Sam Altman, the father of ChatGPT, is in Paris. The CEO of OpenAI, the artificial intelligence (AI) company that created the now famous conversational robot, meets Emmanuel Macron this Tuesday afternoon May 23 and has an appointment with the Minister of the Economy, Bruno Le Maire . He should also see the Minister Delegate for Digital, Jean-Noël Barrot. These interviews are part of a vast “world tour” launched last week by the leader of the American start-up: after passing through Toronto, Washington, Rio, Lagos and Lisbon, this series of diplomatic visits continues this week with Madrid, Warsaw, Paris, London and Munich.
The reason for this institutional trip is the multiplication, all over the world, of calls to regulate AI in general, and in particular generative, which creates texts or images, and of which OpenAI is one of the champions. The Elysée also intends to discuss with Mr. Altman the development of AI in France and Europe, but also the supervision of the sector.
In Paris, Bruno Le Maire intends in particular to mention two avenues of supervision during his interview with Mr. Altman: first, the need to remunerate, in one way or another, data and sources, images or texts, which are used to train generative AI models. Certain press organs, but also photographers or illustrators, as well as computer coders, have indeed expressed concern that software feeds on their content, protected by copyright, in order to then produce texts and images that are almost free and likely to replace their work…
A posture of good will
In addition, the Minister of the Economy intends to defend with Sam Altman the need to indicate when content has been produced by an AI. The idea, already pushed by some, is to distinguish these texts and synthetic images from those made by humans. “These are two avenues we are working on”, we say to the minister’s office, which recently embarked on a phase of “consultation” about AI and has already spoken with Mr. Altman over the phone.
On Monday, Sam Altman met with the president of the Spanish government, Pedro Sanchez. And the previous week, he had, in Washington, lived for the first time the test of the parliamentary hearing, in front of senators.
“We don’t want to repeat the mistakes of the past”had launched Richard Blumenthal, the Democratic chairman of the Committee on Legal Affairs, advocating to regulate AI from its inception, unlike social networks and the Web… He had listed the problems posed by this technology: “risks of bias and discriminatory algorithmic decisions”, ” disinformation “, ” lack of transparency “, “job losses”… But the general tone was less hostile than in past hearings of executives from Facebook (Meta), Google or Amazon.
You have 36.88% of this article left to read. The following is for subscribers only.