In 2013, Spike Jonze's sci-fi romance film "Her" captivated audiences with a story about a man falling in love with an AI assistant voiced by Scarlett Johansson. Just 11 years later this film seems to have become an uncanny reality.
OpenAI, the company behind the massively popular ChatGPT, has said that they will work to ‘pause’ their voice assistant model ‘Sky’ after users repeatedly pointed out the similarity to the Her actress's voice. The company has stated that any similarities are unintentional, however many remain unconvinced.
‘Her’ Film Confirmed to Have Inspired ChatGPT
During the Dreamforce conference in 2023, Sam Altman, OpenAI’s CEO, stated that Her is one of his favourite films and in fact went as far as to say it ‘inspired’ and ‘influenced’ the company.
"The number of things that I think "Her" got right that were not obvious at the time, like the whole interaction model with how humans are gonna use an AI this idea that it is gonna be this conversational language interface that was incredibly prophetic and certainly more than a little bit inspired us [...]. It's not just like a prophecy, it's like an influenced shot,” the CEO told Marc Benioff during a panel at the event.
This raised tech enthusiasts' eyebrows for the blatant similarity when the updated Voice Mode was publicized during a GPT-4o update live stream event.
What is ChatGPT Voice Mode?
ChatGPT’s voice mode is a feature that was initially introduced in September 2023 and updated in May in 2024 for GPT-4o. It allows users to interact with the AI through voice instead of just text. Users can speak out loud to ChatGPT and it will respond using a text-to-speech model, creating a more natural conversation and aiding accessibility for users who have trouble typing.
Users can choose from a variety of pre-recorded voices, named ‘Breeze, Cove, Ember, Juniper and Sky’, each of which has a ‘distinct personality’. OpenAI says the advanced voice features that were demonstrated during the live stream update event would be made available to ChatGPT Plus subscribers in ‘the coming weeks.'
OpenAI Denies Intentional Use Of Scarlett Johansson's Voice
OpenAI states that they selected the voices for their Voice Mode by working with ‘industry-leading casting and directing professionals to narrow down over 400 submissions before selecting the 5 voices.’
The company actively denies that the voice was chosen to mimic the voice and experience of Her, directly addressing in their post about how they chose the voices, explaining that ‘Sky’, the Voice Mode persona said to be similar to the star voice, was ‘not an imitation of Scarlett Johansson but belongs to a different professional actress using her own natural speaking voice.’
The AI giant goes as far as to bring up the ethics of AI in no uncertain terms, stating that they ‘believe that AI voices should not deliberately mimic a celebrity's distinctive voice.’
Read: What are Deepfakes and Why are They Dangerous?
However, it is difficult to believe that this was not noticed at any stage in the development process or was, in fact, an intentional choice. This flame was fanned by CEO Sam Altman, who seemingly referenced the film soon after the Spring Updates event in a post on X.
her
— Sam Altman (@sama) May 13, 2024
The withdrawal of the Sky voice mode highlights the ongoing conversation around AI ethics.
As AI technology continues to develop and integrate more realistic human-like features, questions arise about responsible development and user trust which must be addressed directly by companies as well as by implementing XAI tools.
OpenAI has not confirmed that the Sky persona will be offline indefinitely, and it seems unlikely that it will not make a reappearance in some form given Altman’s obvious enthusiasm for film.
