Public debate

Laurence Devillers

Robots must respect our human values for our well-being.

Laurence Devillers specialises in the application of robotics to the social sciences and encourages our societies to examine the risks of emotional artificial intelligence.

Published on 17/07/2019

2 min

After studying computer science at the Université Paris-Sud, Laurence Devillers began teaching computer science and artificial intelligence at the Université Paris-Saclay and then at the Sorbonne, while also working as a researcher at the Digital Laboratory for Mechanics and Engineering Sciences (LIMSI) at the French National Centre for Scientific Research (CNRS).

In an environment which remains predominantly male, Laurence Devillers is a recognised leader in her field: the emotional and social dimensions of spoken interactions.

For this researcher focused on human-machine interaction, it is crucial that robots’ empathic behaviour is guided by ethical commandments. Although Laurence Devillers is not against these artificially intelligent machines, she nevertheless warns us that these pseudo-emotional objects have a real societal impact.

Already present in our daily lives, social robots and other home assistants like Google Home or Siri are in the process of being improved in order to detect emotions in the human voice. With this analysis of non-verbal language, the risks associated with attachment to machines and the potential for manipulation become ever greater. The professor thus feels that vigilance is called for.

In her book Robots and Humans: Myths, Fantasies and Realities (“Des robots et des hommes : Mythes, fantasmes et réalité” Edition Plon, 2017), Laurence Devillers invokes fiction, from Mary Shelley to Wall-E, Isaac Asimov and Goldorak, to reveal the dangers of artificial intelligence, while reassuring the reader about these creatures: they should not frighten their users and are not dangerous to humans.

Her ethical research and numerous scientific publications have led her to become an essential voice in the debate on human-machine dialogue. She has notably participated in Allistène’s Review Commission on Ethics in Research on Digital Science and Technology (CERNA), the DATAIA Institute’s BAD Nudge-BAD Robot project, and the Global Initiative on Ethics of Autonomous and Intelligent Systems (IEEE). Her desire to provide education in the face of advances in emotional robotics will be on display in January 2020 at a seminar in Dagstuhl, Germany, with colleagues from Universities in Kyoto (Japan), Sheffield (United Kingdom) and Medford (United States).

  • 2004


    Laurence Devillers leads a team dedicated to the “emotional and social dimensions of spoken interactions” within the LIMSI/CNRS.

  • 2011


    She teaches artificial intelligence at the Sorbonne, while continuing her research at the CNRS.

  • 2015


    She becomes a member of CERNA, Allistène’s Review Commission on Ethics in Research on Digital Science and Technology.

  • 2017


    Her book Robots and Humans: Myths, Fantasies and Realities is published by Plon.

  • 2020


    Laurence Devillers will organise the seminar “On Spoken Language Interaction with Virtual Agents and Robots” in Dagstuhl, Germany.

The Institut français and the researcher

Laurence Devillers spoke in Japan in 2018, as part of a series of intellectual debates entitled: “Humans and Machines” These discussions are supported by the Institut français as part of the d'Alembert Fund.

Find out more about the d'Alembert Fund here.

L'institut français, LAB