That does sound creepy, what do they do after the planet is gone.

Terminator has a very basic what if AI scenario, like Ultron. To save humanity from threat it eliminates it so currently the biggest threat of then all is us, so error code of some sort. Terminator is not Terminator's Movies it's James MaFffackin Cameron. It also deals with time travel and all sorts of hard shit to write.

Cyber Terrorism is the best bet for this scenario. Also you know how life I often weirder than science fiction? Were going to be like using strands of DNA for memory storage and Graphene microchips, ferrfluid coolants and harddrives, and sicilene and carbyne transistors, it's going to be wierd.

If they are for destroy they will consume earth.
If they are for help the humans against cancer or AIDS, I think the opposite.
For example:
Nano-Robots will have a reproduction system wich the nano-robots they create, they never will reproduce again.
and the Reproduction Robots will have a limit of creating nano-robots.

The planet wouldn't be gone, just have all of its organic matter turned into nano-bots. Here's a handy graphic summarizing the process:



Image source: http://science.howstuffworks.com/gray-goo.htm

well yeah this is a really dumb thing people think on. robots can never destroy humanity coz i dont think someone will make something have that much of smart to get evolved by itself with an error that make it free to think in stuff like human mind.

oh buddy, you may be in for a surprise there, unless you have some viable evidence to back that statement up.

anyway, op's question; they probably could, but, it'll be probably over 10 years before we'll even think about that truly happening -just my opinion.

edit; 'If mankind was dumb enough to make a computer able to get hostile towards humanity's existance; the computer would be too smart and calculating to realize that most of us would be too stupid to be killed by it.'

what if someone decided it was a good idea to use robots to fight their wars, you know, like using robots to kill humans, since robots could be programmed to drive tanks and other vehicles.
also the last part of the sentence makes no sense at all, care to explain?

I asked politely for some evidence, not opinions.
I agree with zelda on this, even though we have no johnny 5 from short circuit, we have AI's, a chain of information forever growing from experiences, like how humans learn, and are getting nearer and nearer to a artificial brain that could eventually be used by an artificial body.

if you want some evidence about the brain, head over to www.cleverbot.com
a site where you can 'talk' with a robot, where it has information stored from past experiences with other people so it may get cultural references, remember your name, or be able to describe how a certain food may taste.
it's obviously not perfect, but this could be a stem of something huge given time.

I see what you are saying with this, but, the biggest issue at the moment is the memory. We have not yet been able to get a good memory to work with these robots, therefore its like talking to a database with anterograde amnesia, it will only remember a couple of new lines you've said before losing its train of thought and switching the topic to something random.

If your A.I is going awol you just revert the code back to an earlier stage where it didn't do that... Nothing Hollywood about it

Odlov and his image has a point that can't just be overlooked.

Also, anyone think that a super computer could potentially see humans as a threat and set off nukes?

spitballing

Without a doubt they're going to go that far. Hell it might be like world war whatevs. Then some scientists want to create Terminators. Able to infiltrate the ranks then assassinate. We have no clue how it's going to happen if it even will, but an actual AI being able to decide for it's self in such ultimatums as yourself to save 2 other people or save yourself. Watch an AI save himself instead the 2 humans.