Descript began with a comparison Andrew Mason made when introducing it in 2017: writing has the word processor, while creating spoken audio and video still demanded complicated editing tools. Improving speech recognition offered a way to change that interface.
The product made the transcript part of the editing process, allowing people working with voice recordings to shape the media through text. The aim was to spend less effort learning editing machinery and more time developing the story, making talk-driven media easier to create and revise.