IDEA #1: Multiple callers-to-text transcription software


Don’t you hate sitting through 60-minute podcasts only to hear 3-minutes of worthiness?

Jason Calacanis is joined by Doc Searls, Michael Arrington, and Dana Gardner among others for a virtual roundtable conversation available here. However, that webpage tells you nothing more about the discussion — other than it’s a 59-minute podcast. To know what they’re discussing, I’d have to listen to the 59 minutes.

The problem is that there’s so much content/opinion on the web these days, that I can’t possibly consume it all. The author should post an outline of topics discussed in the podcast, that’s a given. Taken a step further, and here’s the idea — I’d love to see a transcript of the podcast. But if you were to pass that podcast through a typical audio-to-text converter, it’s going to spit out one big mass of text. That would help, but ultimately I’d like to know who specifically said what and when.

I envision the participants of this virtual roundtable to dial into a conference call bridge and each audio portion is recorded separately by the proposed software — the software then converts the audio of each participant in the roundtable into text and mashes the text into chronological order to display much like a screenplay or IM conversation might appear.

Here is the napkin-sketch visual (hope to do these better in the future):

Honestly, there’s likely not much of a market for this idea. How many companies/people are doing virtual roundtable podcasts — or plan to in the future? However, there might be a corporate need for this type of utility — recording internal conference calls and providing the entire call as an audio archive and text transcript.

Are there any other uses for this technology? Any ideas you have to extend or better this idea? Discuss related technology, software, keywords, competitors, etc.