In May, Altman’s company released its latest generative AI model called ChatGPT-4o, which came with a setting named Sky. It so happens that Sky sounded very familiar to Johansson — “eerily similar,” according to the actress.
If OpenAI CEO Sam Altman fancies trying his luck on the big screen, there may be an opening for him.
The head of the world leader in generative AI has made a formidable enemy out of Marvel’s Black Widow herself, Scarlett Johansson, who has suggested that Altman could have potential as a big bad.
Johansson recently spoke to The New York Times about technology and Altman, confirming that he could make a good Marvel villain, “maybe with a robotic arm”.
The quip comes against the backdrop of an ongoing feud between the two figures, centred around the alleged nonconsensual usage of her image and voice.
In May, Altman’s company released its latest generative AI model called ChatGPT-4o, which came with a setting named Sky. It so happens that Sky sounded very familiar to Johansson — “eerily similar,” according to the actress.
Altman posted "her" on X, which many saw as a reference to Johansson’s role as an AI operating system in the 2013 Spike Jonze movie, Her.
Johansson released a statement saying that she was “shocked” and “angered” that OpenAI mimicked her voice when she originally turned down the company’s offer to be the voice of its ChatGPT system. She added that Altman used a voice so close to her own that “friends and news outlets could not tell the difference”.
According to Johansson, Altman reached out to her last September about the company hiring her to provide the voice for ChatGPT-4o. She declined for “personal reasons,” but Altman allegedly contacted her agent and “asked me to reconsider”.
“Before we could connect, the system was out there.”
She confirmed her lawyers contacted OpenAI to have the voice of Sky pulled down.
OpenAI ended up complying amid backlash but maintained that “Sky’s voice is not an imitation of Scarlett Johansson but belongs to a different professional actress using her own natural speaking voice.”
Scarlett Johansson with her husband Colin Jost, attending the world premiere of 'Fly Me to the Moon'
In the NY Times interview, Johansson said she had “actively avoided being a part of the [AI] conversation, which was what made it so disturbing. I was like, ‘How did I get wrapped up in this?’ It was crazy. I was so angry.”
“I felt I did not want to be at the forefront of that,” Johansson said about turning down OpenAI’s original request. “I just felt it went against my core values. I don’t like to kiss and tell. He came to me with this and I didn’t tell anybody except my husband.”
“I also felt for my children it would be strange. I try to be mindful of them,” Johansson added.
OpenAI halts using ChatGPT's Sky voice after Scarlett Johansson said it was 'eerily' like hers
ChatGPT, Grok, Gemini and other AI chatbots are spewing Russian misinformation, study finds
The actress went on to call deepfake technology a “dark wormhole you can never climb your way out of”.
“Once you try to take something down in one area, it pops up somewhere else. There are other countries that have different legislation and rules. If your ex-partner is putting out revenge, deepfake porn, your whole life can be completely ruined.”
“I think technologies move faster than our fragile human egos can process it,” she concluded, “and you see the effects all over, especially with young people. This technology is coming like a thousand-foot wave.”
Conjuring thousand-foot-waves could be one of the evil Dr. Altman’s superpowers, should he or Marvel get any ideas following Johansson’s comments.
Johansson currently stars in the rom-com Fly Me to the Moon and recently confirmed that she will be starring in the latest Jurassic World entry.
