“Reinforcement Learning From Advertiser Feedback”

That’s the gist of this century thus far, isn’t it? :P