Youtube introduces automatic captioning for the deaf- Why should you get excited about it?

Youtube has come out with a brand new feature - Automatic captioning of its video to produce subtitles for the benefit of the deaf. In a nutshell, the service will produce subtitles for videos which have been spoken in fluent English. So for example , a deaf person watching a video of Obama making a speech can read about it in real time as he/she watches the video. That's cool!
But this technology if fully developed has some really deep implications on the way we use our online content. Essentially it is an extension of audio based content retrieval. Some of them I enlist here, others you are free to dream about:

Multiple language captioning: A video which has content in English language can be understood by a French person if the technology allows the bot to translate languages. That means a deaf(or otherwise) French guy can view Obama make his speech and understand it completely, all in real time.
Social Media: We already have been using videos on social networking sites for ages. How about using it for social networking? Videos uploaded online could be therefore processed for references of people within ones social circle. These references could then be used for all the millions of applications now being built on the social media platform. For example , if I'm uploading a video of my friends Ross and Chandler, then a alert could shoot off onto their (and possibly their friends) Facebook pages.
Brand management: Companies can derive useful information from user videos regarding the impact of their product has created. Obviously this application is weak in the sense that users will rather opt for textual medium(at least for a few years more) to put their feelings out in the open.
Marketing: This is one that is accompanied with any innovation in the field of information retrieval. Videos often seem to have content that is different from what their titles state. For eg, a video titles "The best hangout place in Mumbai" could talk about something different , say sea food at one particular place. In such a case it would be better to use the video itself to generate ads.
Clustering based on content: All of the above applications are more of audio based content ones. One use that is based on video content(Obviously this does not relate to this particular innovation) is grouping videos based on the spectral patterns of the content. Something similar to Image Swirl by google(Check it out here) . So outdoor videos can be separated from indoors, etc.

Well I have gone on for more than a googol pages it seems . But I'm sure you see that this technology could have far reaching effects. I'm particularly interested in the social media aspect of it since video based social networking is really not into the picture yet.. Something to watch out for...