What to remember from Villani's report

Spring 2018. AI cannot be a new machine to exclude.

Interview with Cédric Villani, French mathematician, Fields Medal and La République en Marche MP for Essonne, in le Monde, on March 28, 2018, by Morgane Tual and Vincent Fagot.

For a total or partial reproduction of the article Le Monde, contact syndication@lemonde.fr.

Excerpts from the article

Cédric Villani delivered, on March 28, his report on artificial intelligence (AI), commissioned by Prime Minister Edouard Philippe: 420 experts interviewed, the report deals with various issues and offers a series of recommendations.

An ethical artificial intelligence

AI cannot be a new machine to exclude”, writes Villani. Between the biases of certain programs – which reproduce racism and human sexism for example –, the lack of diversity in research teams and the social risks that these programs can represent, AI can cause concern. He calls for “new methods of production and distribution of value”. The report also insists on the need for feminization of AI – women are only 10% in computer science engineering schools. “This lack of diversity can lead algorithms to reproduce biases”, explains the document, which recommends an “incentive policy” aimed at achieving a rate of 40% of female students by 2020. More broadly, the MP recommends the establishment of an AI ethics committee, responsible for organizing public debate and providing opinions and recommendations “in complete independence”. He also wants public authorities to be able, “in the context of legal proceedings or upon referral to the Defender of Rights”, to examine these “black boxes” to understand them.

Unquantified proposals

Mr Villani’s report offers a large number of very concrete recommendations, but how to finance them? The document does not advance any figures. “We made these calculations, and rather three times than once”, assured Mr. Villani to le Monde. “But we didn’t want them to be in the report to avoid public expression rushing on them”. (…) On the occasion of a summit on AI, organized at the Collège de France in Paris this Thursday, the President Emmanuel Macron will present his plan for artificial intelligence.

There is a question of sovereignty, which concerns all of Europe.

Interview with Cédric Villani in le Monde, on September 9, 2017, by Morgane Tual.

For a total or partial reproduction of the article Le Monde, contact syndication@lemonde.fr.

Excerpts from AI is everyone’s business

The government has tasked the mathematician and LRM deputy Cédric Villani with a new mission on artificial intelligence. His mission is to draw a roadmap for the government in the years to come… with the idea that AI should not be seen as a specialized field, but as everyone’s business.

Nathanaël Ackerman, the coordinator of the France AI report, which planned an envelope of 1.5 billion euros, will be closely associated with [our] work… he shows great enthusiasm.

The France AI 🇫🇷 report was published in March 2017 by the ministry of Economy and Finance and the ministry of Higher Education and Research, less than 3 years after the appointment of Macron as Minister of the Economy, Industry and Digital in 2014.

Do you already have any priorities in mind for France’s action in this area? A very important issue is how to ensure that artificial intelligence benefits everyone, is associated with a strengthening of democracy and not the opposite. A number of examples show that in some cases the use of AI can have devastating effects on economic issues and the fabric of democracy. A book made a big impression on me, that of Cathy O’Neil, Weapons of Math Destruction (2016, untranslated), which shows that AI is already benefiting very strong companies, leads to social regressions, errors, is used for advertising targeting for purposes that are not acceptable…

March 2023. Are we still in a democracy? Blast, le souffle de l’info.

How can France compete against giants like GAFAM, which are investing massively in AI and hiring the best experts, particularly French? France is at the forefront in artificial intelligence research. We also have institutions like Inria, CNRS, researchers like Yann LeCun… I have other colleagues who went to Facebook, Google… It will be one of the major questions in the report… There is also a question of sovereignty, which concerns all of Europe… who leads projects…

March 2023. ChatGPT: Why the city of Montpellier prohibits its agents from using artificial intelligence. France-3 Regions.
March 2023. ChatGPT banned in Italy: the first domino falls in Europe. La Tribune

Where will we find the necessary money? I’m not clear on this, it’s a major problem…. On some projects , Europe has shown itself capable of investing phenomenal amounts, including for very audacious scientific projects, such as the Human Brain Project. But under very complicated conditions, with considerable governance problems… It is also a question of trust.

Earlier this month, Vladimir Putin declared that the country “that will become the leader in this sector [AI] will be the one that will dominate the world”. This shows the enormous expectations that are placed on these new technologies… It may also be that the legal environment plays an important role in whether there will be an upheaval or not… the law can influence the speed, the violence or the ease of a revolution.

Elon Musk, the boss of Tesla, echoed the statement by saying that “States’ race for AI superiority would likely cause World War III.” It’s fundamental for his business model.


How did the share of women in high school in math evolve since 2018 in France?

It dropped back to 1994 levels. L’étudiant.

Who invaded Ukraine on February 24, 2022?

Vladimir Poutine.

Who is behind Facebook’s AI model announced on February 24, 2023?

Yann Lecun.

Who is behind the ChatGPT model used on March 21, 2023 at the National Assembly to amend the law?

Elon Musk. BFM Business. March 21, 2023.