[Dwolb]

Mentioned previously in this thread, better is a broad term. From a customer perspective, there are thousands of use cases to look over and examine. Maybe one use case could be worth exploring and then have the concept applied to other use cases.

Let's take the use case of 'instructional video'. My dad doesn't know how to change his brakes, but is functional at using tools. He goes onto Youtube and looks for 'How to change brakes'. He finds a video for what he wants to do with a car that looks similar to his and watches this video. He watches the video 2 or 3 times and then goes out to his car with his laptop. He begins playback until the first step is completed then he tries to complete the first step himself. etc...

The current Youtube implementation could be improved the following way: Youtube knows what car my dad drives and pushes the corresponding 'How to change brakes' video to the top of the search accordingly. My dad clicks the video and the video caches to allow my dad to walk outside, where Wifi is spotty, and watch the video. The video has several break points where my dad knows not to continue to the next step in the 'How to change brakes' video until he's completed the first step. Finally, the system is able to recognize my dad has completed the first step and continues cached video playback to continue to the next step. Finally, Youtube follows up with a video to watch to verify brakes have been changed properly. It might even be useful to show an advertisement of a local brake expert.

Just a thought experiment. There's some vision for making Youtube better and there are a lot of different ways to execute the vision.