This year, open-domain conversational AI systems have finally reached the general public: they are widely accessible, excel w.r.t. the naturalness and fluency of their generated output, and provide helpful responses for many of the users’ requests. However, there are still many open challenges — in particular, relating to ethical aspects of their design: for instance, conversational AI systems are prone to encode and amplify unfair stereotypes and are exclusive to many speakers identifying as members of underrepresented cultural and subcultural groups. In this talk, I will present some of these challenges and discuss potential solutions towards responsible AI-based future communication.
This event is part of a series of lectures and tutorials on data science topics hosted by the Platform for Challenges in Data Science in the excellence cluster "Quantum Universe" between DESY and Universität Hamburg. It is intended specifically for the PhD students in the cluster but younger and more senior members are of course also welcome.
Gregor Kasieczka, Matthias Schröder