A musician who lost his ability to sing due to amyotrophic lateral sclerosis has returned to the stage with the help of artificial intelligence, highlighting how emerging technologies are reshaping creative expression for people living with severe medical conditions. Patrick Darling, 32, was diagnosed with ALS at the age of 29. The progressive motor neuron disease gradually affects the nerves that control muscles, eventually limiting movement, speech, and breathing. For Darling, who had been immersed in music since his teenage years and performed with multiple instruments including guitar, piano, mandolin, and banjo, the most devastating loss was his singing voice. As a member of the Ceili House Band, he regularly performed Irish folk music, building strong connections with fellow musicians and audiences. However, as symptoms advanced, he began struggling with balance, then lost control of his hands, and ultimately found it difficult to speak and breathe simultaneously, forcing him to step away from live performances in 2024.
After exploring traditional voice banking options, Darling realized that recordings made during the later stages of his illness would not accurately reflect the voice he once had. He was later introduced to AI voice cloning technology developed by ElevenLabs, a company specializing in synthetic speech and audio tools. Working with speech therapist and researcher Richard Cave from University College London, Darling used older audio clips recorded before his diagnosis to reconstruct a digital version of his original speaking voice. ElevenLabs has launched an impact initiative offering free access to its tools for individuals who have lost their voices due to ALS, stroke, or other medical conditions. The recreated voice closely mirrored Darling’s natural tone and inflections, preserving the personality and subtle imperfections that define human speech. For him, hearing the AI generated voice for the first time was both emotional and reassuring, as it sounded remarkably similar to how he had spoken before the illness progressed.
Rebuilding his singing voice proved more complex because high quality studio recordings were limited. Engineers relied on audio taken from live performances and informal home recordings, including clips captured in noisy environments. Despite these constraints, the AI system successfully generated a singing voice that retained slight raspiness and minor tonal variations, making it sound authentic rather than artificially polished. In addition to voice cloning, ElevenLabs offers an AI powered music generator called Eleven Music, which enables users to compose tracks using text prompts and stylistic inputs. Darling collaborated with Cave over several weeks to produce a song dedicated to his great grandfather, carefully refining lyrics, arrangement, and vocal output. The process allowed him to remain actively involved in composition and production, even though he could no longer physically perform with instruments.
The project culminated in a live performance at a technology summit in London, where Darling appeared on stage alongside his former bandmates while his AI generated singing voice played through the venue. The event marked his first return to the stage in two years and carried deep emotional significance for both the musician and his colleagues. While the technology does not alter the physical progression of ALS, it has enabled Darling to continue creating music and participating in artistic life. His experience reflects a broader shift in how artificial intelligence is being applied beyond commercial efficiency, offering new avenues for communication, identity preservation, and creative continuity for individuals facing life altering health challenges.
Follow the SPIN IDG WhatsApp Channel for updates across the Smart Pakistan Insights Network covering all of Pakistan’s technology ecosystem.
