In the persistent fight for personalized streaming experiences, content recommendations cannot presume to know better than the user, according to Google TV’s Shobana Radhakrishnan, Senior Director of Engineering. Google is attempting to bring Google search-level effectiveness to Google TV discovery, added Radhakrishnan, speaking on a Streaming Media Connect panel focusing on AI and streaming personalization. The era of recommendations being solely based on previous watch history is over. Alex Gruber, CPO at Deltatre, noted that incorporating surrounding contextual information is important to assessing mood, notably for advertising. Personalization training, in addition to using AI to assess a users’ mood, should also take into account context such as weather, or even the political climate in a certain region to make sure…