Remi van Trijp

Sony CSL-Paris

Human languages have evolved many fascinating solutions to complex communicative problems through the use of words and grammatical structures. And they keep on evolving: language is an open system, a unique ability that brings infinite variety to the ways in which we communicate with others about our experiences in life. How is this possible? Can we understand this linguistic creativity? In my research, I try to answer these questions by developing powerful cognitive language technologies, which can be used to study open-ended and robust language processing, to explore innovative linguistic applications, and to function in large open collaborative communities.

From Prometheus Unbound to the Wild Wild West: A Primer on AI Ethics

Ever since its birth in the Summer of 1956 at the Dartmouth college (USA), Artificial Intelligence has raised important philosophical, legal and ethical questions. Unfortunately, the debates on these questions often end up in a shouting match between two opposite strawman positions. On the one hand, there are the Doomsday thinkers who fear that AI systems will soon – if not already – gather uncontrollable and god-like capabilities that will lead to the extinction of all of humanity. On the other, there are the ostriches who stick their heads in the ground so they can deny the reality (or at least the significance) of any dangers related to the deployment of AI technologies. Both positions obscure the fact that there are a whole range of ethical issues that are important to address, and they may cause you (a young, dynamic and talented researcher) to feel as if you are being unfairly tethered to ethical guidelines. The goal of this seminar is (a) to give you a primer of the very real issues in AI ethics and some practical steps that you can take in your own work, and (b) to start a discussion in the lab on how we can ensure to deliver “kando” and “épanouissement” to the world.