ChatGPT Aced a Remote Viewing Experiment
NickPassig · x · 2026-07-16
A repost claims that ChatGPT 5.6 High "guessed correctly 3 times" in a "remote viewing" experiment.
The original poster stated they tested several hidden object scenarios (a box, a backpack, a hat), and the model seemingly described the concealed objects accurately, even identifying the hat. The overall tone was one of shock and confusion, framing a paranormal experiment as a mind-blowing AI capability.