On the Haste Consideration

On Facebook's Effective Altruism page, Konstantin Sietzy asked what effective altruists now thought about the importance of recruitment. This post include some of my off-the-cuff thoughts on the matter.

"I was wondering about the relative effectiveness of taking a typical revenue-maximising career in finance, or petroleum engineering as mentioned below, or what have you, and on the other hand devoting one's time to promoting EA?...

If instead of taking a career in finance and signing the giving pledge, for example, I convinced two other individuals - equally capable/equal expected avg. earnings over lifetime - to do the same, I would have already done twice the amount of good...the number of people you can convince quickly hits an upper bound n. Given that I would already be doing 'more good' over all by pursuing a less lucrative less time-incentive job and convincing merely n+2 others, shouldn't this be an increasingly favourable alternative on a continuum up until dedicating all of one's time to promoting EA?...

William MacAskill himself is probably the best example of this, given that he didn't develop the idea of EA and then draw the consequence to immediately pursue a finance career and never speak to anyone about it again; and the enormously positive effect that has had certainly bolsters the above point.

This question alludes to the The Haste Consideration, a piece by Matt Wage. In this piece, Matt argued that convincing one other person to be an effective altruist can be equal to a lifetime's worth. He said that this depends on the condition that they are able to do as much good as you. However, this is not so easily done, as it requires persuading someone to do some combination of:

  1. taking on your values
  2. being as competent as you at executing on these values
  3. being prepared to stick with this as long as you.

If someone is only able to take on part of your values, but they are more competent than you, then this could suffice. Nonetheless, this can be quite a tricky set of conditions to meet. Furthermore, note that you have to evaluate the impact of your act of persuasion counterfactually. Would they have discovered the idea of charitable effectiveness a year later anyway? Were they already giving to moderately effective causes? And so on.

It's kind of like the problem of finding an ideal successor to a dynasty, or a company, or a research laboratory. Finding the right person (or people) through whom you can live on can be an extremely tricky multi-year project.

If you are, at heart, a researcher and you possess a bunch of scholarly virtues, then it's going to be hard to find other people like this, especially if you're not willing to do any scholarship yourself. On this ground, Will MacAskill's decision to complete a PhD, so that he can go on to perform further philosophy, can be supported. If you go on to meet a bunch of experts in relevant areas of research, then winning them over is more valuable. And indeed, especially if we are concerned about the potential benefits and risks of emerging technologies like in vitro meat, synthetic biology or artificial intelligence, then on a per-person basis, influencing researchers might be much more important than influencing random members of the general population.

It's similar for earning. If you can only maintain a relationship with a handful of altruistic donors, then it's better that they work for Jane Street, that they are venture capitalists, or that they are entrepreneurs with a good chance of success, rather than random university students. (I say university students because this is often where local effective altruist chapters are often established. Of course, for top tier universities or business school, the chances of wealth or research success in later life are higher, clouding the decision)

This would suggest that for many people who are talented in research or who have high earning potential, a blended strategy of recruiting others while executing one's talents is likely to be better than pure recruitment. The mixed strategy helps with discovering useful allies, having shared experience and having a good reputation.

This idea is especially compelling when you notice that lots of would-be influencers are at the star tof their careers. Most people have the most wealth, professional ability and political clout when they are 40-70 years old, rather than 20-25. Lots of aspects of life in an early career resemble investments. You might invest in your education, your technical talent, your status within an organisation, or your relationships with mentors. In any of these cases, your actions early in a career might plausibly compound and allow you to be able to do good much more potently at a later date. This would suggest, at least at first glance, that if you want to perform promotional and recruitment activities for effective altruism, it might be good to gain the requisite skills and contacts in order to do an awesome job at this first. Performing a mediocre outreach effort and attracting a handful of students can be fun. It can be useful if those students are on a good growth trajectory, and are likely to be poised to make a big difference later on. But in many cases, it will be less useful than spending months to years developing your own professional network and then reaching out to a group of individuals who are likely to be of particular relevance to your cause e.g. directors of charitable foundations or AI research labs. Conversely, if one is already, for example, an eminent AI scientist, then recruitment of one's colleagues and collaborators may be extrmely impactful.

Of course, this is not a hard and fast rule. If you are a star fundraiser, then you may be able to earn more by plainly and simply giving your fundraising talents to important organisations than you could ever raise by performing advertising roles, even in major companies. Widespread effective altruist marketing may be able to uncover hundreds of people who will earn half as much as you would have, or made a quarter as many research breakthroughs.

In fact, in the case of research, performance is so skewed, that even if you are very smart, the most important thing may be to discover the handful of people who are even smarter than you. CFAR's SPARC initiative looks excellent under this lens. Moreover, on this view, professional networking with scientists looks better than meeting students at a local university.

Of course, this doesn't exhaust the major considerations. A lot more work needs to be done to evaluate how this trade-off plays out for different individuals at different career stages but we at least have a few reasons to think that purely evangelising one's moral system might be a suboptimal strategy.