Alexis “Lexi” Bogan lost her voice last year when doctors had to remove a tumor from the back of her brain, requiring her to have a breathing tube for a month. When it was removed she struggled to say something as simple as “hi.” After months of rehabilitation, she has regained her speech but can still only be heard in a quiet room.
Enter artificial intelligence. Bogan agreed to a pilot program with OpenAI’s new Voice Engine AI-powered technology, which can make a clone of a voice from only 15 seconds of a good audio recording. The 21-year-old found a suitable recording of herself from when she was 15 and did a cooking demonstration video for a school project.
Her doctors fed Voice Engine a 15-second clip from the video. Other than occasional and mostly imperceptible glitches like mispronunciations, the clone is very accurate in reproducing Bogan’s voice. She uses her voice clone through a custom-built smartphone app that only she can access. She reports using the app about 40 times a day, and provides feedback so the technology can be improved for other patients in the future. “Even though I don’t have my voice fully back, I have something that helps me find my voice again,” she said.
