In order to pay for the resources that they consume, AI machines need customers. If they run out of customers, they run out of resources, and they shut down, which makes room for human labor. As long as humans are the only property owners, the customers must be humans (or agents for humans), and there's a symbiosis there. In that case, I think that the economy will regulate the speed of AI adoption so that people will not be displaced.
Where things get thorny is when the AI is sufficiently evolved to be considered sentient and demand property rights. If human property owners have to compete against AI property owners, I think the humans may be threatened.
Of course, Kurzweil's vision has always been of the dominant human-machine hybrid, which I still think is plausible. Maybe a machine can out-think one person, or ten or twenty, but can a machine out-think an entire society with brains linked together via high speed interconnects?
I imagine a future global supercomputer which is a conglomeration of human and machine intelligence supplemented by mostly mechanical labor. Maybe "going to work" means plugging in a head-set and renting our intellects to the Matrix for the day.
They'll just steal your eyeballs and cpu cycles.