David Erickson: This is a this is fantastic. I’ve been meaning to try this out and I finally have so–
BL Ochman: I’ve tried it.
David Erickson: And this is Otter AI, which is a text transcription service. And I just tried it out today with one of our episodes. It’s very clean out of the gate. So if you upload your audio or video file and it sucks it in and spits out transcript after processing for about 15 minutes, I’d say.
The transcript I got was very clean. It identifies speakers, so it recognized that there were two speakers in one of our podcasts and labeled them as Speaker 1, Speaker 2, which you can then go in and re-label B.L. and David, as the case may be.
You can export content. You export the transcript and text, so it exports as a .txt file. You can during the export process toggle between whether you want to include speaker names, include timestamps, merge same-speaker segments, export as a monologue.
And the best thing about it, it’s free for up to 600 minutes a month. So in our case, we’ve got what four episodes a month at, at most an hour long. We’ve got plenty of room to do transcripts for free.
You do have to go in and edit and it has they kind of–you play it and it skips along through the transcript and you edit as you go along. It makes some mistakes and the punctuation isn’t all the best, but I’d say about 90% accurate.
What I’m curious to find out is whether with more uploads of the same content of our episodes, of our podcast, whether it’s going to start recognizing you, B.L., versus me and applying that and learning as it goes. But so far–
BL Ochman: I don’t know if it will do them. In my experience. It’s about 85 to 90% accurate, but it’s certainly a huge help and, you know, 100,000 times easier than trying to type while you’re listening, you know…