San Francisco.- ChatGPT developer OpenAI is entering the virtual assistant business and showcasing new technology that can clone a person's voice, although it says it won't release it publicly yet for security reasons.
The artificial intelligence company unveiled its new Voice Engine technology this Friday, just over a week after filing a trademark application for that name. The company says it can recreate a person's voice by simply having access to a 15-second recording of their voice.
OpenAI notes that it plans to preview it with the first people to test it, “but not widely release this technology at this time” due to the dangers of misuse.
“We recognize that generating speech that resembles people's voices carries serious risks, which must be taken into account, especially in an election year,” the San Francisco company said in a statement.
In New Hampshire, authorities are investigating robocalls made to thousands of voters shortly before the presidential primaries, in which a voice generated by artificial intelligence that imitated that of President Joe Biden was heard.
A number of startups already sell technology that clones voices, some of which are already accessible to the public or to select commercial clients, such as entertainment studios.
OpenAI says the first people to test Voice Engine agreed not to impersonate another person without their consent and to disclose that the voices are generated by AI. The company, best known for its chatbot and image generator DALL-E, took a similar strategy by announcing its Sora video generator, but without releasing it widely.
However, the trademark application filed on March 19 shows that OpenAI is likely to try to enter the speech recognition and virtual assistant business. In the long run, improving such technology could help OpenAI compete with other similar voice products, such as Amazon's Alexa.
#OpenAI #exhibits #technology #clones #people39s #voices