few years, a place where all the internet's different cultures could cross paths. It was where all the Venn diagrams of various cultures collided into one amazing town square.
And then it spilled over into the mainstream and millions more people poured into the town square until it devolved into a place where no real conversation could happen. The town square was literally full of people just yelling at one another. It just turned into one giant scream that made it impossible to get anything of value out of the site.
Oh, and it also turned into the Nazi bar. And I think that, sadly, will be Twitter's legacy. A place that welcomed evil people with open arms.
Are bots and astroturfing campaigns the reason it turned to shit or is that just a convenient scapegoat? I’m a bit terrified that with much of today’s social media, we just look into a social...
Are bots and astroturfing campaigns the reason it turned to shit or is that just a convenient scapegoat? I’m a bit terrified that with much of today’s social media, we just look into a social mirror and that’s just the true nature of people. But I hope genuine abuse is the only problem because that can at least be fixed in theory.
I see this sentiment a lot and really feel like I need to push back against it when I can. Social media is engineered to alter how you act to get as much engagement out of you as possible, and...
a bit terrified that with much of today’s social media, we just look into a social mirror and that’s just the true nature of people
I see this sentiment a lot and really feel like I need to push back against it when I can. Social media is engineered to alter how you act to get as much engagement out of you as possible, and conflict is the most reliable way to engage people. It amplifies juicy rumours, it amplifies the worst opinions that are easiest to get mad at. YouTube amplifies conspiracy theories because it's built to maximise viewing metrics and nobody watches more YouTube than someone who's into five different conspiracy theories.
I would compare people's behaviour on social media to behaviour on alcohol or drugs. Some get more combative than normal, some are just goofy and less inhibited, and a small percentage of people are either predisposed to severe addiction or just vulnerable due to circumstances and frequently show very antisocial behaviour.
Honestly I think if a not-for-profit platform like Mastodon had the tech to scale to the size of Twitter, you wouldn't see anywhere near the same level of antisocial behaviour.
I don't think it's solely a money thing. Because the large scale ones are good at gaming things out to be addictive, they will always be where most of the activity is.
Honestly I think if a not-for-profit platform like Mastodon had the tech to scale to the size of Twitter, you wouldn't see anywhere near the same level of antisocial behaviour.
I don't think it's solely a money thing. Because the large scale ones are good at gaming things out to be addictive, they will always be where most of the activity is.
I'm also skeptical about "the true nature of people." It's hard to say what that is, since people adapt to their environments. But I have to push back on the "engineered" part. People getting into...
I'm also skeptical about "the true nature of people." It's hard to say what that is, since people adapt to their environments.
But I have to push back on the "engineered" part. People getting into online fights predates social media. Similar arguments happened on Usenet, mailing lists, and BBS's. The bad parts of social media will likely happen at scale unless the system is engineered to suppress them. And it's definitely happening on Mastodon - there are people I've unsubscribed from because they're doing the same old Twitter things.
I think gossip might be closest to being "human nature." It happens in most communities, and it's "amplified" by many people because they enjoy it. Writing down gossip and publishing widely can itself can have weird effects. Newspapers have been doing this for centuries, but now people can do it themselves.
There are also behaviors that only a small percentage of people do and yet seem to inevitably happen in cities around the world. Are pickpockets human nature?
The difference between a Usenet and a Twitter as I see it, is that we now have an attention economy, which creates a financial incentive for platforms to be able to demonstrate as much activity on...
The difference between a Usenet and a Twitter as I see it, is that we now have an attention economy, which creates a financial incentive for platforms to be able to demonstrate as much activity on their site as possible. So they serve recommended tweets chosen to get as much attention as possible from you, and whether that involves juicy gossip or blinding indignant rage, the algorithm doesn't know or care. Simply put, conflict is a good way to get their metrics up, so it happens more. Suppressing certain bad behaviours, like brigading via quote tweets, is not good for business, so it doesn't happen. Network effects will amplify gossip no matter where you are, but an algorithm full of recommended content is another beast entirely, one that can serve you infuriating takes that aren't even getting attention yet, simply because they're similar to previous tweets that caused three days of bitter arguing.
To use your analogy, in a normal city a small percentage of people react to incentives in their environment and are driven to pickpocket. The city takes action to prevent and discourage this, but no solution is perfect and so some pickpocketing happens. Can't fault them for that, they're putting in as much effort as as could be expected.
