I was just at a conference where I had a nice conversation with a new, but already profitable company that builds face recognition for advertisers. Their custom

I was just at a conference where I had a nice conversation with a new, but already profitable company that builds face recognition for advertisers. Their customer base basically all wanted the same thing: to be able to target ads to the people walking past them, such as at bus stops (it's live in Singapore, if you're wondering). They eventually admitted they were working on voice recognition after strong customer demand, because that was much more promising; the machine vision stuff could only detect things like age, gender and income level. Yes, that's mining the content of conversations in the public space (like aforementioned bus stop) and generating ads on nearby real estate accordingly. Launching in Asia in 2015. Also had a 1 hour conversation in a private meeting room at another conference a couple years ago with three sharp representatives from a certain Asian TV manufacturer that may or may not have been mentioned in this thread. I was stuck by how far they had already gone applying machine learning to their customer activity. They were multiplying their efforts and investments in the space and were very excited about the new generation of TVs that go beyond passively showing you content, because of all the data they would be able to get as a result...