This talk, given by Sciences Po sociology professor Dominique Boullier, is organized by the Maison des Sciences de l’Homme SUD (MSH SUD) research platform Numerev, and will be made available via a live webinar on MSH SUD TV.
Boullier, who is also a former researcher and lecturer in the Digital Humanities Institute in EPFL’s College of Humanities and former director of EPFL’s Social Media Lab, will speak about the risks and opportunities associated with the use of machine learning results. According to the talk abstract:
“If machine learning users keep moving in the direction of extracting artificial intelligence from organizational design and expert contributions, they run the risk of overestimating the value of their predictions. At the same time, they will gain a real advantage in terms of accurate and in-depth knowledge of social behaviors, which will put their organizations in a position to bypass all institutions and laws. Another possible route is to develop “human-machine learning” that will empower organizations and citizens, not only when using machine learning results, but also at the very heart of their design.”