OpenAI, the creators of ChatGPT, has announced that they are entering the voice assistant industry with their new Voice Engine technology. This technology can clone a person’s voice using just 15 seconds of recorded speech, although it is currently not being released to the public due to safety concerns. The company is only previewing the technology with early testers for now.
The decision to keep the technology out of widespread release stems from the potential risks of misuse that come with creating speech that closely resembles someone’s voice. OpenAI is aware of the serious implications, especially in an election year, as seen in a recent incident in New Hampshire where AI-generated robocalls imitated President Joe Biden’s voice.
While there are already companies selling voice-cloning technology, OpenAI is taking a cautious approach with their Voice Engine. Testers of the technology have agreed not to impersonate individuals without consent and to disclose that the voices are generated by AI. This responsible approach is similar to how the company handled the release of their video-generator, Sora.
An indication of OpenAI’s future plans in the voice assistant industry was revealed through a recent trademark application for the name Voice Engine. This move suggests that the company is aiming to develop speech recognition and digital voice assistant technologies. By advancing in this field, OpenAI could potentially pose a challenge to existing voice products like Amazon’s Alexa.
As part of their partnership with the Associated Press, OpenAI has access to AP’s text archives. This has allowed them to continue developing cutting-edge technologies while respecting ethical considerations surrounding voice cloning and other AI innovations.
:quality(75)/cloudfront-us-east-1.images.arcpublishing.com/elcomercio/3Z35VNBPCBGUHHMEZC3M4NIVTU.jpg)