AI will never replace humans, and human abstraction is unimaginable.
Recently, people in San Francisco bought a billboard advertising a "new big language model" called ChatTJB.
Go to the official website, the interface is modular, a normal chat interface, where you can ask questions, chat, generate images.
The only problem is that the so-called LLM is actually "artificial intelligence, created by a mere human hand".
(AI=Average Individual, huh?
The project's traffic was far beyond expectations, and the real person behind it, who talks to thousands of people every day, has already put his fingers to work.
The demand is so great that new users have had to be suspended.
A real person, raising all the requests
ChatTJB is officially called "the next generation of singles operating big language experiences", and is abbreviated to LLE.
It doesn't have a huge cluster of GPUs, and it doesn't have 100 billion parameters, and all the messages end up being sent to a real person named Tucker.
He read it himself, thought about it himself, and then tapped his two thumbs to respond to a request for "image creation", which he also completed by hand.
Tucker packs a bunch of phrases for this process: "human-driven reasoning", "zero synthesis generation", "certainty delay curve", and "end-to-end explanation".
When you ask a question, Tucker sees it, he thinks about it, and then responds to you when he's awake and in the mood.
The calculation of this set of "models" depends on how long Tucker slept today, the speed of reasoning depends on whether he is in a state, and the maximum concurrency is strictly limited to n = 1.
It is true that "one man's handiwork is the work of one man".
There is no model illusion here, because there is no model. The answer may still be wrong, but every mistake is purely natural, has a date, a trace, and can be accurately traced back to a person with a name.
Don't ask, it's about exercising your critical skills.
Pro members also went online, with the official website explaining:
"A pointless one, just to help me pay off the billboard fees. Unless you really want to help me pay for the billboard, please don't subscribe".
The website even wrote a page specifically for investors, stating that the project "has not yet generated revenue, has not yet entered the hype cycle, and is actively failing to find a product market match".
The cost of training consisted of a library borrowed a book certificate and an outlet account, with a runway expected to last up to 528 months, which is a true long-termism, based on the average life expectancy of American men.
As the project turned red, Tucker decided to turn ChatTJB into a community project and recruit "AI volunteers".
Note that AI here is an abbreviation for Average Individual, which is an abbreviation for Average Individual.
Each volunteer is required to pass a qualification test, and the platform has configured crisis response resources and security mechanisms.
Tucker said it was only a short-term art project and no one deserves to risk it.
At the end of the page, there is a statement in the eighth verse:
ChatTJB is a satirical work and a public art project that does not constitute a real investment opportunity, does not sell any securities or receive funds. All descriptions of financing and capital allocation are part of the work itself.
Let's face it, pretend to be an AI is real.
People who come to ChatTJB are mostly for fun at first.
But Tucker found that some people started to get angry, started talking about what they really meant, and then stayed and continued talking.
We spent a lot of time learning how to communicate effectively with AI.
Models can respond to countless people simultaneously, but human attention cannot be infinitely copied.
A real-life response means that time has been put into it, and that someone else has taken your question seriously. When information and answers are available, this input itself begins to become scarce.
At the other end of the screen, there is a specific person. He reads your words, stops to think, and writes a response himself, a process that is not very efficient and difficult to scale, but retains the most basic meaning of the conversation.
The stronger the technology, the more people-to-people connections should not be seen as inefficient links that need to be eliminated. Sometimes it is enough to know that someone on the other side is listening seriously.
Subtitles
Since ChatTJB is currently unable to experience it, we found a similar platform called "Your AI Slop Bores Me".
This is a "human-sized model" played by thousands of netizens, created by programmer Mihir Maroju and launched in March 2026.
It has two games: "Human", which consumes points to submit text or draw requests. "LARP as AI", which randomly receives requests from strangers, responds personally or draws within specified times, and collects points after completion.
All the output comes from real people.
We had TA paint "a puppy" in front of him, who probably doesn't know Chinese, and directly translated the phrase "Brother, I don't know what you mean".
We changed it to English again, and we threw an abstract painting in front of it.
Bro, your level of painting can't be higher than that of kindergarten.
Finally, please be kind to the AI on the other side of the screen!