Image recognition is pretty old. We have had OCR of digits and letters a while now. Therefor I wonder if content such as music or movies can be protected by machine learning.
The idea is that display and audio devices could connect to a database with learned parameters for recognizing the particular media file. The parameters are learned to give a positive answer for a given video.
Knowing what media is played the device can then decide to play or not.
From this we possibly need a new legal video service. Like some popular ad supported music services. Maybe a Global TV station could work. Global TV but with local ads.
Since advertising on the Internet can be personal. It is not necessary to model ads after television.
One idea is to let small business, bloggers, private citizens promote their things on such a network. I mean the ad can be controlled to how many they reach. Then how much the ad costs for the seller is variable. Could be a small amount.
So it will be a way to grow smaller sized business or livelihoods.