it's valid in some sense, it shows that given any n you can write a proof under some reasonable assumptions(eg peano arithmetic or some second order logic), it's not a valid inference under the usual first order logic formulation of the natural numbers.
by the way/ My logic exam is on the 8th...