Scarlett Johansson published a scathing response to OpenAI allegedly stealing her voice for one of their new products. In a statement sent directly to NPR, Johansson said she was "shocked, angered, and in disbelief" after OpenAI created a voice called "Sky" for its generative product ChatGPT that sounds remarkably similar to hers.
After the statement, OpenAI both denied any wrongdoing and removed the voice from the product. At the same time, several high-profile employees have left the company citing safety and accountability concerns.
A Little Background
For those who don't know, OpenAI is a controversial company dedicated to bringing about artificial general intelligence, a concept that can most easily be summarized as "the type of humanoid AI you see in movies," and that many in the AI field disagree is even feasible. The company started out as a non-profit, but quietly shifted to a for-profit model several years ago.
OpenAI, under the leadership of Sam Altman, eventually released ChatGPT, an AI product that generates text based on different prompts. This, along with an image-generating counterpart known as Dall-e, are what ultimately brought OpenAI to the general awareness of the average person.
However, the technology was immediately controversial, both for creating potentially dangerous and incorrect text and for allegedly stealing from potentially millions of copyrighted works in order to "train" its model. OpenAI now faces several high profile lawsuits and government investigations.
Despite these troubles, OpenAI is valued at $80 billion and presses on with new iterations to its products. The latest is an audio "voice assistant" feature to go along with ChatGPT.
Where Scarlett Johansson Comes Into Play
Scarlett Johansson voiced an AI assistant in the 2013 movie Her. This movie is Sam Altman's self-proclaimed favorite movie, and one he is "obsessed" with. (Yes, it's a bit ironic that the movie is ultimately a warning against the very technology Altman aims to bring about).
When OpenAI debuted its new voice models within ChatGPT, one model, dubbed "Sky," sounded almost identical to Scarlett Johansson's voice. In response, Johansson sent a statement to NPR on May 20th, revealing that Altman had approached her about using her voice for Sky, but that she ultimately declined. Here is her statement, in full:
Last September, I received an offer from Sam Altman, who wanted to hire me to voice the current ChatGPT 4.0 system. He told me that he felt that by my voicing the system, I could bridge the gap between tech companies and creatives and help consumers to feel comfortable with the seismic shift concerning humans and AI. He said he felt that my voice would be comforting to people.
After much consideration and for personal reasons, I declined the offer. Nine months later, my friends, family and the general public all noted how much the newest system named "Sky" sounded like me.
When I heard the released demo, I was shocked, angered and in disbelief that Mr. Altman would pursue a voice that sounded so eerily similar to mine that my closest friends and news outlets could not tell the difference. Mr. Altman even insinuated that the similarity was intentional, tweeting a sing word "her" — a reference to the film in which I voiced a chat system, Samantha, who forms an intimate relationship with a human.
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 out there.
As a result of their actions, I was forced to hire legal counsel, who wrote two letters to Mr. Altman and OpenAI, setting out what they had done and asking them to detail the exact process by which they created the "Sky" voice. Consequently, OpenAI reluctantly agreed to take down the "Sky" voice.
In a time when we are all grappling with deepfakes and the protection of our own likeness, or our work, our own identities, I believe these are questions that deserve absolute clarity. I look forward to resolution in the form of transparency and the passage of appropriate legislation to help ensure that individual rights are protected.
The OpenAI Response
So, yeah. It certainly doesn't look great for a company that already has an alleged "steal first, ask questions later" attitude. But a Washington Post article notes that the company did start producing the Sky voice before ever meeting with Johansson and that she was not directly referenced as an influence prior to Altman meeting with her.
This, of course, only raises more questions. Why did Altman ask to meet with her if they already had an indistinguishable clone? Why ask her to reconsider two days before releasing a demo of what they already made? And of course, if there really is nothing to hide, why take the voice down?
The scenario certainly feels like another example of questionable ethics from a company whose own board fired Altman for allegedly being dishonest — before he was ultimately reinstated. And it may all just be a big coincidence, but it's not a good look alongside the list of higher-ups at the company who are choosing to leave.
OpenAI Losing Employees
OpenAI has now lost three high profile employees in the last week and five in the last few months. Ilya Sutskever, Jan Leike, and Gretchen Krueger all independently departed the company recently after concerns about safety and accountability.
Sutskever was chief scientist and co-founder, and Leike was another executive. Together the pair headed the now-disbanded "Superalignment" team, which was tasked with general safety, protocols, and making sure any developed artificial general intelligence couldn't become dangerous.
In response to these recent resignations, OpenAI basically started enforcing unusually strict non-disclosure agreements. According to Vox, the company has begun employing hardball tactics to force former employees into staying quiet. You know, just the kind of behavior you want from a company trying to create humanoid artificial intelligence.