I follow everything about artificial intelligence because it’s my job. But I’d check out AI news even if that weren’t the case. And we have certainly had plenty of AI developments in the past few weeks.
First, ChatGPT got the massive GPT-4o upgrade. Google unveiled even more ways it’ll integrate Gemini into its products, even if that ruins traditional search experiences for many people. Microsoft doubled down on Copilot by unveiling a slate of AI PCs and an already controversial Windows 11 feature that can recall everything you did on your PC.
But the biggest AI development in the past few weeks is the ChatGPT scandal with Scarlett Johansson. It clearly seems like OpenAI is in the wrong here, assuming the actress’s claims are accurate. But it still turned out to be a smart marketing move by the company, as it trumps everything else happening in AI tech right now.
To briefly recap, Johansson announced earlier this week that Sam Altman reached out to her in September to hire her to voice ChatGPT. The actress declined. OpenAI contacted her again a couple of days before the big GPT-4o demo last week that involved a presentation of ChatGPT’s voice capabilities. Before she could answer, GPT-4o and its Sky voice was released.
Tech. Entertainment. Science. Your inbox.
Sign up for the most interesting tech & entertainment news out there.
By signing up, I agree to the Terms of Use and have reviewed the Privacy Notice.
That’s when she learned about the similarities between Sky and her voice:
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 single word, “her” – a reference to the film in which I voiced a chat system, Samantha, who forms an intimate relationship with a human.
You should rewatch the Sky GPT-4o demo below if you haven’t already seen it to understand why the actress is upset:
Better yet, listen to Sky reading Johansson’s statement and you can immediately tell how blatantly obvious the similarities are.
Here’s the official statement released by Scarlett Johansson, detailing OpenAI’s alleged illegal usage of her voice…
…read by the Sky AI voice, because irony. pic.twitter.com/cJDlnA0hTP
— Benjamin De Kraker ️ (@BenjaminDEKR) May 20, 2024
As someone who uses ChatGPT and follows all the developments, I wish OpenAI chose a different route rather than going through a scandal that only exacerbates its behavior concerning personal data and copyrighted content when training its AIs. But maybe this was a calculated move from OpenAI, one meant to draw tons of attention to ChatGPT.
Think about it for a second. OpenAI engineered a two-pronged marketing attack. First, the company announced a last-minute AI event on the eve of Google’s AI-centric I/O 2024 keynote. It then demoed the GPT-4o upgrade that Google had no answer for. Well, Google has a similar multimodal assistant in the works, and we saw a demo of it. But it won’t be available until sometime later this year in Gemini, while GPT-4o is out now and the new voice experience is coming in a couple of weeks.
However, only people who use AI products and care about them would have followed these developments. Meanwhile, the Scarlett Johansson fiasco puts this news in front of countless more people. That’s the second part of the marketing tactic. I’m speculating here, as I do not know whether OpenAI planned for this specific course of events. We’ll never truly know.
People who don’t use ChatGPT or don’t follow all the AI news will have heard by now that a tech company’s AI is so good that it cloned Johansson’s voice. OpenAI did it so well that the actress has hired lawyers to deal with the matter, and she already got a first win. The Sky voice is now gone from ChatGPT, although OpenAI says it was trained on the voice of a hired voice actress, not on Johansson’s voice.
Even without the actress reacting, people were already comparing Sky to Samantha’s voice from Her. This must have been the intention of OpenAI all along.
Now that Johansson has threatened legal action, this sort of PR mess will spark additional interest in ChatGPT. Some people will want to see GPT-4o’s voice capabilities in action to decide if they’re as good as they seem. Others might start using ChatGPT because of it. At the very least, OpenAI will make it into even more conversations. That’s the kind of promotion OpenAI needs at the expense of Google and even Microsoft, which is a close partner.
This is the kind of promotion that money can’t buy.