1
I think it depends. There are certainly those who are most interested in ML itself and would rather not „waste“ time on improving the setup. Lately I think that it is often hard to justify investment in tools and abstraction when there is pressure to deliver value.