“Do you want to clone your voice?” asked speech pathologist John Costello. I was sitting with him in his office at Boston Children’s Hospital, where for decades Costello has practiced his specialty: saving patients’ voices. Thanks to Costello and the clinicians around the world he’s trained in his methods, several thousand people, most of them ALS patients who know they will eventually lose control of the muscles in their mouth and larynx, have been able to preemptively preserve the ability to communicate using the sound of their own voice. He is an international leader in the field of “augmentative communication,” and voice cloning is the latest development.

Now Costello was offering to show me how it works, using my own voice. I read a prepared script—a few storybook-style sentences about sunshine and rainbows—into a voice recorder. When I finished, Costello uploaded the sixty-second audio clip to his computer and then typed out a few words that were different from anything I had actually said. “This is the voice we created with one minute of recording,” a voice suddenly boomed through a speaker, saying the very words Costello had just typed, and sounding uncannily like me. Costello then explained that with a longer recording, his AI-assisted program would sound even closer to my actual voice, better capturing my vocal tones, inflections, and speech patterns.

There are other tools out there that help speechless folks communicate using robotic-sounding voices and generic language. Costello’s efforts, though, are unique. He wants patients to sound like themselves. After all, to lose one’s distinctly individual voice, with all its unique color and character, is to lose a piece of identity. “The voice is an acoustic fingerprint,” Costello said. “There are turns of phrase you use that the people around you appreciate and recognize—and it’s not just the words, it’s also the delivery.”

Costello directs the ALS Augmentative Communication Program, an adult program at Children’s. The three voice preservation techniques he’s developed (see “Sound Methods,” below) have also been used to help patients with other temporary or permanent loss of speech, from those recovering from craniofacial surgery to a woman whose jaw was removed due to cancer. Costello, who is an American Speech-Language-Hearing Association fellow, also leads an annual clinical course at a foundation in Italy for children with communication disabilities, and he has been invited to share his work at trainings in nineteen other countries.

While majoring in psychology at BC, Costello found a summer job through the Campus School that involved shadowing speech pathologists at Children’s Hospital. Around 1989, a few years into his own career at the hospital, he learned of the fear, panic, and frustration many young patients experience upon waking from tracheostomy procedures and finding themselves unable to speak. He came up with the rudimentary solution of distributing devices with prerecorded phrases that allowed patients to communicate on a basic level. That sparked an idea: What if he worked in advance to capture the individual voices of patients scheduled for surgery, or those with degenerative conditions?

Today, it’s not only patients who ultimately benefit from his work. It’s also their family and friends, who won’t have to live without the sound of their loved one’s voice. Costello pointed to research that shows the voice of a mother can raise levels of oxytocin, the “love hormone,” in a child. “Voice is really powerful,” he said, “and really personal.”