Your guess is as good as mine, but I’m just afraid of the “gradually, then suddenly” phenomenon. Right now, we can profess to be in control because at the core of it all, Generative AI is just a sophisticated predictive tool in spite of its impressiveness. But what if developers turn to programme traits like empathy into it, just like how Japan has been doing for decades with its robots? When Generative AI learns to read human emotions and respond to them, who knows when’s the crucial turning point when all the dots connect and it starts thinking according to its will? It will probably happen faster than we all would like.