Answer by I-Kang Ding:
The advise I have for my younger self is not specifically related to data science, but more related to career development in general.
- Be prepared to re-learn new knowledge and re-invent yourself during your entire career. The age of stable and permanent employment is over.
Example: I chose chemistry over CS in my college major, thinking that "CS technologies change every few years, I'll have to re-learn something new! But with chemistry, what I learned will always be useful because the field is not updated as quickly!" Silly me.
- Don't fall to. It is not helpful to dwell on how much effort you have spent on a particular project. You can not recoup that effort. Instead, consider the opportunity cost, and ask yourself the question, "given what I know now, what is the best course of action"? Sometimes the best course of action is to change projects, even if you may be attached to the previous project.
Example: I worked on a project longer than I should have in grad school because of Sunk Cost Fallacy. I am still trying to keep this in mind when I evaluate projects.
- You may be overestimating the risk in your head. I heard an interesting theory years ago (which I can't find the source now) that we as humans have an over-sensitive aversion to risk, from our animal days. Back then, the risk associated with a bad decision could lead to starvation, or being hunted and eaten by predators… etc.
But right now, I'd argue that we (Quora readers) mostly like have our "food and shelter" needs satisfied, so the decisions that we need to make (should I join company A or company B? should I choose major A or major B?) are not that "risky" at all. The risk may only exist in our head, so don't be afraid of taking a change!
Example: it took me a long time to decide to make a career transition as a materials scientist (with industry jobs and all) to become a data scientist. But once I thought through the real risk in the grand scheme of things, I realize that the risk wasn't material (pun intended) at all.
Hope that helps!