Why must data scientists treat their conclusions with responsibility and uncertainty?

"Before we dive into the nitty gritty, let me leave you with one more general thought. It’s actually an application of something Spiderman once said: “with great power comes great responsibility.” Here’s the deal. The skills you’ll learn in this book are so powerful and (still!) so rare, that when you demonstrate them, people will think you can walk on water. If you continue in the discipline, you’ll become highly sought-after (and well paid). People will constantly be asking you to work with new data, to produce plots, predictions, and insights, and basically to do your magic. You’ll be treated as a guru: the oracle people go to when they want the scoop. This is ultra-cool, but also dangerous. Why dangerous? One simple reason: because when you make a data-related claim, people will believe you. Pretty much unquestioningly. Most of your colleagues won’t have the expertise or understanding to double-check your snazzy results. And it wouldn’t occur to them to do that anyway – after all, you’re the wizard. The truth of the matter is that data science lives on the knife edge of uncertainty. With our crystal ball, we can make non-obvious assertions about the past or present and even predict the future, but as with all “knowledge,” we must always hold it tentatively. We may be 95% confident that men are paid more than women...but that’s only 95% confidence, not 100%. We may have reason to believe that raising t"