From my personal experience, ChatGPT doesn't fail at the fringe either. I would really like reproducible errors because I tend to trust this kind of usage almos

From my personal experience, ChatGPT doesn't fail at the fringe either. I would really like reproducible errors because I tend to trust this kind of usage almost completely