Ask HN: What would you do if you could detect objects in video?

https://news.ycombinator.com/item?id=6497579

Regarding the above..

If there were some way to detect objects instances in video (that is I show the algorithm what I want to detect before hand, so I only care if it detects 'my' chair, not 'all chairs'), how would you use it?

It seems like this technology is inevitable..