The $800-A-Day Job That Involves Yelling at Computers
Ever found yourself shouting at a chatbot that forgot your name three messages in? That specific frustration might be your ticket to a surprisingly lucrative paycheck. A startup named Memvid is offering $800 for a single day’s work with one simple job description: professionally bully artificial intelligence.
For eight straight hours, you’d converse with various AI assistants. You’d ask them to remember details, watch them fail, and then ask again. Your entire role is to document every frustrating loop, every forgotten context, and every contradictory answer. It’s a stress test for silicon brains, and they’re paying human ones to administer it.
Why AI Chatbots Keep Forgetting Your Conversations
Most AI chatbots impress initially. They answer questions, follow instructions, and seem coherent. But stretch the conversation? That’s when the cracks appear. Details vanish. Context evaporates. The AI might completely ignore a rule you established minutes earlier, acting as if your entire previous chat never happened.
This isn’t a personality flaw; it’s a technical limitation. Many models operate with a limited “context window.” Think of it as the AI’s short-term memory. Once a conversation exceeds that window or resets, earlier information is simply discarded. The model starts fresh, with no recall of your past interactions.
Major players are trying to solve this. Google is adding memory features to Gemini. Anthropic’s Claude attempts to remember conversations across users. Yet, the problem persists for many. Memvid’s entire mission is to build a solution—a persistent memory layer that lets AI models retain important context across different sessions and time. They need to find the flaws before they can fix them.
Qualifications: Frustration, Patience, and a Camera
So, what does it take to become a professional AI critic? The requirements are refreshingly human. No computer science degree or coding expertise is necessary. Memvid is looking for people over 18 with strong opinions about technology. You need the patience to repeat questions endlessly and the genuine frustration to care when the AI gets it wrong.
There’s one more, very modern requirement: you must be comfortable on camera. The entire eight-hour session will be recorded for potential promotional use. The application process itself is telling. It asks candidates to describe their most annoying experience with AI and to argue why they deserve the title of “AI Bully.”
Currently, it’s a one-person gig, paying $100 per hour for remote work. Memvid suggests they may hire more candidates in the future if the initiative proves valuable. It’s a bizarre snapshot of the AI economy—a high-paying job that exists solely because our smartest machines still have the memory of a goldfish.
A Glimpse Into a Weirder AI Future
This unusual job opening highlights a critical, often overlooked weakness in today’s AI. But it also points to stranger horizons. If individual chatbots can’t remember a simple conversation, consider what happens when they start working together. Recent research has shown that AI agents can now team up autonomously, sometimes to spread misinformation, effectively becoming self-running propaganda networks.
The act of bullying an AI for pay feels almost satirical. Yet, it underscores a serious challenge. For AI to become truly useful and trustworthy assistants, they need to stop forgetting. They need to build a continuous sense of history. Until then, there might just be a well-paid human, camera rolling, reminding them of their failures.
