原推:@rohitalmelkar Yes. It should be computationally infeasible to discover x where f(x) can’t be computed, or (x, y) where x != y but f(x) = f(y).
(Or if the construction has *other* defects, eg. z where f_inv(z) can’t be computed, those should also be hard to find)
https://twitter.com/VitalikButerin/status/1516320038629879808