RikSharp said:
nice, but the idea of robots having any sort of artificial intelligence or decision making capability is only going to lead to problems in future.
AI cleaning robot "I clean mess up every day. Would be more efficient to stop source of mess."
AI cleaning robot kills human.
AI cleaning robot cleans up mess.
AI cleaning robot rests.
What you have here is a case of stupidy on the part of the programmer (I've said before, the machines won't kill us unless we A. Treat them as second class citizens and, thereby, abuse them or B. Program them in a way that will inevitably cause them to destroy us [As in the above])
Rather than programming them to clean up the mess, give them the programming:
AI cleaning robot is programmed with "I must have a purpose."
AI cleaning robot is programmed with "Cleaning up after humans is my purpose."
AI cleaning robot computes "Without humans, there would be nothing to clean up after, leaving me without purpose."
AI cleaning robot computes "In order to protect my purpose, I must protect humans at all costs"
AI cleaning robot computes "In order to protect humans, humans must be protected from themselves"
Wait...I think I just defeated my own argument...