Post #4015654

Since my comments appear to disappear on the blog of Terence Tao: I think that the recent Jacobian counterexample could have been done as a classical optimisation problem, if the search space is bounded. You just look for roots in an anulus away from a root that is kept fixed, and bound the search space. If there is a solution in your search space, and you start optimising from the right place, you basically have to find the solution.