Groupon’s Andrew Mason is back with “a word processor for audio”

In 2013, Andrew Mason was famously fired as CEO of Groupon, a company he cofounded. His second act, Detour, was a startup that created podcast-like audio tours. That involved producing a boatload of spoken-word content. And Mason discovered that working with speech in existing sound-editing programs was a hassle.

“Trying to edit waveforms, even once you get over the learning curve, is tedious and requires you to disengage your editorial brain and engage your engineering brain. It makes it a non-delightful, non-creative process,” he explains.

Detour ended up building its own editing tools. And now Mason is leaving Detour and—with $5 million in seed funding from Andreessen Horowitz—is spinning out that technology into a new startup called Descript. The company is launching a Mac app and web service which transcribe spoken audio into written text, and then let you edit the audio by making changes to the text using a word processor-like interface, including moving sections around or deleting them altogether. Descript charges 7 cents a minute for automated speech-to-text conversion based on Google technology and 99 cents a minute for “white glove” human transcription, and lets you share your work with other folks for their comments.

Mason likens Descript to Photoshop, YouTube, and other tools that democratized various aspects of content creation and distribution. As new technologies open up possibilities such as synthesizing speech that sounds like it was spoken by a particular person, he plans to take advantage of them. “We’re on the precipice of a new age of media production where artificial intelligence is going to have a huge impact,” he says. “We see Descript as a reset button that sets that up for that future.”