A San Francisco-based chatbot called ChatTJB is answering user questions by having a real person respond instead of using artificial intelligence. Created by 32-year-old Tucker Bryant, the project is described as a blend of art and social experiment. Bryant, a former Google project manager and artist, launched ChatTJB in April but saw significant attention after a $6,000 billboard went up in July. The project has since received over 30,000 queries, with peak traffic reaching about 5,000 prompts per hour. Bryant said the project was initially meant to be a short-term experiment but has since grown due to public interest.
Bryant answers questions himself, whether it’s drawing an image of a skunk wearing sunglasses or giving advice on how to fry an egg. He told WIRED that the project was an opportunity to comment on the current state of technology and human interaction. He does not consider himself anti-AI but is concerned about the risks of 'cognitive surrender,' where people rely too heavily on AI without critical thinking. He reflected on his own use of AI, like asking for advice on weather, and realized he could have relied on his own experience instead.
Bryant emphasized that he is not a skilled artist or expert advisor, and his advice is no better than any other person’s. He described the project as 'the worst LLM of all time' on YouTube, but he believes people value the human interaction it provides. The project has attracted over 3,000 applicants for volunteer roles, though only 10 have been selected. Bryant also launched a 'ChatTJB Pro' tier for supporters, which he calls 'a worthless tier' as it offers no additional features beyond supporting his art. He said he did not create the project to make money but to encourage people to reflect on their relationship with technology.
Source: wired
