Recent posts by Larry Wasserman and Vincent Granville are fantastic additions to the data science conversation.
Here's the short version (with links to original posts):
There is a lot of merit in their advice and frankly, if I didn't live and breathe it, I wouldn't promote it.
Two things to add, Irving Wladawsky-Berger and Greta Roberts have equally compelling arguments for data scientists: be creative and curious. In the business world, this is not only extremely important if you want anyone to listen and pay attention to you but are drivers of innovation.
What does this mean practically? For us statisticians, it means there are a lot of signals out there about the 'how' and 'what' of being a data scientist but acquiring that expertise isn't going to be straightforward or easy (that's my experience). That being the case, I would also add: always be prepared to pivot in your professional endeavors and be a strong practitioner of continuous learning.
On a related note, if you have the time, I would also recommend this video posted by Joseph Rickert at Revoluation Analytics.