Artificial Intelligence: XL Budget, Training, Public Awareness…French Battle Plan Revealed

Don’t miss the train that’s already on its way. In an extensive report presented this Wednesday to the President of the Republic, a committee of experts in the field of artificial generative intelligence (AGI) called for the development of an ambitious national strategy to remain in the face of intense international competition. With a telling number: France must invest 5 billion euros a year for five years if it wants to be on par with the United States and China. Through 600 hearings, their work resulted in 25 recommendations to strategize and accelerate the development of generative artificial intelligence in France, epitomized in the public eye by the ChatGPT phenomenon.

This committee, created in September by former Prime Minister Elisabeth Bourne, brought together about fifteen experts on issues related to AGI, such as clean technology, and intellectual property law or its economic implications. “This is a realistic financial effort, this amount represents 0.3% of government spending,” said economist Philippe Aghion, co-president of the Generative Artificial Intelligence Committee, after the meeting at the Elysee Palace. “Despite commendable efforts, France currently invests three times less in AI than the United States relative to GDP,” he recalled.

Avoid Delays

“The risk is to be left without companies specialized in artificial intelligence and to see our companies (…) lose competitiveness,” while the government’s goal is to make France a major player in the sector. This should help accelerate economic growth, which the report estimates could double in the coming years thanks to AI.

“We have a chance to take a seat in this race because the United States only has one to two years of hindsight on these technologies,” says an Elysian adviser. Generative AI can become a “growth lever at the same economic level as industry GDP.”

France is building a robust higher education system in artificial intelligence, carbon-free energy and nuclear energy, which could support the creation of the data centers needed to use artificial intelligence models. And the comparative advantage inherited from centralization: billions of publicly available data useful for training and improving algorithms.

Education will also need to be carried out to dispel the perception that AI is destroying jobs. “Companies that adopt AI increase employment more than similar companies that do not,” insisted Philippe Aghion, although he acknowledged that there are inequalities between professions.

Beyond economic issues, the report highlights the need to raise awareness among the general public and provide training “immediately, massively and continuously” in AI-related jobs. He suggests, for example, developing teaching in higher education or encouraging teachers to use it. “Without this, our other recommendations will not be able to have their full effect,” said Anne Bouvereau, president of the board of directors of the École Normale Supérieure (ENS), who also chairs the committee.

Among these 15 personalities, we found such famous world figures as Yann LeCun, director of the Meta artificial intelligence laboratory, and Joël Barral, scientific director of Google. As well as representatives of the dynamic French ecosystem, such as Arthur Mensch, co-founder of the startup Mistral AI, champion of France and Europe in the field of artificial intelligence, and Cédric O, former Secretary of State for Digital Technology and consultant to Mistral AI.

Other main areas being explored include ensuring a country’s cultural influence by training models on French data, finding a balance between innovation and privacy, or even developing models that use less energy. Experts also want to make France and Europe a “major center of computing power” by promoting the installation of private computing centers necessary for the proper functioning of AI. “A start is necessary,” they warned, emphasizing Europe’s dependence on the rest of the world.