Please cite that as the execrable film so it's not confused with the incredible stories by Asimov. 

Updated.  Though I rather like the movie.  Perhaps because I didn't read Asimov.

Count me in as one of the people who has concerns about the rapid development and ubiquity of AI. I don't envision doomsday scenarios like The Matrix or Terminator, with evil machines rising up to take over the world. But I do predict that in the not-too-distant future, in the same way that advances in medicine and biotechnology have resulted in the field of bioethics, advances in AI will result in discussions regarding machine ethics. Consider the following:

1) Between social networks, search engines histories, e-commerce sites, and various "free" websites that provide a service in exchange for your personal information, there's a lot of information about each of us floating around out there. Now suppose an artificial intelligence has access to all this information. Can we trust it to use that information responsibly, or in a secure and private manner?

2) AIs will soon be placed in areas where decisions involving human safety will be made -- autonomous vehicles being the most imminent application. Supposing there are emergency situations where loss of life, or at least severe injury, will occur, do we trust machines to make decisions for us as to what is the "best" outcome? If either the driver or a pedestrian's life is threatened, whose life does the AI assign greater priority?

3) As machines become more and more lifelike, it's inevitable that the sex and pornography trades will develop artificial sex workers. Does this cross the line from porn to prostitution? And if so, what kinds of laws will govern it? While it's arguable as to whether human prostitution is a victimless crime (I would say in most instances it is not), is it more convincingly "victimless" if the prostitute is a machine? And what will be the societal ramifications if we allow unfettered access to robot prostitution on the grounds that it is no different that using any other device for our entertainment?

4) Related to #3, lifelike machines with personalities and emotional responses will raise the question of people wanting to assign human qualities to those machines. Whether or not it's actually possible for machines to truly emote, have a soul, be sentient, etc etc, the more realistic they become the more we will *want* to think they are. In the same way that people attribute human qualities to animals with a consequent belief in things like animal rights, will there arise a movement for AI rights?

I can't fault his taste...