I think LingQ would benefit by promoting the transcription of copyright free audio by learners. For example, as a Japanese or French learner I can transcribe some Japanese or French audio. Then it can be checked by a native speaker and submitted to the library. I pay the checker by the word, but I get to submit it and earn points if possible.
Another idea is to use online software like http://en.podcastle.jp/
Users submit audio which is converted to text with speech to text software, with errors of course. But users can correct errors with an ingenious interface that crowd-sources the correction. Correction is done without payment. And final transcripts can join the library.