I come down on Chris's side. Robots don't have to be as good or dispace 100% of jobs to have the potential for serious economic harm to quite large segments of the population.
While AI has langished for some time, "expert machines" have been taking sudden jumps lately. I tuned into the original DARPA Grand Challenge fully expecting the utter failure we saw. But then the next go-around had five finishers. Then they moved to the Air Force base roads complete with human drivers that deliberately broke the rules sometimes and still had many sucessful completions. I never expected those vehicles to advance in their capabilities so quickly.
It was a reminder that reality may outstrip my ability to imagine it.