Be data-driven, and question the provenance of your data all the time. Otherwise you will end up like economics, a field with prettier models and more mathemati

Be data-driven, and question the provenance of your data all the time. Otherwise you will end up like economics, a field with prettier models and more mathematics than almost every engineering field, and yet gets every major prediction wrong.