Eugene Kirpichov. Towards an ML-augmented Programming Stack

Making machine learning a first-class citizen in application programming is one of the "holy grails" of the domain. Packaging ML into abstractions practical for everyday development involves challenges at different layers of the programming and research stack, as well as developing new layers and new kinds of abstractions. We present a vision of how such a stack could look like, and recommend directions of necessary work to make it a reality.

slides (https://docs.google.com/presentation/d/1fDTYQ9MnZsJPPbDQkMi6vIh4RDwxl1q842jCUHDaS1M/edit?usp=sharing)

Speaker: Eugene Kirpichov
Software Engineer, Research and Machine Intelligence
Google

Join Zoom Meeting
https://scu.zoom.us/j/95229947801?pwd=MzdqODBwK1Y1UHkwQVdYSjV2NjFrUT09

Meeting ID: 952 2994 7801
Password: 583404

Join by phone:
+1 (669) 900-6833
Meeting ID: 952 2994 7801

One tap mobile
+16699006833,,95229947801#