Podcasting’s lack of metadata
It seems like it would be interesting to put audio feeds on my blog about some topic or another, however the fact that it isn’t really searchable without typing everything twice sucks.
The solution to this problem would be a really good offline speech-to-text converter. You should be able to do much better than the realtime systems because you have the complete context of every word instead of just those in front of the current one. My limited searching didn’t find anything great, but there are a few. I’m really looking for one on the Mac, although most people would probably be happy with a windows version. I just realized this should just be built into Google, rather than everyone having to do it. Get to work Googlers!
Wave To Text