The 'AI' Chatbot That Was Just One Guy

There is a new AI on the block. Only the AI stands for "Average Individual."

A billboard towering over a San Francisco street advertises a service called ChatTJB. The giant ad boasts the "leading chatbot interface powered by AI." Sandwiched between startup billboards and tech bus stop shelters, it looks like any other Silicon Valley launch. But as a small disclaimer — partially obscured by a tree — explains, "AI" here does not mean artificial intelligence. It stands for "average individual."

When you chat with ChatTJB, you are chatting with exactly one human: a man named Tucker Bryant.

The Worst LLM Ever Built

"I just built the worst large language model of all time, and it's about to revolutionize everything," Bryant, an artist and former Google staffer, explained in a social media video. "Picture this: ChatGPT, but on the other end it's just me giving bad advice. Manually."

The project, first reported by the local San Francisco outlet Gazetteer SF, is described on its website as "artisanal intelligence, handcrafted by a single human being." Instead of feeding responses through an LLM, ChatTJB offers what the site calls a "single-operator large language experience" or LLE.

"Rather than sampling from a probabilistic model, each query is resolved through a proprietary biological reasoning substrate, delivering responses that are fully interpretable, non-synthetic, and accountable at the source," the site reads. "In practice: you ask a question, and a person named Tucker reads it, thinks about it, and writes back when awake and motivated."

The deadpan delivery is the point. Bryant wraps his critique in the exact language of AI marketing — "proprietary biological reasoning substrate" is a joke that lands only if you have read one too many press releases about "proprietary neural architectures."

Cognitive Surrender in the Wild

Bryant's motivation traces back to a concept called "cognitive surrender," coined in a study by researchers at the University of Pennsylvania's Wharton Business School. The term describes the tendency of chatbot users to outsource their critical thinking to AI. Users follow confidently stated advice from chatbots even when that advice is demonstrably false.

"We're now so used to confident, fast, and smart-sounding answers from AI that we've largely stopped questioning them," Bryant said. "Some of us are so far gone that the second we drop into that soothing dark interface, see the little thinking animation, and the text coming up letter by letter, critical thinking skills go out the window."

The Wharton research found that this surrender happens across skill levels. Even domain experts in a field will defer to an AI's confidently wrong answer rather than trust their own knowledge. The smoother the interface, the stronger the effect.

ChatTJB is designed as the antidote. Bryant adds friction back into the experience — intentionally, systematically, and by being "infinitely worse" than any real AI.

"My answers will be slow. They probably won't be correct, or maybe they could be, but they'll be very cryptic," said Bryant. "You will have a bad experience with ChatTJB."

Hand-Drawn and Hand-Typed

The satire extends beyond text. When users request image generation, ChatTJB responds with hand-drawn pictures — actual pen-on-paper sketches. The service also accepts a "Pro" subscription to help Bryant pay off the billboard, though a disclaimer clarifies the project is "a work of satire and a public art project," not a real investment opportunity.

In typed responses, Bryant has been known to answer tax questions with haikus about crows. The point is not utility. The point is that the format — the chat interface, the typing animation, the polished confidence of each reply — carries so much persuasive weight that even a deliberately bad experience can make people pause.

"I want you to have such an inconvenient experience with something that aesthetically looks like artificial intelligence that the next time an LLM gives you what sounds like a reasonable answer, you remember the time ChatTJB responded to your tax question with a haiku about crows," Bryant said. "And you become the tiniest bit more likely to pause and remember that just because a smooth-talking chatbot can answer your questions with polish, that doesn't give you an excuse to not think."

What a One-Man Chatbot Reveals

The ChatTJB project works as satire precisely because it is barely distinguishable from a real AI product. The San Francisco billboard, the polished website, the confident marketing copy — these signifiers alone were enough to make it look credible. Someone had to read the fine print to discover the joke.

That is the deeper point. If a chatbot that is literally one guy with a laptop and a sketchbook can pass for AI, then the line between artificial intelligence and performed intelligence is thinner than most people realize. The tech industry's visual language — the dark UI, the streaming text, the "thinking" animation — has become a shortcut that bypasses skepticism.

"The next evolution in artificial intelligence," Bryant added, "is human absurdity."

Sources

• Futurism: New AI Chatbot Turns Out to Be One Overworked Guy Answering All the Messages by Hand
• Gazetteer SF: The new AI billboard locals actually like (original local reporting)
• Ars Technica: "Cognitive surrender" leads AI users to abandon logical thinking, research finds
• Forbes: We Trust AI Over Our Own Brains, Research Finds
• University of Pennsylvania Wharton School: Cognitive surrender research on AI chatbot users and critical thinking