Human intervention and refusal to allow it, as you already mentioned.
It could be a self-fulfilling prophecy, but not necessarily a logical certainty. For example, if I could convince you (and with the same amplitude of conviction that you have right now towards a "smarter"-computer future) that such a future will inevitably result in the slavery of one man by another, and of ever-more powerful dictatorships (that call themselves democracies), this very thought would be the first step towards the non-conception of the singularity. I believe a person called Theodore Kacinzsky has previously argued along these lines, though he chose a more explosive form of argument than mere words :) http://en.wikipedia.org/wiki/Theodore_Kaczynski
One alternative is to restrict research to biological advances that benefit humans (stem cell research, for example) rather than trying to create even-more powerful machine learning tools all the time.
Just because "everyone" eats McDonalds, it does not take away the power of choice from you to never ever visit a McDonalds (for whatever reason), for example. Look at how scarily equipped are the police today, what things man has created for war (weapons that boil your invisibly from a remote distance), and you will understand that technology is fun as long it's not. It doesn't take a genius to look at each invention and decide whether it will most probably be used for the good or result in further loss of privacy and freedom. Two headings from today' ACM "TechNews" newsletter in my inbox are, "The Display That Watches You" and "Predictive Powers: A Robot That Reads Your Intention?". My reaction is : "Are you dumb? don't you see where this is leading?"