Science and Technology Science and Technology
Wed, December 11, 2024

AI can bring back a person's own voice


Published on 2024-12-11 14:42:32 - The Economist
  Print publication without navigation

  • But when Pedro speaks, his lightly accented English flows with ease. His voice is generated by an artificial-intelligence model, trained on clips recorded before he lost his speech. His words, too, are generated, by a large language model ( LLM) fine-tuned on his writing.

The article from *The Economist* discusses how AI technology is being used to restore the voices of individuals who have lost their ability to speak due to conditions like motor neurone disease. This innovative approach involves training AI models on recordings of a person's voice from before their condition worsened. The AI then synthesizes speech that closely mimics the individual's original voice, providing a deeply personal and comforting means of communication. This technology not only aids in personal interactions but also has potential applications in therapy, helping patients maintain a sense of identity and connection with their past selves. The article highlights the emotional and psychological benefits of this technology, emphasizing its role in enhancing the quality of life for those affected by voice-impairing diseases.

Read the Full The Economist Article at:
[ https://www.economist.com/science-and-technology/2024/12/11/ai-can-bring-back-a-persons-own-voice ]
Contributing Sources