[tim333]

Actually Kurzweil has always estimated the singularity at 2045. It's passing the Turing test he estimated for 2029. Not that he really defines the singularity.