🇺🇸▼
Tears filled the audience as a deeply emotional song began to play. Patrick Darling had written this piece for his great-grandfather, a man he never had the chance to meet. However, the performance carried a weight far deeper than the lyrics suggested. It marked Darling's first time on stage with his bandmates since he lost the ability to sing two years prior.
The 32-year-old musician was diagnosed with amyotrophic lateral sclerosis, commonly known as ALS, when he was 29. Like other motor neuron diseases, ALS affects the nerves that control the body's muscles. As the condition progresses, individuals eventually lose the ability to control their muscles, including those used for moving, speaking, and breathing.
Darling's last live performance occurred over two years ago. By that time, he had already lost the ability to stand and play his instruments. He was also struggling significantly with his singing and speech. Recently, however, Darling was able to recreate his lost voice using a specialized artificial intelligence tool. This AI was trained on snippets of his old audio recordings. Another AI tool has subsequently enabled him to use this "voice clone" to compose entirely new songs. Through this technology, Darling is once again able to make music.
"Sadly, I have lost the ability to sing and play my instruments," Darling said during a recent event in London, speaking through his voice clone. "Despite this, most of my time these days is spent still continuing to compose and produce my music. Doing so feels more important than ever to me now."
Darling has been a musician and composer since he was approximately 14 years old. "I learned to play bass guitar, acoustic guitar, piano, melodica, mandolin, and tenor banjo," he explained. "My biggest love, though, was singing."
He met his bandmate, Nick Cocking, more than 10 years ago while they were both university students. Darling soon joined Cocking's Irish folk group, the Ceili House Band. Their first show together took place in April 2014. Cocking noted that Darling, who joined as a singer and guitarist, "elevated the musicianship of the band."
A few years ago, Cocking and his fellow bandmates began to notice subtle changes in Darling's behavior. He became increasingly clumsy. Cocking recalled a specific rainy night when the band had to walk across the city of Cardiff. "He just kept slipping and falling, tripping on paving slabs and things like that."
At the time, he did not think much of these incidents, but Darling's symptoms continued to worsen. The disease first affected his legs. In August 2023, he began needing to sit during performances. Shortly after, he started to lose the use of his hands. "Eventually he couldn't play the guitar or the banjo anymore," Cocking said.
By April 2024, Darling was struggling to talk and breathe at the same time. For that specific performance, the band carried him on stage. "He called me the day after and said he couldn't do it anymore," Cocking recalled, his voice breaking. "By June 2024, it was done." That was the last time the band played together.
Darling was connected with a speech therapist who raised the possibility of "banking" his voice. People who are losing their ability to speak can record themselves. These recordings can then be used to create speech sounds that are activated by typed text. This technology can be operated by hand or with a device controlled by eye movements.
Some users find that these tools sound robotic. However, Darling had a specific problem. "By that stage, my voice had already changed," he said. "It felt like we were saving the wrong voice."
Another therapist then introduced him to different technology. Richard Cave is a speech and language therapist and researcher at University College London. He also serves as a consultant for ElevenLabs, an AI company that develops audio, speech, and music tools. One of their specific tools can create "voice clones." These are realistic copies of real voices made from just minutes or even seconds of a person's recorded voice.
Last year, ElevenLabs launched an impact program. It promised free licenses to these tools for people who lost their voices to ALS or other diseases, such as head and neck cancer or stroke.
The tool is already helping some users. "We're not really improving how quickly they're able to communicate, or all the physical difficulties," said Gabi Leibowitz, a speech therapist who leads the program. "But what we are doing is giving them a way … to create again, to thrive." Users can stay in their jobs longer and "continue to do the things that make them feel like human beings," she explained.
Cave worked closely with Darling to recreate his lost speaking voice from older recordings.
"The first time I heard the voice, I thought it was amazing," Darling said, using the voice clone. "It sounded exactly like I had before, and you literally wouldn't be able to tell the difference. I will not say what the first word I made my new voice say, but I can tell you that it began with 'f' and ended in 'k'."
Recreating his singing voice proved not as easy as the speaking voice. The tool typically needs about 10 minutes of clear audio to generate a clone. "I had no high-quality recordings of myself singing," Darling admitted. "We had to use audio from videos on people's phones, shot in noisy pubs, and a couple of recordings of me singing in my kitchen."
Still, those snippets were enough to create a "synthetic version of [Darling's] singing voice," said Cave. In the original recordings, Darling sounded a little raspy and "was a bit off" on some notes. The voice clone retained these same qualities. It does not sound perfect, Cave said; it sounds human.
"The ElevenLabs voice that we've created is wonderful," Darling said. "It definitely sounds like me—[it] just kind of feels like a different version of me."
ElevenLabs has also developed an AI music generator called Eleven Music. This tool lets users compose tracks using text prompts to choose the musical style. Several well-known artists have partnered with the company to license AI clones of their voices. These include actor Michael Caine, whose voice clone is being used to narrate an upcoming documentary. Last month, the company released an album of 11 tracks made using the tool. "The Liza Minnelli track is really a banger," Cave said.
Eleven Music can generate a song in a minute. But Darling and Cave spent about six weeks fine-tuning Darling's song. Using text prompts, any user can "create music and add lyrics in any style [they like]," Cave explained. Darling prefers Irish folk music. Cave has also worked with a man in Colombia who is creating Colombian folk music. The ElevenLabs tool is currently available in 74 languages.
Last month, Cocking received a call from Cave, who sent him Darling's completed track. "I heard the first two or three words he sang, and I had to turn it off," Cocking said. "I was just in bits, in tears. It took me a good half a dozen times to make it to the end of the track."
Darling and Cave planned to perform the track live at the ElevenLabs summit in London. Cocking and bandmate Hari Ma each arranged parts to play on the mandolin and fiddle. They had a couple of weeks to rehearse before they joined Darling on stage. It had been two years since their last performance together.
"I wheeled him out on stage, and neither of us could believe it was happening," Cave said. "He was thrilled." The song played as Darling remained on stage, while Cocking and Ma played their instruments live.
Cocking and Cave say Darling plans to continue using the tools to make music. Cocking says he hopes to perform with Darling again. But he acknowledges that, given the nature of ALS, it is difficult to make long-term plans.
"It's so bittersweet," Cocking said. "But getting up on stage and seeing Patrick there filled me with absolute joy. I know Patrick really enjoyed it as well. We've been talking about it … He was really, really proud."