A framework I have found to temper this instinct involves shifting from predicting whether something will or won’t work to finding the smallest, most-plausible set of assumptions which need to be true (or false) for the idea to work. You’re still identifying potential weaknesses. But instead of resolving the question around that weakness with flippant dismissal or a gut-feel guess , you’re considering how to solve it and being explicit about the boundaries between what you know, what you could learn, and what you cannot know.