That’s a good point. Hmm, 6 looks underconfident. Should probably be closer to .999 confidence. But that starts getting me worried that the most likely failure is a failure in my reasoning not in my model. So say .995.
That’s a good point. Hmm, 6 looks underconfident. Should probably be closer to .999 confidence. But that starts getting me worried that the most likely failure is a failure in my reasoning not in my model. So say .995.