OpenAI Introduces Voice Cloning Tool, But Waits for Public Release
2024-04-11
LRC
TXT
大字
小字
滚动
全页
1ChatGPT developer OpenAI has introduced a new tool that it says can reproduce human voices with just a short audio sample.
2The tool is among several developed by technology companies that aim to clone voices with a high level of exactness.
3The system is called Voice Engine. OpenAI released details about Voice Engine on March 29.
4The company also released some examples of how the tool performs.
5The company explained in a statement that the system uses written instructions and a 15-second audio sample to produce, or generate, "natural-sounding speech" in the voice of any speaker.
6OpenAI said it has been working on developing the tool since late 2022.
7It plans to continue developing it before releasing it to the public.
8The company said the delayed release represents "a cautious and informed" effort to prevent possible misuse.
9OpenAI said it began testing Voice Engine "with a small group of trusted partners."
10This testing aims to help company officials decide how to improve the tool's safety and learn how it can "be used for good" across different industries.
11There have already been highly publicized examples of false, or fake, voice recordings that sound like well-known people and politicians.
12One such example was an AI-manufactured voice claiming to belong to U.S. President Joe Biden.
13The voice was turned into a message that was sent by telephone to voters earlier this year in the state of New Hampshire.
14In the fake audio, Biden appears to try to urge voters not to take part in presidential primary voting.
15Election officials in New Hampshire are still investigating the incident.
16OpenAI's statement identified such election risks.
17"We recognize that generating speech that resembles people's voices has serious risks, which are especially top of mind in an election year."
18There have been other examples of so-called "deepfakes" being used in political campaigns around the world.
19A "deepfake" is a piece of audio or video created to make it appear that people in it are saying or doing things that they never did.
20OpenAI said it is currently discussing how to limit these and other risks of such voice cloning systems.
21Technology experts have also warned such tools can be used to carry out financial crimes.
22For example, voices can be recorded and then reproduced to represent another individual in an effort to carry out fraud.
23The company said it has been in contact with U.S. and international partners "from across government, media, entertainment, education, civil society and beyond."
24The aim is to get advice from people in these groups about the best ways to develop and deploy the system, OpenAI said.
25OpenAI said its early Voice Engine testers have agreed not to represent a person without their permission and to clearly state that the voices are produced by AI.
26The company is best known for its launch of the AI-powered ChatGPT tool, which was launched in November 2022.
27There are several other companies developing similar voice cloning tools.
28One of the most widely known is called Descript.
29The company explains on its website the AI-powered system can clone any person's voice with a 60-second audio sample.
30Descript also offers other services involving audio and video production.
31The company's website says users can choose a free plan that includes up to 1,000 words of voice cloning.
32For $24, users can get unlimited voice generation, Descript said.
33OpenAI listed several possible uses for the technology tool.
34These include providing reading assistance to non-readers or children, supporting individuals who cannot speak, and helping patients recover their voice.
35The company said its Voice Engine system can work in several languages.
36One example of such use is taking an English voice recording and producing a new audio version in another language.
37In an example released online, OpenAI demonstrated a person's voice and style of speaking can remain the same in other targeted languages.
38I'm Bryan Lynn.
1ChatGPT developer OpenAI has introduced a new tool that it says can reproduce human voices with just a short audio sample. 2The tool is among several developed by technology companies that aim to clone voices with a high level of exactness. 3The system is called Voice Engine. OpenAI released details about Voice Engine on March 29. The company also released some examples of how the tool performs. 4The company explained in a statement that the system uses written instructions and a 15-second audio sample to produce, or generate, "natural-sounding speech" in the voice of any speaker. 5OpenAI said it has been working on developing the tool since late 2022. It plans to continue developing it before releasing it to the public. The company said the delayed release represents "a cautious and informed" effort to prevent possible misuse. 6OpenAI said it began testing Voice Engine "with a small group of trusted partners." This testing aims to help company officials decide how to improve the tool's safety and learn how it can "be used for good" across different industries. 7There have already been highly publicized examples of false, or fake, voice recordings that sound like well-known people and politicians. 8One such example was an AI-manufactured voice claiming to belong to U.S. President Joe Biden. The voice was turned into a message that was sent by telephone to voters earlier this year in the state of New Hampshire. 9In the fake audio, Biden appears to try to urge voters not to take part in presidential primary voting. Election officials in New Hampshire are still investigating the incident. 10OpenAI's statement identified such election risks. "We recognize that generating speech that resembles people's voices has serious risks, which are especially top of mind in an election year." 11There have been other examples of so-called "deepfakes" being used in political campaigns around the world. A "deepfake" is a piece of audio or video created to make it appear that people in it are saying or doing things that they never did. 12OpenAI said it is currently discussing how to limit these and other risks of such voice cloning systems. Technology experts have also warned such tools can be used to carry out financial crimes. For example, voices can be recorded and then reproduced to represent another individual in an effort to carry out fraud. 13The company said it has been in contact with U.S. and international partners "from across government, media, entertainment, education, civil society and beyond." The aim is to get advice from people in these groups about the best ways to develop and deploy the system, OpenAI said. 14OpenAI said its early Voice Engine testers have agreed not to represent a person without their permission and to clearly state that the voices are produced by AI. The company is best known for its launch of the AI-powered ChatGPT tool, which was launched in November 2022. 15There are several other companies developing similar voice cloning tools. One of the most widely known is called Descript. The company explains on its website the AI-powered system can clone any person's voice with a 60-second audio sample. 16Descript also offers other services involving audio and video production. The company's website says users can choose a free plan that includes up to 1,000 words of voice cloning. For $24, users can get unlimited voice generation, Descript said. 17OpenAI listed several possible uses for the technology tool. These include providing reading assistance to non-readers or children, supporting individuals who cannot speak, and helping patients recover their voice. 18The company said its Voice Engine system can work in several languages. One example of such use is taking an English voice recording and producing a new audio version in another language. In an example released online, OpenAI demonstrated a person's voice and style of speaking can remain the same in other targeted languages. 19I'm Bryan Lynn. 20Bryan Lynn wrote this story for VOA Learning English, based on reports from OpenAI, The Associated Press and Agence France-Presse. 21__________________________________________ 22Words in This Story 23sample - n. a small amount of something that gives you information about the thing it was taken from 24clone - v. to create something that is very similar to something else 25instructions - n. advice and information about how to do or use something 26cautious - adj. taking care to avoid risks or danger 27resemble - v. to look like or be like someone or something 28top of mind - idiom. something that occupies a big part of a person's thoughts 29fraud - n. the crime of doing something illegal in order to get money 30style - n. a way of doing something that is typical of a particular person, group, place, etc.