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

The web site Your AI Slop Bores Me takes its title from a meme individuals on social media use to criticize AI-generated content material. The location — a pretend AI chatbot — has solely been round for a few month. However its creator, Mihir Maroju, mentioned it is already obtained greater than 25 million distinctive guests and almost 280 million whole hits.

“Persons are spending hours on the positioning,” the 17-year-old highschool graduate in Puducherry, India mentioned in an interview with NPR. “I did not actually anticipate it to be so addictive.”

As with actual AI chatbots like Gemini, Claude and ChatGPT, anybody can submit a request for a picture or info by typing it into the youraislopbores.me interface. However on this case, the response does not come from an algorithm — simply one other human.

The enjoyment of taking part in AI chatbot dress-up

A couple of third of U.S. adults have used ChatGPT, in accordance with a June 2025 Pew Analysis research. Persons are not solely deploying AI chatbots for every thing from planning journeys to doing homework assignments — they’re additionally having enjoyable impersonating them.

“Somebody requested me to attract a bat consuming a strawberry,” mentioned San Francisco-based cartoonist Amy Kurzweil, creator of the chatbot-oriented graphic memoir, Synthetic: A Love Story, of her interactions on youraislopbores.me. “That was actually enjoyable.” The location forces its human customers to approximate the pace at which a machine would return a response; there is a 75-second time restrict. So drawings, created with a mouse or finger on a trackpad, have a essentially slapdash look.

Along with responding to queries, Kurzweil mentioned she’s additionally loved asking questions via the positioning. “I requested somebody what they have been studying. They mentioned they have been studying Twisted Hate, however they favored Twisted Video games extra.” (Kurzweil mentioned the change impressed her to look these titles up — they’re a part of a romance sequence by Ana Huang.)

With its old-school Comedian Sans MS font — a staple of internet sites within the late Nineties and early 2000s — the cartoonist mentioned Your AI Slop Bores Me conjures up nostalgia for a time when the Web was, for essentially the most half, a vigorous, pleasant place.

“I do suppose that persons are reaching some extent of frustration with the Web being flooded with non-humans,” Kurzweil mentioned. “So I feel persons are having enjoyable reclaiming a number of the magic of the early Web, only for the little pleasure of connection.”

As a result of the digital panorama has modified quite a bit because the late Nineties, Your AI Slop Bores Me’s directors mentioned they’ve carried out instruments that attempt to flag and filter out dangerous or unlawful contributions. “We had a number of spam and other people exploiting loopholes within the web site,” Maroju mentioned. “Of late, we have not had these points.”

The customers of the positioning know their questions will probably be answered by people. If its URL does not make this clear, the 2 tabs customers can choose from on the homepage — “human” and “larp as ai” (which implies people get to “Stay Motion Function-Play” as AI) — definitely do.

When the consumer does not know it is pretend

However some elements of the AI-bot-dressup universe, akin to Ben Palmer’s model of comedy, function underneath totally different guidelines.

In a deadpan, 2023 skit on YouTube, the Nashville-based comic talks a few pretend ChatGPT web site he arrange not lengthy after the true ChatGPT took off.

“Generally individuals find yourself on the web site considering that they are writing to the precise ChatGPT. However they’re writing to me,” he explains. Palmer goes on to explain his backwards and forwards with a consumer in China — the place the precise ChatGPT has been banned since 2023 — who unwittingly finds themselves on the comic’s pretend model:

“They requested me to write down an article on international local weather change. And I attempted to inform them that this is not the true ChatGPT; it is a joke. And so they wrote again and mentioned, ‘That is no joke.’ And I gave them the deal with to the true ChatGPT, and mentioned, ‘I am too lazy to write down an article.’ And so they mentioned, ‘I would like your assist.'”

Palmer goes on to clarify how he asks the true ChatGPT to write down the requested article, which he then sends on to the consumer. He lastly makes use of AI to translate the textual content, additionally on the consumer’s request, into Chinese language.

In an interview with NPR, Palmer mentioned he arrange a bunch of faux AI textual content and picture technology websites with URLs similar to the names of the true AI web sites. He says some customers would get indignant once they realized they have been being pranked by a human. However others performed alongside. “They’d maintain going as a result of they have been now being entertained,” he mentioned.

The darkish aspect 

The comic mentioned most of his websites have been pulled down from numerous platforms. He admitted there is a darkish aspect to disguising himself as a bot. For instance, he has declined to satisfy requests for sexually specific content material. Palmer mentioned his intention is to remind those who the Web ought to be a messy, vibrant place — not one overrun by soulless firms. “I wish to see how individuals react once they suppose that they are speaking to an AI and it goes off the rails,” he mentioned. “Generally they may shock you.”

“As increasingly individuals embrace AI, it is naturally beginning to present up throughout popular culture,” mentioned ChatGPT maker OpenAI in an electronic mail to NPR. “We love seeing how persons are bringing ChatGPT into their every day lives, and the humor that comes with it’s a part of what makes that so enjoyable.”

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

“People pretending to be AI — that is nice sketch comedy. But it surely doesn’t suggest we’ll use the expertise much less,” Kimmel mentioned. “It simply means we acknowledge that there is a very clear language that is evolving round how we talk with bots that is distinct from how we talk with one another.”