When doctors told Sonya Sotinsky that she would need to have her tongue and voice box removed to stop the spread of cancer in her mouth, she faced a devastating reality—she would never speak again. But instead of accepting silence, Sotinsky took a bold step: she sat down with a microphone and recorded every word she ever wanted to say, from everyday phrases to heartfelt confessions, even the curse words she’d never been able to say out loud. She used those recordings to train an AI voice model that could speak for her. The result was a digital voice that not only sounded like her but carried her personality—her humor, her warmth, and even her sharp wit. She named it “Soni,” a tribute to her own spirit. Sotinsky’s journey wasn’t just about regaining speech—it was about reclaiming her identity. She recorded lines from children’s books, favorite songs, and phrases she’d always wanted to say to her family and friends. The process was emotional, raw, and deeply personal. She wanted her voice to reflect not just her past, but who she was becoming. Now, with her AI-generated voice, she speaks freely—laughing, swearing, telling stories, and sharing her thoughts with the world. She’s even used it to record videos for her children, leaving behind a legacy of her voice that they can hear long after she’s gone. Her story is a powerful reminder of how technology can restore what illness takes away—not just function, but expression, connection, and joy. For Sonya, AI didn’t just give her back her voice. It gave her back her life.