Brit Sarah Ezekiel lost her voice after developing motor neurone disease (MND) 25 years ago and ever since has been speaking with what she described as a “posh robot’s voice.”
MND is a progressive neurological condition that damages the nerves controlling muscles, leading to weakness and eventual paralysis. After about five years with the disease, Sarah was able to use eye-gaze technology that allowed her to type and speak with a synthetic voice, similar to the late physicist Stephen Hawking, the BBC reported.
While the artificial voice allowed her to continue to communicate through speech, Sarah said she never really liked the sound of it, and wished she could communicate with her now-adult children using her real voice.
Enter a U.K.-based tech firm called Smartbox, which said recently that it might be able to help her so long as there was an audio recording of Sarah’s real voice that it could work with.
The family found an old VHS tape with footage of Sarah, but the audio of her speaking only lasted eight seconds and was obscured by background noise.
Working with various AI technologies that included ElevenLabs’ Voice Isolator, Smartbox’s Simon Poole was able to clone Sarah’s voice and incorporate it into her speech software. You can hear the impressive result in this BBC News report.
Hearing her speak with her real voice for the first time, daughter Ava told the BBC: “It was amazing. I’m still coming to terms with it. Hearing it now in everyday life, it still surprises me.”
She added: “We can feel who she is as a person — mum isn’t just a disabled person in the corner with a robot that doesn’t relate to her.”
Sarah’s story shows how the latest technology can not only restore someone’s speech, but also, by enabling them to speak in their own — albeit synthetic — voice, help them reclaim their identity and reconnect emotionally with loved ones.
“After such a long time, I couldn’t really remember my voice,” Sarah said. “When I first heard it again, I felt like crying. It’s a kind of miracle.”
