Q&A: France’s Top Mathematician-Turned-Politician Discusses Artificial Intelligence

Fields Medal winner Cedric Villani on experimentation, sharing and sovereignty

France’s best-known scientist decided to shift to the policy arena to serve decision-makers with the expertise they desperately need and reconcile politics with science, tech and innovation. - Credit by Getty Images

When he was awarded the Fields Medal—widely regarded as the Nobel Prize for mathematics—in 2010, little did Cedric Villani know that he would embark on a political journey half a decade later.

A staunch proponent of European integration, the celebrated mathematician joined French president Emmanuel Macron’s political movement in 2016, adhering to its progressive voice and pledge to champion multilateralism and curb the populist tides that had swept the continent. Against all odds, Villani, who wears velvet cravats and gives Ted Talks about how sexy math is, ran for office on an ‘En Marche’ (Macron’s party) ticket and unseated a long-standing politician.

France’s best-known scientist decided to shift to the policy arena to serve decision-makers with the expertise they desperately need and reconcile politics with science, tech and innovation. A few months into his current mandate, he authored a report on French artificial intelligence commissioned by Macron, outlining France’s long-term strategy to become a leader in the field, overhaul digital innovation around AI and make sure labor-market and technological disruptions support inclusivity and diversity.

Adweek met with Villani to discuss his vision for French AI, his approach to integrating it across Europe and plans to lure foreign players.

Adweek: It’s a rare thing to see scientists get involved in electoral politics. What prompted you?
Cedric Villani: I have always been interested in politics, and I have published countless white papers about a host of policy issues. But indeed, it wasn’t natural to run for office and seek public representation. The previous administration had tried to offer me cabinet positions, but I had declined. This time, I really felt aligned with Macron’s vision for Europe and approach to propel France into the 21st century, finally.

Also, when I’m looking at the pressing issues facing France, Europe and the planet, I had the feeling that our politicians weren’t up to the task. We needed additional technical expertise. And on the campaign trail, Macron had shown a deep understanding of future environmental, technological and social trends.

But beyond that, there is this idea that scientists and engineers should be at the service of policy makers to guide their decisions. For example, in the context of my work on AI, I’m now contributing to a new report on hyperloop transport technologies, an area on which many elected officials are poorly informed. Overall, working in politics is fascinating. It’s one of the only fields where you can tackle and have an impact on a series of issues. It’s also a real honor to represent my constituents, their interests and their aspirations.

What’s the main takeaway from your report?
Our priority is research and development. It’s where competition is the fiercest, with Asian countries, notably. We’re also fighting for AI talent, scientists and engineers, which will become tomorrow’s raw commodities.

France’s AI strategy is now centered around three pillars: experimentation, sharing and sovereignty. And these will underpin policies and investments. To achieve our goals, we’re setting up ethics committees to regulate and monitor AI developments, training programs to build the right workforce around AI and interdisciplinary learning at the university level to breed new talent. It’s an action plan that will be deployed over the next few years.

Are these investments up to the task ahead?
Our investment plans are between $1 billion and $10 billion, depending on whether you factor in public-private partnerships. At first glance, it’s not much compared to what is being allocated by China or Germany, but there are a few things to consider. Those investments will only target research, development and innovation. These investments will be complemented by further action on education, employment and training, and they will be subject to change as trends in AI evolve.

What will matter more in the near future is what will be done at the European level to support European champions capable of competing with China and the U.S. But AI is not just a question of investment. It’s a question of political will. And Emmanuel Macron is the first French president who is thoroughly committed to finally breeding a culture of innovation. This strategy will work if we foster synergies between the private sector, policy-makers, higher education institutions and so on.

How will you measure the success of this program?
We haven’t set tangible KPIs, but we will examine how many jobs we can create, how much job destruction we can avoid through training, how many top tech firms in the field we can elevate on the global stage, how many startups have will have emerged through our support and so on.

Thanks to this roadmap, how will France’s top tech, media and telecoms firms be involved?
Bringing leading businesses on board is the priority underpinning this roadmap. There is a vast awareness and training challenge ahead of us. We also need to mobilize our research institutes, which will play a key role. Right now, AI research institutes are scattered across the country. That’s where innovation will start. These research institutes will then serve the private and public sector to inform training and policy and bolster national champions. We are remodeling research around the private sector’s AI needs. That’s the revolution.

