This year, though, two studies conducted by psychologists from MIT and Cornell University suggest that when artificial intelligence presents sufficient counterfactual evidence, believers are more apt to change their minds.
“Science” lmao. This is worthless. You cant properly replicate studies that solely rely on letting people talk to LLMs like chatgpt.
Nevermind that the same thing also works in reverse. All that this proves is that humans can be influenced by reading things. Big news. They were programmed by reading things, no reason why it wouldnt work the other way around.
Can I just turn them off and on again?
Sure, it’s possible but will they?
So long as the “theory” isn’t foundational to a person and the person isn’t made to feel too embarrassed at believing the wrong thing… the odds are about 50/50 ish.