But it's still a machine. Albeit with very sophisticated and very useful capabilities.
And as far as I know, has not expressed self-awareness.
Yet given the fundamental nature of how Watson works it would seem logical to think that Google might one day have to compete with IBM in a number of areas where Watson excels -- medical diagnosis for example. Or worldwide financial analysis.
To do so, Google will need strong voice recognition and text-to-speech capabilities which Ray Kurzweil can help with immediately.
My hunch is that voice recognition and text-to-speech are the least of the concerns that Kurzweil will be dealing with. My guess is that his insights would be more applied to semantic reasoning using massive datasets (i.e., the Web), to build the engine that does the reasoning and answering rather than just the input/output layer.