FRIDAY 14 FEBRUARY > 10:30 am
Auditorium Grenoble INP
French/English simultaneous translation
How can we fight against the amplification of discrimination by artificial intelligence?
Why are artificial intelligences able to lead to discriminatory practices linked to gender or ethnic, social or geographic origin? The algorithms that reign over our AI are not neutral. They simply learn what we show them. They can be affected by numerous biases that have a discriminatory effect on individuals. These biases are introduced by the programmer or by the data that feeds the algorithm. Hence, when we assert that an AI makes decisions, it is the result of a chain of almost instantaneous decision-making events from the production of the algorithm through to its utilisation.
AI is spreading into ever more vast sectors of our lives. So how can we support its development in order to make it fairer and more egalitarian and prevent the different forms of discriminatory practice that could potentially emerge?
Patrick Loiseau, INRIA Researcher and Member of the IT Laboratory of Grenoble. Co-holder of the "Ethical and Responsible Artificial Intelligence” Chair at the Multidisciplinary Institute for Artificial Intelligence of Grenoble (MIAI)
Vytautas Jankauskas, Artist, Designer, Member of the Cached Collective (Cached)
Marie Lorphelin, INRIA légal counsel
Claire Richard, Author, Journalist, Translator, Chief Editor of the Digital Society Forum
Within the scope of the project entitled European Artificial Intelligence Lab project co-funded by the European Union’s Creative Europe programme and IDEX - UGA.