Thanks for setting down in words the misgivings I’ve carried around for some time around the titles, functions, roles, skills and expectations. More importantly, you’ve separated the classification problem from the labelling problem. There is a tendency to dismiss the labelling problem as “soft” or tackle it as an afterthought to “hard” classification, but that misses the two-fold importance of labelling: first, how you want to label entities surely influences how you want to classify them, and second, labels evoke a set of associations in the users’ minds and are extremely useful natural language short forms for understanding and acting on the classification.
There has been a lot of title inflation going on in the industry, but what Lyft is doing is obviously much deeper: recognizing the two classes of roles/functions/skills and giving them evocative labels in keeping with changing industry expectations.
A many to many map between the lists of functions and titles would also be useful.
By the way I had never heard “hammer carpenter” before. A better (equally snide) analogy with “data scientist” would be “wood carpenter”, since data is the medium of every scientist like wood is the medium of every carpenter. Every scientist works with data, the question is what kind of data and what information or knowledge they are trying to build from it; the set of tools is then somewhat secondary.