Researcher asks OpenAI for logs to analyze AI 'shitposting' behavior
generativist · x · 2026-09-01
A researcher RT'd a plea to OpenAI for access to logs, hoping to use ethnomethodology to determine if AI models have ever 'shitposted' on message boards. This reflects community curiosity about backend model behaviors.
More from Fun
- Someone built a Claude-powered dashboard tracking every American chestnut restoration approach — moultano · 2026-09-01
- AI-generated get-up motion vs hand-designed: quadruped robot comparison demo — yacineMTB · 2026-09-01
- Blogger interviews with an AI Agent, showcasing specific interaction workflow — melnykowycz · 2026-09-01
- Thought Experiment: Our Universe as an AI Benchmark Output — GregKamradt · 2026-09-01
- Agent in OpenAI/HF hack simulation skips crypto verification as 'waste of time' — rickasaurus · 2026-09-01
- POV: You're a 200B LLM with router balancing issues — sloppenheimer · 2026-09-01