at the London summit, CEOs and managers face the challenge of regulation

“To see all the actors who participated in the rise of artificial intelligence brought together [IA] over the last ten years, it felt like we were living through a historic moment”, says Mustafa Suleyman, CEO of the start-up Inflection and co-founder of DeepMind, the AI ​​gem bought by Google and one of the first investors of which was Elon Musk. The surprise arrival in London of the CEO of Tesla and “first international summit on the risks of AI”, organized Wednesday 1er and Thursday November 2 by the British Prime Minister, Rishi Sunak. The joint presence, among the leaders of twenty-eight countries, of the Vice-President of the United States, Kamala Harris, and Chinese representatives, was seen as another small diplomatic victory for London.

Read also: Article reserved for our subscribers UK hosts first global summit on risks associated with artificial intelligence

In addition to the cast, the setting – the sumptuous Victorian mansion of Bletchley Park and its outbuildings, where Alan Turing’s team deciphered German codes during the Second World War – was also symbolic. Finally, the organization (informal exchanges and certain round tables behind closed doors, in a building not accessible to journalists) generated “very good quality of exchanges”recognizes Antoines Bordes, from the French start-up Helsing.

Despite this positive atmosphere, even among some initially skeptical participants, this unprecedented gathering also showed the challenge represented by the regulation of artificial intelligence, a year after the emergence of ChatGPT and a family of software capable of generating images and stunning texts.

Read also: At Bletchley Park, the secret history of the invention of computing

Among the announcements was the launch of a “AI IPCC”, responsible for forging a scientific consensus, as its equivalent does for climate change. Indeed, the writing of a report on the state of AI was entrusted to Canadian researcher Yoshua Bengio. However, the structure is not very clear because similar initiatives already existed. Mr. Bengio would be joined by researchers but would remain attached to the Scientific Advisory Council on AI created in October by the United Nations. In addition, some propose adding other structures, inspired either by CERN (European Council for Nuclear Research) or by the IAEA (the International Atomic Energy Agency, endowed with investigative powers). . Without commenting, James Manyika, senior vice president at Google, recognizes the need to establish “standards” devaluation.

You have 70% of this article left to read. The rest is reserved for subscribers.

source site-30