Some people are betting that accurate voice recognition capabilities will be a boon for Video Search. It will give the Search Engines some context for videos beyond the  current meager  set.

This will only be true, however, for videos with spoken word soundtracks. And only for videos where the spoken word soundtrack is pertinent to what the video is really about. Marketers will quickly latch on to the capability of stuffing the audio portion of videos full of keywords, but it won’t likely effect how others produce theirs.

I don’t really see this as having any great significance to video search, but I do fear that SEMs will take it, make it theirs and break it very quickly, so I hope the Search Engines are considering how to prevent this undesired side effect.