One annoying thing about being a generalist is that domain experts in any given area that you need familiarity with can't help but complain about how little you

One annoying thing about being a generalist is that domain experts in any given area that you need familiarity with can't help but complain about how little you know about that domain, ignoring the fact that your job requires equally deep knowledge of several other domains simultaneously. In the case of data scientists, I think the business folks that want them to understand the business domain better generally have the strongest argument, followed by the statisticians - good data scientists need to personally understand both of those things well, while the engineering and ops stuff that data scientists are also expected to do is easier to compartmentalize on other teams. So I agree that we should have more data engineers, but apparently for the opposite reason as most people in this thread.