This assumes that the impacts of the technology will be positive. And they can be, if we find a way to make it safe, but if we create superintelligent AI before we make it safe it will just kill everyone which is not positive
But most representations of it are probably not very accurate