OpenAI Voice Engine clones voices from 15-second clips
OpenAI sparked the artificial intelligence revolution through its text-generating program ChatGPT. However, generative AI development has gone beyond making words as other AI projects focus on audio statements based on cloned voices.
Recently, OpenAI teased its Voice Engine tool in a blog post which said that it uses “text input and a single 15-second audio sample to generate natural-sounding speech that closely resembles the original speaker.”
More info about the OpenAI Voice Engine
OpenAI deems its voice cloning tool too risky for general release https://t.co/j8tpjHv3eC
— Guardian news (@guardiannews) March 31, 2024
On March 28, OpenAI published an article reporting that it collaborated with various third parties to test Voice Engine’s capabilities.
READ: OpenAI removes Sam Altman’s ownership of its Startup Fund
Here’s how these companies used the voice-cloning technology:
- Age of Learning is an education technology company that used the new tool to create educational content tailor-made for various students.
- HeyGen used the OpenAI Voice Engine to create custom human-like avatars with multilingual voices. As a result, it enables businesses to translate their content into multiple languages.
- Dimagi, a software company making tools for community health workers, created interactive feedback in various languages via the AI voice cloning program. Consequently, it delivered essential services to remote areas more efficiently than ever.
- Livox, an AI app for Augmentative and Alternative Communication (AAC) devices, created unique, non-robotic voices for those with speech difficulties.
- The Norman Prince Neurosciences Institute at Lifespan, a nonprofit medical and teaching organization at Brown University, created AI versions of patient voices with the AI tool. As a result, these people can use these AI-powered voices to communicate.
READ: How to avoid AI voice scams
“We hope this preview of Voice Engine both underscores its potential and also motivates the need to bolster societal resilience against the challenges brought by ever more convincing generative models,” the official blog post said.
OpenAI hasn’t released this tool publicly to address potential risks, especially in the upcoming US elections.
OpenAI is working with various businesses and the US government to ensure it will be safe for public use. The AI firm hasn’t confirmed a release date at the time of writing.
