Millions of people are pretending to be AI chatbots — for fun

The website Your AI Slop Bores Me takes its name from a meme people on social media use to criticize AI-generated content. The site — a fake AI chatbot — has only been around for about a month. But its creator, Mihir Maroju, said it's already received more than 25 million unique visitors and nearly 280 million total hits.

"People are spending hours on the site," the 17-year-old high school graduate in Puducherry, India said in an interview with NPR. "I didn't really expect it to be so addictive."

As with real AI chatbots like Gemini, Claude and ChatGPT, anyone can submit a request for an image or information by typing it into the youraislopbores.me interface. But in this case, the response doesn't come from an algorithm — just another human.

The joy of playing AI chatbot dress-up

More than one third of U.S. adults have used ChatGPT, according to a June 2025 Pew Research study. People are not only deploying AI chatbots for everything from planning trips to doing homework assignments — they are also having fun impersonating them.

"Someone asked me to draw a bat eating a strawberry," said San Francisco-based cartoonist Amy Kurzweil, author of the chatbot-oriented graphic memoir, Artificial: A Love Story, of her interactions on youraislopbores.me. "That was really fun." The site forces its human users to approximate the speed at which a machine would return a response; there's a 75-second time limit. So drawings, created with a mouse or finger on a trackpad, have a necessarily slapdash look.

Amy Kurzweil created her drawing of a bat eating a strawberry in response to a request on Your AI Slop Bores Me

(

Amy Kurzweil

)

In addition to responding to queries, Kurzweil said she's also enjoyed asking questions through the site. "I asked someone what they were reading. They said they were reading Twisted Hate, but they liked Twisted Games more." (Kurzweil said the exchange inspired her to look these titles up — they're part of a romance series by Ana Huang.)

With its old-school Comic Sans MS font — a staple of websites in the late 1990s and early 2000s — the cartoonist said Your AI Slop Bores Me inspires nostalgia for a time when the Internet was, for the most part, a lively, friendly place.

"I do think that people are reaching a point of frustration with the Internet being flooded with non-humans," Kurzweil said. "So I think people are having fun reclaiming some of the magic of the early Internet, just for the little joy of connection."

When NPR's Chloe Veltman asked youraislopbores.me a question about pink pompoms, an anonymous human provided a delightful response.

(

Chloe Veltman

/

Screenshot from youraislopbores.me

)

Because the digital landscape has changed a lot since the late 1990s, Your AI Slop Bores Me's administrators said they have implemented tools that try to flag and filter out harmful or illegal contributions. "We had a lot of spam and people exploiting loopholes in the site," Maroju said. "Of late, we haven't had those issues.

The users of the site know their questions will be answered by humans. If its URL doesn't make this clear, the two tabs users can select from on the homepage — "human" and "larp as ai" (which means humans get to "Live Action Role-Play" as AI) — certainly do.

When the user doesn't know it's fake

But some parts of the AI-bot-dressup universe, such as Ben Palmer's brand of comedy, operate under different rules.

In a deadpan, 2023 skit on YouTube, the Nashville-based comedian talks about a fake ChatGPT website he set up not long after the real ChatGPT took off.

"Sometimes people end up on the website thinking that they're writing to the actual ChatGPT. But they're writing to me," he explains. Palmer goes on to describe his back and forth with a user in China — where the actual ChatGPT has been banned since 2023 — who unwittingly finds themselves on the comedian's fake version:

"They asked me to write an article on global climate change. And I tried to tell them that this isn't the real ChatGPT; it's a joke. And they wrote back and said, 'This is no joke.' And I gave them the address to the real ChatGPT, and said, 'I'm too lazy to write an article.' And they said, 'I need your help.'"

Palmer goes on to explain how he asks the real ChatGPT to write the requested article, which he then sends on to the user. He finally uses AI to translate the text, also at the user's request, into Chinese.

In an interview with NPR, Palmer said he set up a bunch of fake AI text and image generation sites with URLs very similar to the names of the real AI websites. He says some users would get angry when they realized they were being pranked by a human. But others played along. "They would keep going because they were now being entertained," he said.

The dark side 

The comedian said most of his sites have been pulled down from various platforms. He admitted there's a dark side to disguising himself as a bot. For example, he has declined to fulfill requests for sexually explicit content.

Palmer said his aim is to remind people that the Internet should be a messy, vibrant place — not one overrun by soulless corporations. "I want to see how people react when they think that they're talking to an AI and it goes off the rails," he said. "Sometimes they might surprise you."

"As more and more people embrace AI, it's naturally starting to show up across pop culture," said ChatGPT maker OpenAI in an email to NPR. "We love seeing how people are bringing ChatGPT into their daily lives, and the humor that comes with it is part of what makes that so fun."

San Francisco-based angel investor Brianne Kimmel, who has backed several AI agent startups, concurs.

"Humans pretending to be AI — that's great sketch comedy. But it doesn't mean we're going to use the technology less," Kimmel said. "It just means we recognize that there's a very clear language that's evolving around how we communicate with bots that's distinct from how we communicate with each other."

Copyright 2026 NPR