The paradigmatic examples of infeasible numbers put forward by strict finitists employ exponential or iterated exponential notations such as n₁^n₂ or n₁^(n₂^n₃).

[53] And from this it follows that such orders cannot be reached from below by a sorties-like sequence of feasible orders of growth \(O(1) \prec O(n) \prec O(n^2) \prec O(n^3) \ldots\) When analyzed according to the Cobham-Edmonds thesis, it hence appears that the ‘naive’ notion of feasible computability does not suffer from the sort of instability which Dummett takes to plague the notion of a feasibly constructible number. These observations point to another theme within the writings of some st