I’m running a prospective test of a new investigative framework called Forensic Epistemology (FE), designed to improve how humans and commercial AI investigate complex claims. I need a small group of ordinary AI users willing to run a simple controlled A/B test in fresh ChatGPT sessions. You do not need scientific, legal, technical, or epistemology expertise. In fact, please don’t study FE beforehand. You’ll receive: - a frozen FE protocol, - a test packet, - exact copy/paste instructions. Your job is simply to run the assigned prompts without helping or correcting the AI, preserve the responses, and return them under anonymous X/Y labels. This is not a test of you. We are testing whether the protocol measurably changes AI investigative performance when its developers are not present. If you use ChatGPT and are willing to donate a little time to an actual blinded experiment, reply here or message me. Failures are useful data too. We are testing the method, not trying to prove it works. Details are in the classroom under a class called "AI Experiment". Thanks! 🔬🐺