In my estimation it is precisely the other way around. There are hardly any applications of blockchains that are not equally well solved by other methods, whereas computer vision alone has a huge number of applications.