I thoroughly enjoyed this talk and thought it would be nice to share.
The bit that starts around 7:45 where he talks about the “clash of worldviews” between computer scientists and domain scientists is particularly great; I’ve been trying to articulate that basic idea for a year and failing.
The point about languages for programming in the large being domain specific was interesting too.
Ignore the poor quality of the audio and sit through this talk (although you can barely hear anything in the Q&A session at the end).
You should come away with the realization that when you describe a system by saying, “I can’t believe this works”, the implicit assumption is that it does, in fact, work.