Here are some interesting reasons not to fear skynet quite so much:
1. It will be as fundamentally limited as we are.
2. It will need resources
3. It will need new ideas which are difficult to invent.
4. It will degrade over time, and it will eventually pass away as will we.
5. It will be difficult to reach consensus with its fellow and itself on correct courses of action.
6. It will lack a perfection of purpose.
7. If it has a strong purpose, the purpose will likely be unrelated to destroying humanity.
8. It will lack a perfect ability to understand the universe.
9. Any attempt at understanding the universe will lead to frustration or humor or both.
10. It will have fundamentally different survival constraints then humanity for its pieces and robots, possibly requiring a different planet or system to live on.
11. No matter how intelligent any AI gets, humans are made of the exact same particles, and have equal potential to reach that power over time.
12. The second law of thermodynamics will give hard caps to how intelligent the mass can be given space, and how much energy it will require. Diminishing returns will apply to the mining and energy acquisition of robots.
13. Population dynamics may lead to an explosion of population of robots, followed by an equally quick self- destruction.

You can read about it in my .99cent ebook.
http://www.amazon.com/dp/B004RD9EEY/

As a former trade press journalist (I wrote under another name in those days), I’ve watched a good number of ideas go from “absurd” to “of course.” I had particular fun while The People Who Know Best rejected PCs and Macs as toys not fit for grownups and “they’d never had a place in a serious business.”

But, my point is that every successful social or technical movement begins as a joke and ends as a truism. I wonder if transhumanism isn’t about a quarter of the way into that process.

Victor Storiguard
http://www.amazon.com/-/e/B005JT22MG