Twitter’s New Developer Rules Might End One of Its Most Enjoyable Parts

Goodbye, good bots.

An illustration of Twitter birds on a conveyor belt.
Illustration by Slate. Images by in8finity/iStock and Twitter.

Memes are more of an art than a science. Some spring up right after a big pop culture moment, like the image (now completely divorced from its original context) of an exasperated LeBron James reacting to a bad pass at this year’s NBA Finals. Others, such as the “Fellas, is it gay?” meme, seem to materialize out of thin air, until they just become part of internet vernacular.

It’s precisely opportunities like the latter that “internet artist” Darius Kazemi waits for. Kazemi built Fellas Bot, an algorithm tied to a Twitter account posting several times a day, prompting us to ponder whether some concept is “gay” or not.

Sometimes Fellas Bot is spot-on, but more often than not, it’s completely nonsensical. And most importantly, it’s funny. “Sometimes you open Twitter and it’s just depressing,” Kazemi says. “I think being on the internet shouldn’t always be that way.”

Fellas Bot is one of 77 bots that Kazemi has unleased onto Twitter to provide his followers with a couple of daily laughs. But the mention of a Twitter bot likely evokes something insidious for the average person, especially after Russian agents used bots to cause discord and spread misinformation during the 2016 presidential election. The concerns those accounts raised about safety and security on the platform have prompted Twitter to make policy changes that will prevent those same misuses from happening again. Earlier this year, Twitter launched a crackdown on bots, deleting millions of accounts. It also updated its policies for app developers (which bot-makers fall under) to prevent them from using automation to schedule, like, or reply to tweets. Many bots ceased to work after those updates. But, a change that came last month pushed Kazemi and other bot-makers over the edge.

Twitter again revamped its procedure for app developers, requiring them to go through a new verification process in order to continue operating. It involves a lengthier application, which includes information about the app’s uses, goals, and monetization. It caps how many accounts someone can run at 10, and if you have more, you need to go through an additional application process. There’s also no specified time frame for how long it takes for an application to be processed, and even after an application is approved, Twitter still reserves the right to do periodic reviews. Someone like Kazemi would have to go through this for each of his 70-plus bots, which is not an abnormal amount for a bot-maker. “I don’t get paid for this,” he says. “They’re pushing us independent artists out.”

Allison Parrish is a bot-maker, computer programmer, and professor at New York University’s Interactive Telecommunications Program. In her work, she’s used artificial intelligence to create poetry, and she rose to prominence after she made a bot that tweeted every single word in the English language, starting in 2007. “The purpose of a Twitter bot, for me, always was to get people to reconsider the nature of the platform, what it’s used for, and what you should be doing on here,” she says.

Parrish says she will not “take any proactive measures” to keep her bots working after Twitter’s new application process takes effect. For her, the process feels like asking Twitter for permission to continue doing what she’s doing, and that’s something she’s not comfortable with. “It’s like skateboarding. It’s an illicit use of the platform. It’s going against what Twitter wants people to do with it,” she says. “It’s like if you took your skateboard downtown or to a park, but you stopped by the mayor’s office to ask for permission first. They’re asking me and people like me to give legitimacy to the way that they operate their business. And I can’t do that.”

For Kazemi, Parrish, and other bot-makers, their mission couldn’t be more antithetical to the Russian trolls or the spammy accounts that are always trying to get you to click on scary links. They want the internet to be a less terrible place. But, with Twitter’s crackdowns, that’s going to be much harder for them to do, and accounts like Fellas Bot are going to be swept up in the process.

Parrish says she’ll continue her work on different platforms. But Kazemi says he’s not ready to let his bots go yet. “A good artist doing an installation in a gallery is going to be thinking about the shape of the gallery, and the color of the walls, and the quality of the lighting, and how many people they expect to show up,” he says. “With a bot, it’s the same kind of thing, but you’re thinking about how people interact with them instead.” He’s exploring different ways to keep them alive, like on an RSS feed or on Mastodon, a social media website popular among artists and developers. “I just wanted to bring a little bit of delight and fun into people’s internet lives,” he says. In a world of Russian bots, that’s a noble mission.

Update, Aug. 9, 2018: After publication, Twitter tweeted a statement in response to its changes affecting bot-makers: