Abstract:
The powerful Machine Learning algorithms that fuel innovations in production and services in all fields are advertised as able to generate so well tested solutions that no decision seems to be left to leaders and experts. Some human activities are already automated behaviours and will easily be replaced but decision making includes many other types of situations. Machine Learning algorithms design needs to be combined with an "organizational design" in order to make humans and machines learn to develop some collective intelligence.

Speaker's Bio:
Dominique Boullier is professor of sociology, member of the Digital Humanities Institute at EPFL (Ecole Polytechnique Fédérale de Lausanne, Switzerland). PhD in sociology (EHESS), in Infocom, and MA in linguistics. Expert in digital technologies, specifically social networks, politics of digital architectures in organizations, digital cities, digital methods in social sciences. He founded and was the director of a company in the 90’s and of many labs (including the medialab at Sciences Po together with Bruno Latour) and test beds in user experience (Lutin, Cité des Sciences, Paris). He published a manual Sociologie du numérique (Paris, Armand Colin, 2016) and many papers on a theory of replications and how to track propagations in order to build a third generation of social sciences.

FORUM NUMERICA is sponsored by the UCA JEDI Academy of Excellence “Networks, Information and Digital Society”.