If instead of that the mayor profits from a high pickpocketing rate (he owns a private prison nearby), doesn't suppress it and in fact has had policy in place for years that every worker gets paid in cash each week, making them easy targets, then I would say the pickpocketing rate is artificially heightened - the environment that people adapt their behaviour to has been shaped by someone who profits from pickpocketing. That means preventable pickpocketing is happening, and people are driven to pickpocketing who wouldn't have even tried under a more reasonable set of policies.
I do sometimes see bad posts appear in my feed by people I'm not following. (I use the realtwitter redirect to avoid it on Twitter most of the time.) Yes, algorithms can boost bad stuff. I'm not...
I do sometimes see bad posts appear in my feed by people I'm not following. (I use the realtwitter redirect to avoid it on Twitter most of the time.) Yes, algorithms can boost bad stuff. I'm not sure we've seen good evidence that this is intentional, though? It's usually assumed without evidence.
I think a lot of it is people resharing stuff they hate, but feel like they need to comment on. I have friends whose feeds I can't stay subscribed to for long, even though they're otherwise good people, because they like to reshare stuff to trash it.
Also, Mastodon doesn't have quote-tweets, but sometimes people take screenshots of posts they dislike and then it gets boosted, so it has the same effect.
It can also happen indirectly via blogs and news site opinion pieces about things happening on social media sites. Then you don't need quote-tweets or screenshots, a link will do.
Bots and astroturfing are an outcome of the culture becoming noxious. One of the challenging things with moderating out bots for just being bots is that the behaviors of bots or paid trolls is...
Bots and astroturfing are an outcome of the culture becoming noxious. One of the challenging things with moderating out bots for just being bots is that the behaviors of bots or paid trolls is largely indistinguishable from the behaviors of a specific clade of noxious internet users.
It used to be that it was a well known rule in the internet that you never read the comments on a news article because it would be dominated by just the most hateful, angry, unhinged people in the world going there to spout nonsense at a functionally captive (as in they can't tell you to shut up) audience. When social media ate everyone else's websites as portals to content, that specific type of internet user metastasized and became the dominant vote on the platforms. The hateful people individually don't do much, but the upvote/boost dynamics let them functionally act as "assignment editors" for the content stream, elevating the profiles of the sorts of posters who appeal to them. These basically end up being more polished and articulate versions of the same noxious newspaper comment guy.
Into this system come the bots. Because newspaper comment guy is such a predictable, sheep of a person without their own ideas or opinions, they mostly interact in fairly robotic ways. So it's easy to mimic their behaviors with literal robots or random people working from a script. This, in turn, makes it hard to moderate out bots. You have to discourage that kind of participant from being active in your platform at all, because the bots rely on them providing too much cover for themselves to sneak around unnoticed.
This isn't necessarily a partisan or ideological thing. This sort of dynamic happens across the political spectrum, but one specific spoke on the political compass is so heavily astroturfed (because big-money has an interest in propping it up) that any attempt to address the problem will basically stamp that sector out. Other parts of the political spectrum would be impacted in ways that are more mixed, it would probably change the internal dynamics of these movements rather than scraping them away entirely.
I'm a big believer in the dead internet theory because of bots and astroturfing. I think a lot of users are not legitimate and it's actually rare to have real discussions online these days. Try...
I'm a big believer in the dead internet theory because of bots and astroturfing. I think a lot of users are not legitimate and it's actually rare to have real discussions online these days.
Try using any online dating platform in 2023 like... Okcupid. Tell me how many legit users you match with.
Twitter have been historically inconsistent when it's come to stamping evil out. Ban Trump but allow the Taliban to remain? Curb hate speech but welcome views which outright glorify pedophilia?...
Twitter have been historically inconsistent when it's come to stamping evil out. Ban Trump but allow the Taliban to remain? Curb hate speech but welcome views which outright glorify pedophilia? Twitter has a growing MAP ("minor attracted person", their terminology, not mine) community that has caught the ire of mainstream news outlets for a while now...
Sure, Musk may be turning Twitter into a Nazi bar, but the site was a borderline Pedo bar before this. Not even Reddit or 4chan allowed some of the stuff you saw posted on Twitter just two years ago.
Last year I read a fantastic book called Digital Minimalism, which spelled out a lot of the ill effects social media has on us and advises how to cut your personal usage down to healthy but...
Last year I read a fantastic book called Digital Minimalism, which spelled out a lot of the ill effects social media has on us and advises how to cut your personal usage down to healthy but non-luddite levels. I never liked Twitter to begin with but the book brought me to the conclusion that Twitter needed to die for the good of society. Musk bought it like two months later, which gave me a lot of hope I would get my wish. So far I think it's going nicely.
