"Hackenburg found that models trained to become more persuasive also ended up being less truthful." It sounds like the arguments were basically AI slop halluci

"Hackenburg found that models trained to become more persuasive also ended up being less truthful." It sounds like the arguments were basically AI slop hallucinations.