ChatTJB: A Human-Powered Chatbot Goes Viral as Commentary on AI Dependency

A Human Replaces the Algorithm

Tucker Bryant, a 32-year-old San Francisco-based artist and former Google project manager, has launched ChatTJB, a chatbot interface with a deliberately disruptive premise: instead of deploying a language model, he personally answers every user query. According to Wired AI, the service received over 30,000 queries by August 6, 2026, just 10 days after a $6,000 billboard in San Francisco announced the concept on July 27. At its operational peak, Bryant reported receiving approximately 5,000 prompts per hour.

The service invites a spectrum of requests—some absurd (a skunk wearing sunglasses), others practical (how to fry an egg), and some personal (advice on birthday celebrations or dinner ideas). Bryant sketches images by hand and provides written responses in conversational tone, often declining to answer or taking time to reply, creating friction that a traditional chatbot would eliminate.

Reframing an “AI” Interface as Genuine Human Interaction

The billboard’s tagline—“the leading chat interface powered by AI” with AI redefined as “average individual”—signals Bryant’s conceptual goal: subverting the industry’s vocabulary while exposing user assumptions. Wired AI notes that Bryant built the ChatTJB website using Lovable, an AI-assisted development tool, underscoring that his critique is not anti-AI absolutism but rather skepticism about unthinking reliance on algorithmic systems.

Early user behavior shifted over time. Bryant observed that novelty-seekers initially dominated the user base, but as engagement deepened, requests became more intimate and sincere. People began asking questions they might pose to a friend rather than a search engine—a behavioral shift that suggests users experience dialogue differently when guaranteed a human is reading and responding.

The Cognitive Surrender Problem

Bryant articulated his artistic motivation as pushback against what he calls “cognitive surrender”—the compulsion to defer judgment to AI systems even when personal knowledge or intuition would suffice. In an interview with Wired AI, he described a telling moment: asking Claude whether to wear short or long sleeves on a 65-degree San Francisco day, despite seven years of living in the city. “And yet, it felt like an easier thing for me to defer to a LLM for advice on that,” Bryant reflected, characterizing the impulse as worth interrogating publicly through art rather than dismissing as mere humor.

Why This Matters

ChatTJB’s viral growth demonstrates latent cultural appetite for human judgment in an era of scaled algorithmic systems. While the project’s sustainability depends on Bryant’s personal availability—a constraint that renders it unsaleable and unmaintainable at scale—it functions as a mirror held up to how dependent users have become on instant, zero-friction recommendations. The 30,000-query throughput, achieved without algorithmic optimization, implies that human-powered interaction carries psychological weight that LLM responses, regardless of quality, cannot fully replicate. For AI product teams, the implicit lesson is worth examining: what friction are we removing, and what cognitive responsibility are users surrendering in exchange?