Thanks for posting this NaraVara and thanks for posting it in a way that made it simple to read without needing a paywall bypass. I remember a few months ago I DM’d you asking if you felt like...
Thanks for posting this NaraVara and thanks for posting it in a way that made it simple to read without needing a paywall bypass.
I remember a few months ago I DM’d you asking if you felt like Twitter had lost its cultural significance. So it’s nice to have a whole article written about exactly that.
I was thinking about it and was wondering, if Trump was not a Twitter addict would Twitter have been at the forefront of cultural stuff? Like yeah a lot of journalists used it before trump and whatnot. But it seems like it really became more of a thing with the 2016 election. Like what if Trump had been addicted to a different social media, would that have been the cultural force Twitter was in the late 2010s.
Twitter was, for the first few years, a place where all the internet's different cultures could cross paths. It was where all the Venn diagrams of various cultures collided into one amazing town square.
And then it spilled over into the mainstream and millions more people poured into the town square until it devolved into a place where no real conversation could happen. The town square was literally full of people just yelling at one another. It just turned into one giant scream that made it impossible to get anything of value out of the site.
Oh, and it also turned into the Nazi bar. And I think that, sadly, will be Twitter's legacy. A place that welcomed evil people with open arms.
Are bots and astroturfing campaigns the reason it turned to shit or is that just a convenient scapegoat? I’m a bit terrified that with much of today’s social media, we just look into a social mirror and that’s just the true nature of people. But I hope genuine abuse is the only problem because that can at least be fixed in theory.
I see this sentiment a lot and really feel like I need to push back against it when I can. Social media is engineered to alter how you act to get as much engagement out of you as possible, and conflict is the most reliable way to engage people. It amplifies juicy rumours, it amplifies the worst opinions that are easiest to get mad at. YouTube amplifies conspiracy theories because it's built to maximise viewing metrics and nobody watches more YouTube than someone who's into five different conspiracy theories.
I would compare people's behaviour on social media to behaviour on alcohol or drugs. Some get more combative than normal, some are just goofy and less inhibited, and a small percentage of people are either predisposed to severe addiction or just vulnerable due to circumstances and frequently show very antisocial behaviour.
Honestly I think if a not-for-profit platform like Mastodon had the tech to scale to the size of Twitter, you wouldn't see anywhere near the same level of antisocial behaviour.
I don't think it's solely a money thing. Because the large scale ones are good at gaming things out to be addictive, they will always be where most of the activity is.
I'm also skeptical about "the true nature of people." It's hard to say what that is, since people adapt to their environments.
But I have to push back on the "engineered" part. People getting into online fights predates social media. Similar arguments happened on Usenet, mailing lists, and BBS's. The bad parts of social media will likely happen at scale unless the system is engineered to suppress them. And it's definitely happening on Mastodon - there are people I've unsubscribed from because they're doing the same old Twitter things.
I think gossip might be closest to being "human nature." It happens in most communities, and it's "amplified" by many people because they enjoy it. Writing down gossip and publishing widely can itself can have weird effects. Newspapers have been doing this for centuries, but now people can do it themselves.
There are also behaviors that only a small percentage of people do and yet seem to inevitably happen in cities around the world. Are pickpockets human nature?
The difference between a Usenet and a Twitter as I see it, is that we now have an attention economy, which creates a financial incentive for platforms to be able to demonstrate as much activity on their site as possible. So they serve recommended tweets chosen to get as much attention as possible from you, and whether that involves juicy gossip or blinding indignant rage, the algorithm doesn't know or care. Simply put, conflict is a good way to get their metrics up, so it happens more. Suppressing certain bad behaviours, like brigading via quote tweets, is not good for business, so it doesn't happen. Network effects will amplify gossip no matter where you are, but an algorithm full of recommended content is another beast entirely, one that can serve you infuriating takes that aren't even getting attention yet, simply because they're similar to previous tweets that caused three days of bitter arguing.
To use your analogy, in a normal city a small percentage of people react to incentives in their environment and are driven to pickpocket. The city takes action to prevent and discourage this, but no solution is perfect and so some pickpocketing happens. Can't fault them for that, they're putting in as much effort as as could be expected.
