The focus of effective altruism has been shifting from giving to doing. When EA started out it was mostly about encouraging people to give more and give effectively. The initial organizations, GiveWell and Giving What We Can, were focused there, and aside from "effective altruism" the other contenders for the name of the movement I saw were "effective giving," "smart giving," "optimal philanthropy," "smart aid," "high impact philanthropy," and "efficient charity". People who were thinking about the movement did recognize that doing good through your career should be included [1][2][3], but even in March 2013 Peter Singer's TED talk on effective altruism was about giving, the obligation to give, making giving more effective, and earning more to give more.

Over time, however, doing good via your career has been becoming a bigger part of effective altruism, with yesterday's why you should focus more on talent gaps, not funding gaps being closer to how people are thinking about things now.

I think this is very positive. We've been doing well enough at earning to give and influencing foundations that it's getting harder to find giving opportunities that are primarily bottlenecked on funding. We could create more excellent giving opportunities, though, by founding charities that GiveWell would like to recommend. Growing the EA movement, improving government policy, or researching in a high priority field are also really valuable.

There would be some really nice things about building a movement just around donating:

• Charity research is general: anyone can read GiveWell's recommendations and make giving decisions based on excellent evidence; the best career for you is going to be a complicated interaction between your preferences, abilities, and interests.

• Charity research is more mature: in choosing what charities to recommend, GiveWell can draw on decades of studies and theory by economists and other researchers looking at various ways of helping people; evidence-based career advice that considers counterfactual benefit and really tries to find the best option all-things-considered is just getting started.

• Donations are legible: if someone donates 10% of their income, that's good. If it's 5%, that's less good, 15% is more good. Numbers can be shared. This lets us help keep each other on track. With career choice it's a lot harder to tell if someone is making good decisions. What if "the computer science majors have convinced each other that the best way to save the world is to do computer science research"? [4]

• Donating is more adjustable: many people can fit donating 10% into their existing lives, and the percentage can be scaled up or down based on how altruistic you are. Taking a career for altruistic reasons, potentially in another city, can be a lot more demanding. Though much less when you're starting your career.

• Donating is a shared experience: a movement of people who donate means everyone has something in common. They can talk about giving opportunities, ideas for earning more, and ways to spend less to have more to give; people going into widespread fields for EA reasons have less in common with each other after just a few years.

This isn't enough to justify pushing donations beyond what we need, though. We can't respond to more funding being available by continuing to recommend earning to give and donating, not if we want to have as large a positive impact as possible.

(This isn't saying "don't donate," but instead "if you're trying to decide between donating and doing direct work, lean more towards the latter than we had been.")

[1] When founding CEA the Oxford group wanted a name that explicitly was not limited to doing good via donations, and chose "effective altruism".

[2] When Julia asked for advice on career choice in August 2011, thinking about earning more so she could give more, Carl Schulman mentioned he was doing research for High Impact Careers, which would become 80,000 Hours. Similarly Jonah Sinick suggested Julia might do a lot of good through influencing, networking, or working at a foundation.

[3] Earlier that August I'd written about how someone could make being a Giving Advocate into a full time job. At the time I thought nearly everyone should be donating and just a few people should be in influencing and fundraising positions.

[4] I don't think this has happened. The sort of skills MIRI wants in researchers are a lot more mathematical than the people they most appeal to have. But this is an example of the sort of conflict that "A earns the money, chooses B to spend the money doing X" gives you less of than "A does X because they think it's the best thing to do".