I think he is wrong: STEM (especially programming) is excellent at proving the person to have failed.

You cannot create software without experience failure a hundred times a day.

Compare this to humanities like literature: how many times a day is a literature teacher proven to have made a mistake? How much real experience does this person have with failure? How many times have your literature teacher admitted "I made a mistake"?

So the STEM person is likely to be confident because s/he has a lot of experience with failure and know how to handle them, and how much they can delay progress.