OpenAI has just announced a modest view for its latest tool, Voice Engine. This technology, known as voice cloning, can replicate any speaker’s voice by analysing a 15-second audio portion. The company claims that it creates speech that sounds natural, with emotive and genuine tones.
AI Voice Engine
The OpenAI Voice Engine has been in development since 2022 and depends on the company’s existing text-to-speech API. OpenAI already includes a version of this toolset to power the preset voices in the present text-to-speech API and Read Aloud function.
A series of samples posted on the company’s official blog show incredible authentic versions that bear an undeniable resemblance to genuine human voices.
For Translation
According to OpenAI, this technology could help in reading aid and language translation, as well as gain those with challenges.
The technology is now helping people with trouble speaking by using a Voice Engine clone based on audio recorded over a school project. This has been achieved through Brown University’s pilot the programme.
However, in spite its potential benefits, there is induce for focus in terms of malicious users. Voice Engine could easily be used to create highly realistic counterfeits of people, which is already common in the industry.
Privacy Concerns
This shows that the Voice Engine is not yet fully ready for general release, and some privacy concerns must still be addressed.
OpenAI acknowledges this issue, saying that the tech has “serious risks, which are especially top of mind in an election year.”
The startup is currently in the process of taking in user feedback from “US and international partners from across government, media, entertainment, education, civil society and beyond.” OpenAI has just released an availability date for Voice Engine.
To read our blog on “OpenAI to rehire Sam Altman as a CEO with new board members,” click here
