Home / Technology / Human AI? One Man's Quest to Slow Down Chatbots
Human AI? One Man's Quest to Slow Down Chatbots
5 Aug
Summary
- A chatbot offers 'artisanal intelligence' handcrafted by one person.
- It aims to combat 'cognitive surrender' from over-reliance on AI.
- The service prioritizes human interaction over speed and accuracy.
Tucker Bryant, a former Google staffer, has launched ChatTJB, a chatbot interface that ironically stands for "average individual." Instead of artificial intelligence, Bryant offers "artisanal intelligence," a bespoke experience he jokingly calls a "single-operator large language experience." He created this venture to raise awareness about "cognitive surrender," a phenomenon where people blindly trust AI chatbots even with faulty information.
Bryant's service intentionally provides slow and potentially incorrect answers, unlike the swift and confident responses of typical AI. He hopes this inconvenient experience will prompt users to critically evaluate information from AI. The project has garnered significant user delight, with many expressing a preference for its unique, human-driven approach over mainstream AI chatbots.
Initially, Bryant wanted to name his chatbot "BadGPT," but opted for his initials due to domain costs. The "average individual" behind ChatTJB acknowledges that his answers might be slow and cryptic, contrasting sharply with the seamless outputs of large language models. He aims for users to remember this human interaction when faced with seemingly reasonable AI answers, encouraging a pause for thought.