If instead of that the mayor profits from a high pickpocketing rate (he owns a private prison nearby), doesn't suppress it and in fact has had policy in place for years that every worker gets paid in cash each week, making them easy targets, then I would say the pickpocketing rate is artificially heightened - the environment that people adapt their behaviour to has been shaped by someone who profits from pickpocketing. That means preventable pickpocketing is happening, and people are driven to pickpocketing who wouldn't have even tried under a more reasonable set of policies.
I do sometimes see bad posts appear in my feed by people I'm not following. (I use the realtwitter redirect to avoid it on Twitter most of the time.) Yes, algorithms can boost bad stuff. I'm not sure we've seen good evidence that this is intentional, though? It's usually assumed without evidence.
I think a lot of it is people resharing stuff they hate, but feel like they need to comment on. I have friends whose feeds I can't stay subscribed to for long, even though they're otherwise good people, because they like to reshare stuff to trash it.
Also, Mastodon doesn't have quote-tweets, but sometimes people take screenshots of posts they dislike and then it gets boosted, so it has the same effect.
It can also happen indirectly via blogs and news site opinion pieces about things happening on social media sites. Then you don't need quote-tweets or screenshots, a link will do.
Bots and astroturfing are an outcome of the culture becoming noxious. One of the challenging things with moderating out bots for just being bots is that the behaviors of bots or paid trolls is largely indistinguishable from the behaviors of a specific clade of noxious internet users.
It used to be that it was a well known rule in the internet that you never read the comments on a news article because it would be dominated by just the most hateful, angry, unhinged people in the world going there to spout nonsense at a functionally captive (as in they can't tell you to shut up) audience. When social media ate everyone else's websites as portals to content, that specific type of internet user metastasized and became the dominant vote on the platforms. The hateful people individually don't do much, but the upvote/boost dynamics let them functionally act as "assignment editors" for the content stream, elevating the profiles of the sorts of posters who appeal to them. These basically end up being more polished and articulate versions of the same noxious newspaper comment guy.
Into this system come the bots. Because newspaper comment guy is such a predictable, sheep of a person without their own ideas or opinions, they mostly interact in fairly robotic ways. So it's easy to mimic their behaviors with literal robots or random people working from a script. This, in turn, makes it hard to moderate out bots. You have to discourage that kind of participant from being active in your platform at all, because the bots rely on them providing too much cover for themselves to sneak around unnoticed.
This isn't necessarily a partisan or ideological thing. This sort of dynamic happens across the political spectrum, but one specific spoke on the political compass is so heavily astroturfed (because big-money has an interest in propping it up) that any attempt to address the problem will basically stamp that sector out. Other parts of the political spectrum would be impacted in ways that are more mixed, it would probably change the internal dynamics of these movements rather than scraping them away entirely.
I'm a big believer in the dead internet theory because of bots and astroturfing. I think a lot of users are not legitimate and it's actually rare to have real discussions online these days.
Try using any online dating platform in 2023 like... Okcupid. Tell me how many legit users you match with.
Twitter have been historically inconsistent when it's come to stamping evil out. Ban Trump but allow the Taliban to remain? Curb hate speech but welcome views which outright glorify pedophilia? Twitter has a growing MAP ("minor attracted person", their terminology, not mine) community that has caught the ire of mainstream news outlets for a while now...
Sure, Musk may be turning Twitter into a Nazi bar, but the site was a borderline Pedo bar before this. Not even Reddit or 4chan allowed some of the stuff you saw posted on Twitter just two years ago.
Last year I read a fantastic book called Digital Minimalism, which spelled out a lot of the ill effects social media has on us and advises how to cut your personal usage down to healthy but non-luddite levels. I never liked Twitter to begin with but the book brought me to the conclusion that Twitter needed to die for the good of society. Musk bought it like two months later, which gave me a lot of hope I would get my wish. So far I think it's going nicely.
Thanks for posting this NaraVara and thanks for posting it in a way that made it simple to read without needing a paywall bypass.
I remember a few months ago I DM’d you asking if you felt like Twitter had lost its cultural significance. So it’s nice to have a whole article written about exactly that.
I was thinking about it and was wondering, if Trump was not a Twitter addict would Twitter have been at the forefront of cultural stuff? Like yeah a lot of journalists used it before trump and whatnot. But it seems like it really became more of a thing with the 2016 election. Like what if Trump had been addicted to a different social media, would that have been the cultural force Twitter was in the late 2010s.
I'm glad it's appreciated. I was feeling self-conscious about the URL-gore I was subjecting everyone to haha
Great article that is a rumination on Twitter, what it was, and what impacts it has had on "discourse."