Even if you can individually help the most people by becoming a doctor, you can probably do better by paying someone with a comparative advantage at doctoring to become a doctor or do more doctoring (while you focus on what you're best at).
This may be good advice when the socially-beneficial job in question is something less expensive than medicine. But doctors are expensive. Most of us can't afford to pay for someone else to be a doctor.
But, still, let's consider someone whose skills and interests do make it feasible for them to do so. Let's say they could work as a doctor (earning, let's say, $150k/year, and perhaps costing $200k/year "fully loaded" -- for the avoidance of doubt, all numbers here are completely made up) or as some kind of financial analyst (earning, let's say, $3...
A high school student with an interest in math asks whether he's obligated on utilitarian grounds to become a doctor.
The commenters pretty much say that he isn't, but now I'm wondering-- if you go into reasonably pure math, what areas or specific problems would be most likely to contribute the most towards saving lives?