So basically he's using the path integral formalism, and since that requires you to specify both the start and end points of the path, he's claiming that physic

So basically he's using the path integral formalism, and since that requires you to specify both the start and end points of the path, he's claiming that physics has to work that way? That seems like a very extravagant claim given that we can compute the same results without using the path integral formalism, and without having to know anything about future end points.