I wondered what careers or the like help other people the most. Tyler reposted my question, adding:

Let’s assume pure marginalist act utilitarianism, namely that you choose a career and get moral credit only for the net change caused by your selection. Furthermore, I’ll rule out “become a billionaire and give away all your money” or “cure cancer” by postulating that said person ends up at the 90th percentile of achievement in the specified field but no higher.

And answered:

What first comes to mind is “honest General Practitioner who has read Robin Hanson on medicine.” If other countries are fair game, let’s send that GP to Africa. No matter what the locale, you help some people live and good outcomes do not require remarkable expertise. There is a shortage of GPs in many locales, so you make specialists more productive as well. Public health and sanitation may save more lives than medicine, but the addition of a single public health worker may well have a smaller marginal impact, given the greater importance of upfront costs in that field.

I’m not convinced that the Hansonian educated GP would be much better than a materialism educated spiritual healer. Tyler’s commenters have a lot of suggestions for where big positives might be too – but all jobs have positives and many of them seem important, so how to compare? Unfortunately calculating the net costs and benefits of all the things one could do with oneself is notoriously impossible. So how about some heuristics for what types of jobs tend to be more socially beneficial?

Here some ideas, please extend and criticize:

Low displacement: if someone had to be hired, you only add the difference between your ability and the second best candidate (plus the second best candidate’s efforts to another job at random). The same goes for what you produce. Even if creating beautiful music doesn’t knock another musician out of business, people listen to your new song instead of older songs, which are not seemingly any worse.
Big gains to a marginal person being better: careers that fail the above can still rate highly if this is so. This is a hard route because if candidate quality matters more there will generally be stronger selection so you will be more average for who they let in. Your best bet here might be in something important to you which usually attracts people with poor abilities.
Goal discretion: one way a displacing person can make a relatively big difference is to have a better purpose than the person they displace in a position where that matters. This is more likely in jobs that aren’t meant to be altruistic. e.g. if you head an NGO for delivering fresh water to poor households, you only add to the cause to the extent that you are better at it than the alternative guy. If you are an academic with some choice in research topics and no imperative to be altruistic, you can add a person’s worth of effort to an important topic at the expense of an unimportant topic without having to be any better at researching. Similar for shop owners and heads of state.
Profits: if you are making huge profits, it’s probably good.
Positive externalities: same.
Ongoing effects: usually a kind of positive externality. Nice cake lasts minutes whereas nice ideas last a long time (though someone else would have thought of them eventually).
Negative externalities
Fueling status races or other zero sum games: arguably things people buy to look better than other people do not cause any overall benefit, so you shouldn’t read too much into the profitability of selling such things. Similar for weapons in at least some situations. Same for selling any other negative externalizing capabilities.
Speeding trends: culturesome things tend to cycle through new variants continually. Contributing a song, movie, or meme variant for a current trend certainly makes a lot of people like you (briefly). But it’s not clear what people value here: each new item while it is new, or being up to date with the current trends in general. If the former you are doing them a service by adding your bit of culture, if the latter you are just making it more expensive for them to keep up.
Effects on galaxy colonization chances and speed: this matters so much you can ignore all the other things.

What jobs do well on most of these things?