Much has been written about how federated learning (FL) can enable training much better models and help build more intelligent systems. This essay instead discusses a more personal aspect: Why working on FL as an individual is interesting, meaningful and fun. You can either take this as an advertisement for why one should work one FL, or as a personal reflection on why I am still passionate about it after working in the space for a few years. Fast and strong feedback loops One thing that init...
