"I know I make mistakes. I have a compiler that tells me so every day."
Of course, that's the trivial case. That's easy and usually easy to fix. Many things are harder.
I try to keep track of the facts. Facts are important, because you've got a much better chance of avoiding mistakes if you're operating from the right set of facts. Part of that is being able to assign the right set of probabilities to the "facts" that you're working with, because some things that once were facts (the continents don't move) turn out to be not-quite-facts (ok, the continents move, but not very quickly on a human time scale).
So what I end up with are little piles of evidence that I keep handy and use to inform my opinions. Some times a piece of evidence turns out to be bogus and is discarded; other times it turns out to be well-founded and ends up being given more weight.
What this means is that I'm seldom certain that I'm right. (That may come as a shock to some of you.) I'll operate based on the best evidence that I've got, but my natural instinct is to be careful about perturbing the system, because you don't know what's going to cascade out the other end of the process when you do. You may think you know what's going to happen, but you could be really, really wrong.
And that doesn't mean that I'm never in favor of doing something, just that I am going to worry about the results. If doing something looks like the right alternative given the available evidence, well then, we may need to give it a try and keep our fingers crossed.
But I live with uncertainty, with the knowledge that I might be wrong, and that I may not ever know everything that I need to know to make me comfortable with making a particular decision.
I suspect that many of you don't feel that way. I'd like to say that I would love to be that certain about things.
But I don't want to be that certain. I don't want to automatically discard evidence that challenges my beliefs, nor do I want to preface my statements with "always" and "never" instead of the -- in my experience -- usually more accurate "usually" and "seldom". And recognizing that I may have the piles of facts misweighted and that -- depending on the exact situation that I'm looking at -- what I thought to be "usually" and "seldom" might be exactly reversed in the real world.
Certainty is overrated.