The movie is not quite as thought-provoking as the original story deserves, but it is much, much, better than the trailers made it appear. With the exception of a few scenes which thankfully do not dominate the movie, Bicentennial Man is not the Disneyish comedy that the lousy marketing made it seem. It is an often touching film with much affection for Asimov's tale.
Bicentennial Man tells the story of a robot who over the course of two hundred years endeavors to become a human being, trying everything from replacing his metallic body with flesh to humanizing his behaviour. Robin Williams gives an affecting performance. The script features some occasionally annoying humor, but the writing contains some notable strengths as well. The foreshadowing is subtle and well-placed, and the ending which could have been merely tragic feels triumphant. This movie deserves to be more successful than it probably will be thanks to the marketing.
Yet neither this movie nor Asimov's story addresses what I feel to be the real difference between humans and machines: conciousness. Of course, in both versions we are shown everything from Andrew-the-Robot's viewpoint, so we as an audience know he is conscious. However, it bothers me that none of his human associates tries to examine the question of whether he is conscious and use that as a basis for deciding whether to consider him human. To me, a robot that merely acts like it is thinking can be treated justly as a kitchen appliance, but robot that is actually aware should be treated as a living being.
On the other hand, perhaps there could be no way to determine whether Andrew was truly aware or simply acting as though he were. Personally, I consider the Turing Test highly untrustworthy.