But the current AI is already very helpful in a lot of applications, none can deny that? I am not sure I get your point, I think its about developing a useful tool and as an unexpected outcome we may create something unforeseen. For me even if we are early it just feels, like more than just simple machine-learning and statistics already.