Institute of Formal and Applied Linguistics

***We regret to announce that the lecture has been cancelled. We hope to reschedule this event to a later date as soon as possible.***

I sketch a number of bottlenecks for progress in fair and explainable NLP. Some of these result from foundational problems, some from lack of commonly accepted experimental protocols, and others from apparent impossibilities. I will present examples of recent and ongoing research that attempts to address these challenges.