I don't think the danger is so much a super-brainy robot/computer deciding to eliminate humanity. I think the real danger is the power that robotics can give to already existing human brains that already have not only the will to survive but the "will to power" (to borrow a phrase). Robots don't have to learn to hate --- as long as there is just one human master the potential is there for all the usual troubles - except made worse by technology.
As far as 'brain like' thinking, the real danger may come from fusing actual brains to technology. Already there are experiments in using insect-brains to control 'robots'. You'd better believe that a lot of people are hot for human/machine interfaces that, combined with 'telepresence', could combine the intelligence of humans with the survivability of machines. So it's not like we would have to develop brains from scratch, rather we could just meld them with technology to create intelligent machines, or machine intelligences, or whatever we would call them. Such a creation might have the worst of two worlds.
