When I wrote TheAPIBook I concentrated a lot on the concept of cluelessness. To be honest, I derived that observation from two sources. The original Martin Rinard presentation and a book by Petr Vopěnka analysing why we believe our geometry is correct. Of course with such preparations, I watched with a great enthusiasm following Nathan Marz presentation discussing why we believe our software is correct. It is the best keynote I have ever seen I and recommend it to everyone who wants to develop good software:
I believe Nathan's conclusions are similar to mine. Especially the Chapter 11's path of a warrior leads to the same conclusions. The only certainty about a software we produce is the part that we (automatically) test. The more certainty we want, the more (automatic) tests we shall produce. So much about truth...