Scarlett Johansson says she was irritated by ChatGPT’s voice similar to hers

Actress Scarlett Johansson said in a statement sent to CNN on Monday (20) that she was “shocked, angry and incredulous” that OpenAI CEO Sam Altman would use a synthetic voice released with a ChatGPT update “so eerily similar” to hers.

The statement comes after OpenAI announced that it is pausing the update following comparisons with a fictional voice assistant played by Johansson in the 2013 near-dystopian film “Her.”

The withdrawal by OpenAI follows a backlash to the artificial voice, known as Sky, which critics described as overly familiar to users and sounded as if it emerged from a male developer’s fantasy. She was widely mocked for her flirtatious tone.

“We’ve received questions about how we choose voices on ChatGPT, especially Sky,” OpenAI said in a post on X on Monday. “We are working to pause the use of Sky while we address these issues.”

Johansson said Altman offered to hire her last September to voice the ChatGPT 4.0 system. She said she turned down the offer for “personal reasons.”

“Two days before the ChatGPT 4.0 demo was released, Mr. Altman contacted my agent, asking me to reconsider. Before we could connect, the system was already out there.”

Johansson said she has hired legal advisors and that OpenAI “reluctantly agreed” to remove the “Sky” voice after its lawyer sent two letters to Altman.

“At a time when we are all dealing with deepfakes and protecting our own image, our own work, our own identities, I believe these are issues that deserve absolute clarity. I look forward to a resolution in the form of transparency and the passage of appropriate legislation to help ensure that individual rights are protected,” Johansson wrote.

The company said in a blog post on Sunday that the voice in question is not derived from Johansson’s but rather “belongs to a different professional actress using her own natural voice.”

Altman reiterated the company’s position that “Sky” was voiced by a different actress in a statement Monday following Johansson’s allegations.

“Sky’s voice is not Scarlett Johansson’s, and it was never intended to sound like hers,” said Altman. “We chose the actress behind Sky’s voice before any contact with Ms. Johansson. Out of respect for Ms. Johansson, we have paused the use of Sky’s voice in our products. We apologize to Ms. Johansson for not communicating better.”

OpenAI said that with each of its AI voices, it has tried to create “an accessible voice that inspires trust,” one that contains a “rich tone” and is “natural and easy to listen to.” ChatGPT’s voice mode that used Sky’s voice had not yet been widely released, but videos of the product announcement and teasers of OpenAI employees speaking to him went viral last week.

Some who heard Sky ridiculed it for being perhaps too easy to listen to. Last week, the controversy inspired a segment on The Daily Show in which senior correspondent Desi Lydic described Sky as a “pervert baby robot voice.”

“This was clearly designed to feed the guys’ egos,” Lydic said. “You can really tell that a man built this technology.”

Even Altman seemed to recognize the widespread parallels users were drawing with Johansson when she posted to X on the day of the product announcement: “her.” Johansson said Altman used that post to imply that “the similarity was intentional.”

“Her” is the title of the 2013 film in which Johansson plays an artificially intelligent voice assistant with whom the protagonist, played by Joaquin Phoenix, falls in love, only to be heartbroken when the AI ​​admits that she is also in love with hundreds from other users and subsequently becomes completely inaccessible.

Watch the trailer for “Her”

Leadership questions

The criticism surrounding Sky highlights wider social concerns about the potential biases of a technology designed by technology companies, largely led or funded by white men.

The announcement came after OpenAI leaders were forced to defend their security practices over the weekend after an employee who left the company questioned the company’s priorities.

Jan Leike, who previously led a team focused on the long-term safety of AI but left OpenAI last week along with co-founder and chief scientist Ilya Sutskever, posted a follow-up to X on Friday claiming that “in recent years , security culture and processes took a backseat to flashy products” at OpenAI. He also raised concerns that the company was not devoting enough resources to prepare for a possible future of “artificial general intelligence” (AGI) that could be smarter than humans.

Altman quickly responded, saying he appreciated Leike’s commitment to “safety culture” and added, “He’s right, we have a lot more to do; we are committed to doing so.” The company also confirmed to CNN which, in recent weeks, began to dissolve the team led by Leike, integrating its members into various research groups. A company spokesperson said this framework would help OpenAI better achieve its security goals.

OpenAI President Greg Brockman responded in a longer post on Saturday, signed with his name and Altman’s, detailing the company’s approach to long-term AI security.

“We have been raising awareness about the risks and opportunities of AGI so the world can better prepare for it,” said Brockman. “We have repeatedly demonstrated the incredible possibilities of scaling deep learning and analyzed its implications; we called for international AGI governance before such calls were popular; and we helped pioneer the science of evaluating AI systems for catastrophic risks.”

He added that as AI becomes smarter and more integrated into people’s everyday lives, the company is focused on having “a very tight feedback loop, rigorous testing, careful consideration at every step, world-class security and harmony between security and capabilities.”

ChatGPT outperforms doctors in making a diagnosis in consultation, says study