He argues that “if/then” rewards work only for very simple tasks. Further, for more complicated tasks, ones that require creativity, rewards actually harm outcomes. He argues that for these more complicated tasks, an approach organized around autonomy, mastery and purpose are what leads to the best outcomes.
It seems like this philosophy would lend itself to motivational interviewing. It emphasizes tapping internal vs. external motivation. At one point he says that rewards are helpful for obtaining compliance but not engagement. If these lessons have any application to addiction treatment, this makes one wonder about the long term value of contingency management. Time will tell, but I found this to be a pretty interesting talk.