I think what Khaneman was getting at is that it's hopeless futile to try and recognize your own biases in a meaningful. So why I understand your statement, I'm skeptical that's it's possible if someone at the forefront of cognitive science admits he can't do it. I'm still enough of an optimist to try and use mindfulness myself to accomplish that though.
Of course, maybe he's just really bad at being mindful ;-)
