ChatTJB arrives with the shape of a familiar AI chatbot, but its core mechanism is deliberately low-tech: a person reads prompts and replies. Created by 32-year-old Tucker Bryant, a public speaker, artist, and former project manager at Google, the project has become a pointed joke about automation, trust, and the ease of outsourcing small decisions to software.
A San Francisco billboard calls ChatTJB “the leading chat interface powered by AI.” The twist is that, in this project, AI means “average individual.” That substitution is the whole premise: instead of Claude or ChatGPT generating a response, Bryant sits down and answers himself.
How ChatTJB Works
ChatTJB does not use an LLM trained to imitate Bryant’s writing. It is not a model wearing a personal brand. The answer, image, or suggestion comes from Bryant directly, at least when he can get to it.
The requests can be absurd or ordinary. If someone asks for an image of a skunk wearing sunglasses, Bryant may draw it. If someone wants to know how to fry an egg, he may offer a suggestion. WIRED asked ChatTJB how to become a morning person and, at the time of writing, had not received a response. The site notes that “thousands of people are writing in right now.”
That delay is part of what makes the project different from a commercial chatbot. Modern AI tools promise instant availability, fluent answers, and scale. ChatTJB makes the constraint visible: there is a person on the other side, and that person has limited time.
From Art Project to Viral Queue
Bryant says he launched ChatTJB back in April. The project did not take off immediately. According to Bryant, the shift came on July 27, when his $6,000 billboard went up in San Francisco. News coverage followed.
By August 6, Bryant says ChatTJB had received over 30,000 queries. At its busiest, he says he was receiving about 5,000 prompts per hour. Those figures help explain why a service built around one person answering strangers could quickly become overwhelmed.
Bryant initially expected the project to last only a couple of days. After seeing the response, he says he plans to keep it going. He has brought in 10 volunteers, selected from more than 3,000 people who applied.
The economics are intentionally unusual. ChatTJB Pro lets supporters pay $5 per month, but Bryant describes the tier plainly on the website as “this is a worthless tier.” Users receive no special benefit beyond supporting the art. So far, he says, about four people have paid, meaning the billboard cost is far from recovered. Bryant says making money was not the goal.
Why an Average Individual Became the Point
Bryant says he considers ChatTJB “first and foremost” an art project. It is also a commentary on the moment created by AI tools that can answer almost anything in a confident interface.
He is not presenting himself as anti-AI. The ChatTJB website was vibe-coded using Lovable. His concern is more specific: the risk of “cognitive surrender,” or the habit of accepting AI outputs with too little scrutiny.
That concern came from his own experience with LLMs. Bryant says he once asked whether he should wear short or long sleeves on a 65-degree day in San Francisco. He later reflected that he had been living in the city for “like seven years” and already knew how to handle the weather, yet it still felt easier to defer to an LLM.
That is the project’s clearest argument. The danger is not only that AI systems may be wrong. It is also that they can make deferring feel normal, even for small choices where personal judgment would be enough.
What People Ask a Human Chatbot
Many requests began as jokes. Bryant has recently drawn a frog eating spaghetti, San Francisco mayor David Lurie as Godzilla, and two black cats eating grass on a table. The image prompts fit the playful premise: if a chatbot can generate anything, why not ask a person to try?
But Bryant says the written messages have gone beyond gags. Some people ask what to do for dinner or how to use ingredients already in their fridge. Others bring more personal questions, such as not wanting to have a birthday party in two weeks and wondering what they should do.
That shift matters. Once users realize there is room for actual conversation, the interaction changes. Bryant says some messages become more personal, sincere, and genuine.
He does not frame himself as an expert. He says he is not a good artist, and his advice is no better than any other person’s. On YouTube, he described the project as “the worst LLM of all time.” Yet that limitation may be part of its appeal. Bryant interprets the value as people getting a human to bounce questions off, rather than receiving expert guidance.
The Joke Lands Because the Question Is Serious
ChatTJB works because it borrows the language of AI while refusing the central promise of automation. It has a chat interface, a paid tier, a billboard, and a queue of users. But behind the interface is an “average individual,” responding with ordinary attention.
That makes the project both funny and uncomfortable. It asks why a person might consult an LLM about sleeves, dinner, a party, or a drawing request when another person could answer imperfectly, slowly, and openly as a person.
Bryant says he wants people to take a second look at their own relationship to technology, and at the world’s relationship to technology. He also sees the project as a way to add levity while people are surrounded by companies claiming to sell world-changing technology and by frightening coverage of a looming AI bubble.
His closing hope is modest: “I hope that that just makes a weird time sit a little bit lighter,” Bryant says, “because I think sometimes that's the best that we can do.”