Knowledge Requires Testing

The idea of knowing what is true is closely related to the idea of being able to make accurate predictions. This is a topic Nate Silver discusses at length in "The Signal and the Noise: Why So Many Predictions Fail-but Some Don't"

One of the points he makes is that while some predictive sciences get better over time, others don't. He concludes that to get better at prediction, we need to make lots of predictions that we then test against what actually happens.

In my own life, I started writing computer code about 40 years ago, and I notice that I still make the same mistakes and typos I did when I was in high school. What saves me now is that we have much better testing facilities and I've been able to discipline myself to use them. I still make mistakes, but I design processes specifically to expose the weaknesses in my code.

If we want to have an accurate picture of the world -- a shared view of reality that we can use to make decisions about society -- we have to get used to asking those questions and seeking answers that are based on something other than depth of feeling.