You have to follow the incredible showdown between Scarlett Johansson and OpenAI on the alleged cloning of the actress's voice, you have to do it because it can teach us a lot. It's the perfect microcosm of the philosophy behind the generative AI gold rush: an unstoppable future built on data collected online without the consent of creators and copyright holders.
A future where society's rules don't apply, because the stakes (a hypothetical super-intelligence that would rewrite the course of human history) are too high to worry about details like attribution and privacy. Welcome to the world of OpenAI, where the fate of AI is already written, whether you like it or not.
The bare facts: the voice of Scarlett Johansson and the message of OpenAI
According to what was reported by the lawyers of the American actress, the CEO of OpenAI Sam altman he allegedly contacted her nine months ago to license her voice for a new digital assistant. Scarlett Johansson would have refused: just two days before the system launch event GPT-4o, Altman allegedly insisted that the actress reconsider the proposal. Although Johansson never gave her consent, OpenAI debuted with Sky, a program with a voice very similar to that of the actress. The name itself, Sky, has a similar sound to Scarlett (in OpenAI they ate bread and fox).
Altman's response to the "cloning" accusations? Rather ambiguous: while denying having cloned Johansson's voice, he admitted that he would have removed Sky's voice "out of respect" for the actress. An ironic post on X, in which Altman referenced the film “Her” (in which Scarlett Johansson voices an AI), shows that the CEO was well aware of the similarity.
The “messianic” vision of OpenAI
Beyond specific responsibilities, the Johansson-OpenAI case is emblematic of the mentality that drives the development of generative AI: a technology built on data collected online without the consent of creators and copyright holders, in the name of a hypothetical greater good.
OpenAI's stated goal has always been to create aAGI (Artificial General Intelligence) which would lead to an unthinkable revolution in productivity and prosperity. A utopian world where work disappears, replaced by a universal basic income, and humanity makes quantum leaps in science and medicine. (If you want the dystopian version, machines cause the end of life on Earth as we know it. We are for one in between, we root for Protopia).
In this messianic narrative, the stakes are so high that they justify any means to accelerate progress. As Altman explained in an interview, the development of AI is a geopolitical race against autocracies like China. Better to root for the success of OpenAI than for that of "authoritarian governments", he argues.
An OpenAI engineer, Jeff Wu, was even more explicit in a video last year: “It's deeply unfair that a bunch of people can just build AI and take everyone's jobs away, and in a way there's nothing you can do about it. stop them right now,” he admitted. Then he added: “Join us and you will have one of the few jobs left.”
Scarlett Johansson (and the rest of the world), hear? “There's nothing you can do.”
This is the crude, paternalistic logic of OpenAI. That such a small group of people is tasked with building civilization-changing technology is inherently unfair, they admit. But they will move forward anyway, because they have both a vision of the future and the means to try to make it happen. Wu's proposal is revealing: you can try to fight, but you can't stop them. Your best bet is to get on board.
The same dynamic is reflected in the content licensing agreements that OpenAI has signed with platforms such as Reddit and publishing groups. One tech executive likened it to a hostage situation: AI companies will still find a way to “comb” publishers' websites if they don't cooperate. Better to get a measly rate while they can.
Scarlett Johansson's accusations only confirm these suspicions. If true, Altman's reasoning for commissioning the actress' voice would have been that her familiar timbre could be "comforting" to those who find AI assistants disturbing. Not an aesthetic, but a recruiting tool for a technology that many didn't ask for and seem to distrust. The excellent Pier Luigi Pisa on LinkedIn he summarized this mindset in an admirable way (If you don't follow it, start doing it). Once again, OpenAI's logic is this: move forward, consensus or not, because the stakes are too high to change course or wait.
The hubris of generative AI
Hubris and a sense of entitlement are intrinsic to the development of any transformative technology. A small group of people must feel confident enough in their vision to bring it into the world and ask the rest of us to adapt. What did the other guy say? Be hungry be crazy? Here, this. I don't know how crazy these guys are, but the hunger is evident.
Generative AI pushes this dynamic to the point of absurdity, however. It is a technology that requires a manifest destiny mentality of domination and conquest. If you believe that the future belongs to you, you are not stealing: you are building it. Intercom Indian tribes.
But is this really the future we want? A future in which a handful of tech companies unilaterally decides to overturn society, bypassing consent, copyright and privacy in the name of a singularity yet to be demonstrated?
The Scarlett Johansson-OpenAI case presents us with a choice: passively accept the "manifest destiny" of AI, or demand that its development is subject to rules, limits and responsibilities like any other technology.
Because if it is true, as OpenAI claims (and I believe it is probable), that AI will change the course of human history forever, it is even more important to ensure that it does so while respecting the rights and will of everyone, not just who has the power and means to impose it. Otherwise, the risk is that the artificial intelligence revolution turns into a new form of oppression and exploitation. And that the bright future promised by OpenAI turns out to be, in reality, a waking nightmare.
