Yesterday Elon Musk spoke in a Twitter Space aimed at mollifying nervous advertisers about his plans for verification on the platform he recently acquired. The hourlong chat between Musk and Robin Wheeler from Twitter’s ad team was wide-ranging — he floated plans for some sort of payment system and explained why car advertisers shouldn’t be worried.
Largely, he focused on explaining his vision for an $8-per-month Twitter Blue subscription and its verification system. “I’m struggling with the question of how do you deal with millions of bots and troll farms, including malicious state actors,” the Chief Twit said.
Wheeler asked him about content moderation and brand safety — specifically hate speech (advertisers, for good reason, don’t want their ads running next to hate speech). Musk said that this would be solved by the $8 paywall. “The propensity of someone to engage in hate speech if they've paid $8 and are risking their account,” Musk said. “Think about it… how much hate speech do you encounter if you're at a party, or at an event?”
Unfortunately, Musk is dead wrong.
He is falling for an old fallacy that if people use their real names online, they won’t say terrible things. Anyone who has ever observed boomers in a local Buy Nothing Facebook group knows that people have no problem saying nasty things under their real names.
Musk’s plan also flies in the face of several academic studies on online behavior. A well-known 2016 study from the University of Zurich showed that using real names actually makes people more likely to post hateful and aggressive comments. One reason for this is that people who were trolling sensed the approval of their peers.
Making people use real names has also been shown to fail at scale. In 2007, in an effort to reduce cyberbullying, South Korea passed a law requiring real names for commenters on large websites. The law was scrapped by 2012, in part because a study showed it only reduced aggressive comments by less than 1%.
“The last thing you want is for a platform to require authenticity because that limits your speech, especially for a marginalized community like queer people.”
Musk’s plan to rely on a paywall to curb hate speech creates another issue: Even if real display names aren’t required, Twitter will be collecting more identifying data about the paid account holders. “One problem with this kind of backend authentication is that it opens up platforms to requests,” Sareeta Amrute, a researcher at Data & Society and associate professor at the New School. “It’s extremely frightening. All types of surveillance can be used against you, and in most cases it’s going to be law enforcement requesting those records.”
Although Musk is not proposing a real name requirement (and yesterday’s rollout of paid verification led to a number of troll impostors anyway), the $8 payment is meant as a half measure of verifying identity — something that can have chilling effects. “The last thing you want is for a platform to require authenticity because that limits your speech, especially for a marginalized community like queer people,” said J.M. Grygiel, assistant professor of communications and social media at Syracuse University. “It’s not going to limit hate speech, but it will limit expression.”
Musk seems to think that hate speech is entirely tied to the bot issue — stemming from someone creating a bot army. This may be partly true — some hate speech is from bots. But he isn’t understanding how many real people are willing to be racist or hateful on main, mask off.
A side effect of selling blue checks to anyone willing to pay is it gives prominence to certain influencers — people who were mostly using their real names already — who were skirting the line of getting banned already, and often incite hate speech from their (also real) followers.
So much of what has happened over the last week at Twitter appears to be the result of someone who has never thought about a social platform inventing it all from the ground up, going through all the mistakes, and then changing tack. (See Mike Masnick’s speedrun of content moderation timeline.) Like reinventing the blue check, then adding the gray “Official” check only to get rid of it hours later.
Musk’s approach to dealing with hate speech seems woefully uninformed. Which is bad for Twitter users, but also bad for Twitter’s bottom line, since a cesspool situation will make advertisers flee. Hopefully, as he is wont to do, Musk will pivot quickly.