Get involved as a citizen and help shape the future of our democracies

Interview with Raphaël Lasseri, PhD in Physics, founder of Magic LEMP

How did your time at ENS Paris-Saclay prepare you to create your own start-up in the field of artificial intelligence?

My training at ENS Paris-Saclay was fundamental in my journey. It provided me with a solid foundation in science and taught me the importance of interdisciplinarity, essential in the field of AI. The ability to combine technical knowledge with an understanding of societal and ethical issues prepared me to venture into artificial intelligence, with a particular focus on applications that can strengthen our democratic fabric and always a consciousness of the public interest. ENS is also a breeding ground for ideas, talents, and associations (I think of Cr@ns in particular) that allowed me to learn a lot humanly, technically, and even administratively. Finally, it was also on the school benches that I met those who later became the first to join the Magic LEMP adventure.

Can you tell us about Magic LEMP?

Magic LEMP is the result of a different vision of the researcher's role. Today, we have two main areas of development. On the one hand, supporting companies by offering custom AI solutions, accompanying them both technically and strategically to make the right choices and uses. Our very "normalien" difference: scientific excellence in the design and development of solutions, as well as a strong emphasis on training and popularizing solutions to users. On the other hand, our second focus is developing AI technology blocks to better represent and understand heterogeneous, multimodal data, allowing for simplified and better representation of complex systems and issues.

What challenges have you faced in developing a company focused on artificial intelligence?

The main challenges lie in maintaining a balance between technological innovation and respecting democratic values. This includes ensuring transparency, privacy protection, and fairness of our solutions. Moreover, it is crucial to work closely with institutions to ensure our tools truly meet citizens' needs and strengthen the trust between them and the institutions, even in areas like security or defense where the boundary is always thin. Practically, one of the main challenges is to ensure that AI is used ethically and responsibly, respecting democratic values such as transparency, fairness, and privacy protection. It can be difficult to reconcile these values with the technical and practical requirements of AI, as well as stakeholders' expectations.

Another major challenge is ensuring that AI is accessible and understandable to all citizens, regardless of their technical knowledge. This requires designing user-friendly and intuitive interfaces and providing clear and comprehensive documentation. It is important to raise citizens' awareness of the benefits and limitations of AI so that they can make informed decisions about its use.

Finally, developing an AI-focused company in a democratic context requires working closely with public institutions and stakeholders. This can be challenging, as decision-making processes in the public sector can be slow and complex, and it can be difficult to convince decision-makers of AI's usefulness and relevance.

Do you think your start-up already has a positive impact on democratic processes?

Yes, Magic LEMP has had a significant positive impact on democratic processes. For example, we collaborated with the Île-de-France region to develop an AI version of the application "Ma Région, ma Commune," which uses AI and generative AI approaches to facilitate access to and understanding of public texts and decisions, often written in complex ways and not well-suited to the general public. Thanks to this application, citizens can follow the region's projects and initiatives in real time, participate in public consultations, and give their opinions on regional policies. But that's not all: we also integrated a tool to quantify the impact and outcomes of public policies, thus enabling intelligent, data-centered decision-making. By making information more accessible and facilitating citizen participation, we have helped strengthen the transparency and accountability of regional institutions. Citizens can now better understand regional issues and actively participate in the democratic life of their region. And this is just the beginning: our goal is to scale up nationally and make public information accessible to all French citizens.

Another notable example of our positive impact on democratic processes is our platform Pluralisme.fr. By providing real-time tracking of politicians' speaking time and all political positions, we promote pluralism and diversity of voices in the public debate. We also help citizens better understand political issues and make informed decisions during elections.

Finally, by combating misinformation, fake news, and deep fakes, we help strengthen citizens' trust in democratic institutions. We are convinced that our work can help reinforce democracy and encourage more active and informed citizen participation.

How do you envision the future evolution of your company?

Among other things, we aim to continue developing innovative and useful AI solutions to strengthen democratic processes. We also want to continue our collaboration with public institutions and companies to promote responsible and ethical use of AI. We are indeed convinced that AI can play an important role in promoting transparency, accountability, and citizen participation, especially in such uncertain times for our democracies. We must invest in the informational space and fight with all available tools to preserve our democratic values. In particular, we plan to develop AI tools to facilitate access to public information, encourage citizen participation, and improve democratic decision-making. We also want to explore new areas of AI application in the democratic context, such as combating misinformation and promoting the diversity of voices in the public debate.

What role do you think large companies should play in the responsible use of artificial intelligence to strengthen democracy?

We believe they have a key role to play. As mentioned earlier, large companies have the responsibility to develop and use AI ethically and responsibly, considering social and ethical issues. However, innovation and research should not be stifled! Certainly, it is a balancing act, but the current version of texts like the AI Act is not only poorly adapted but can pose a real risk of hindering the innovation of our companies while favoring the growth of our foreign competitors, many of whom do not share our ethics or respect for democracies!

Do you have a message for other young innovative entrepreneurs?

I would tell them this: be bold, creative, and persistent. Do not be afraid to take risks and explore new areas of technology application. But above all, stay true to yourselves! For former ENS students, don't hesitate to embrace your intellectual heritage, your identity as researchers; there is no need to conform to traditional start-up codes and master the business plan, pitch, and deck perfectly to succeed and to succeed on your own terms.

Finally, get involved! Not necessarily politically, but as citizens, you can, by undertaking, become actors and shape the future of our democracies concretely, and right now, it is urgent to respond to this call.