OpenAI For the second time in as many months, theText-to-speech toolsExplain and re-emphasize that the tool is not currently widely available and may not be in the future.
Image source: Pexels
"Whether or not we end up deploying this technology on a large scale, it's important for people around the world to understand where this technology is going," OpenAI Inc. said in a statement posted on its website Friday, "which is why we wanted to explain how the model works, how we're using it for research and education, and how we're implementing security measures around the technology."
According to IT Home, OpenAI shared its "voice engine" with a small number of external users at the end of last year.Using text input and a 15-second audio clip of a human voice, the engine "generates a natural-sounding human voice that closely resembles the original speaker."The tool can create fake multilingual character voices. At the time, the company, which can create realistic multilingual character voices, said they chose to preview the technology rather than release it on a large scale in order to "increase social resilience" to the "increasingly realistic AI-generated models" that pose a threat. Threats.
As part of these efforts, OpenAI says they are actively phasing out the use of voice recognition for bank account verification, exploring policies to protect the use of personal voices in AI, educating the public about the risks of AI, and accelerating the development of technology to track audio-visual content so that users can recognize whether they are interacting with a real person or synthetic content.
However, despite these efforts, concerns about the technology remain. Bruce Reed, President Joe Biden's head of artificial intelligence, has said that voice cloning is one of the reasons he can't sleep at night. The U.S. Federal Trade Commission said in March that fraudsters are using AI technology to increase the credibility of their scams by using voice cloning tools that make it harder for people to distinguish between AI-generated voices and human voices.
In an updated statement on Friday, OpenAI attempted to alleviate these concerns, "We continue to engage with our U.S. and international partners from government, media, entertainment, education, civil society, and other sectors to ensure we incorporate their feedback into our build process." The company also pointed out that once the Voice Engine is equipped with its latest model, the GPT-4o, it will also pose new threats. The company said that they are actively "'red team testing' GPT-4o internally to identify and address known and unknown risks from areas such as social psychology, bias and fairness, and misinformation."