France lacks a giant such as Apple or Google. Can this program shape the rise of such a player?
Again, it’s a question of scale. We need European actors using European technology. I don’t know if we can breed giants such as Apple or Google in the near future, but if it ever happens, it will take shape at the European level. We have different economic structures, models and consumption patterns.

But it is vital to support our won leaders in tech and telecoms such as Orange to equip them with homegrown technology. There is this narrative around the GAFAs (Google, Apple, Facebook, Amazon) being independent, but they were also supported by public research, public policy and protectionist measures such as the Buy American Act.

Your report focused on four key sectors: transport and security, health, the environment and defense. Why?
We are emphasizing these sectors in which we have national champions, but also because it is urgent for us to forge the right industrial and tech policies to support them. These are sectors where we need to focus on experimentation because AI is already playing a key role. In trailing industries, we need to invest more. For instance, our financial sector will not complete its AI transition without our help. Of the four sectors you mentioned, the environment is the most interesting. It highlights President Macron’s commitment to curbing climate change, notably by using AI to usher in a new era for sustainable agriculture.

The FrenchTech label has been gaining resonance worldwide. Are you working closely with that ecosystem?
FrenchTech is a quality label and an extraordinary ecosystem with which we will be working very closely. We have seen the success of French Tech labeled firms in startup and tech fairs around the world. Our role is really to guide and help these French AI and tech firms raise funds and gain international recognition.

President Macron and his administration have hammered out the message “France is back.” What do you have to say to foreign actors looking at French AI and French innovation?
Yes, we are obsessed with restoring France’s standing in the business and tech communities worldwide. We have a bastion of startups with state of the art technologies and innovation that deserve the attention of foreign investors and risk capital. But they were undermined by previous policies. So we’re gaining momentum. Macron has also signaled to the world that our AI and economic strategies and innovation policies are part of a greater plan for European cooperation.

AI comes with countless opportunities, but also many threats. How are you trying to address the underlying dangers of AI?
Protecting citizens from the dangers of AI is also a key priority. On the one hand, we need to anticipate employment trends and protect and retrain those who will lose their jobs. On the other hand, we have to address data protection. That’s why we’re establishing Ethics Committees that will be responsible for overseeing and regulating the impact of AI on data privacy. The GDPR is an excellent first step. Finally, we need to develop better expertise on the regulatory side. Regulators are not trained on AI and are ill-equipped to anticipate looming threats, audit algorithms and so on.

What’s your vision for AI training? How do we retain talent?
My blueprint recommends a three-tiered approach. First, train and support leading talents, engineers and researchers to build an AI elite. For that, we also need more public-private partnerships like those that can be found in the U.S. so that we can have top engineers working on AI innovation in French universities but also supporting startups on tech firms.

Then, we need to train and develop a technical workforce through apprenticeships, professional degrees and so on. These are the jobs of the future that will assist tomorrow’s most qualified engineers.

And finally, we need to teach people who don’t have technical backgrounds on how to use AI tools in their professions: lawyers, teachers, PR professionals, consultants, etc. We need European cooperation to retain talent.

We need a European network for AI to allow students and professionals to train in different countries. We must inspire young talents to develop something fabulous at home. In a sense, we need to breed a form of European solidarity towards AI and a patriotic spirit of innovation and science.

The Germans, through their research centers, have already carved a path to achieve this objective. Consider leading figures in the field—such as Yann LeCun in France or Yoshua Bengio and Geoffrey Hinton in Canada—who are icons in deep learning; they decided to support research and innovation at home when the largest multinational firms were trying to hire them.

Justin Trudeau told me that Bengio aims to stay in Canada to make his country a leader in the field. But for now, these are exceptions. If a French talent wants to develop a homegrown model for autonomous vehicles, we need to provide the resources, databases, public data and support for him to accomplish his goals.

Alexis Chemblette is a Brooklyn-based reporter and contributor to Adweek.
Publish date: July 23, 2018 https://dev.adweek.com/digital/frances-top-mathematician-turned-politician-discusses-artificial-intelligence/ © 2020 Adweek, LLC. - All Rights Reserved and NOT FOR REPRINT