High Impact Medicine, 6 months later - Update & Key Lessons
With thanks to Jacob Arbeid, George Rosenfeld, Jacob Smith, Thomas Cohen, Jake Mendel (Hi-Med Oxford), Pauline Scheuber (Hi-Med Germany), Erik Jentzen (Hi-Med Germany) and Rahul Shah (Hi-Med Cambridge) for comments.
6 months ago, we launched High Impact Medicine. In this post, we share an update on our progress and the lessons we learnt in running a community building organisation focussed on a particular profession. We hope that this might be relevant to others interested in community building, particularly in professional and workplace groups.
High Impact Medicine is a group whose goals are to bring together medics who want to have a wide reaching positive impact, alongside and learning from clinical practice, and explore the opportunities for them to do so. We have 3 main paths to impact:
Outreach & community building: our chapters run talks and socials for medics of all grades, from students to doctors working in hospitals or outside of clinical medicine
Knowledge building around impact in medicine: we run a fellowship, a podcast and talks. We are planning to run follow up ‘senior’ fellowships as well.
Career advice: we help members explore how they can have a large impact alongside, and learning from clinical medicine
We set out to foster local, engaged communities of medical students and doctors including those not familiar with EA, focussing on outreach to new interested medics, working alongside EA Medicine who facilitate a global network of medics who are already EA aligned.
Since launching:
High Impact Medicine has grown to approximately 100-120 members, at least 60 of whom are moderately to highly engaged (definition).[1] Our rough split has been 70% medical students and 30% doctors.
Our inaugural fellowship attracted 7 cohorts with a total of 54 fellows
We have run many talks and socials in Cambridge, Oxford and London with an average of 15 attendees per event, with speakers including Professor Sir Andrew Pollard
We have launched our Podcast
Several (≥4) members have taken the GWWC pledge, and ≥3 others have made changes in their career direction as a result of being part of the organisation.
We have expanded significantly beyond our initial base in the UK:
We recently launched High-Impact Medicine Germany
We are running a cohort of our fellowship for medical students in Tehran, Iran
We have members in the USA, Australia and Singapore.
Our key lessons
1. Having an open framing allows for effective outreach
We chose to frame Hi-Med openly—a chance for medics to find out about ways to ‘have a widespread positive impact’ and explore areas such as ‘global health, health policy, biosecurity and mental health’. Our objective was to try and get a lot of people interested in considering their own personal impact and how they could do a lot of good alongside and learning from their medical backgrounds. This was reflected in the fact that a significant number (approximately 25 out of 50) of the fellowship participants were entirely new to EA. Many of the ideas we discussed in the fellowship stem from the Effective Altruism community (e.g. ITN framework, EA-aligned cause areas) but we also drew on material from the wider scientific community [2]in our fellowship, talks and podcast.
Our approach was intended to provide medics with information to make their own decisions and enable an honest approach. Hearing ideas without them being exclusively attached to a wider philosophy and in more neutral (rather than persuasive) fashion may help avoid putting people off and allows individuals the freedom to pick aspects they found more meaningful to explore. This helps encourage a ‘truth seeking’ approach that focuses on ‘learning useful tools and asking important questions [… ] and helping [individuals] have the biggest impact they can’ rather than the impression that EA was a large, take-it-or-leave-it all-encompassing philosophy you must be persuaded of. The objective was not to subtly trick people into becoming more EA but to provide information about ideas that we thought were important, worthy of discussion and that people may be receptive to. We found that some of our members did in fact become more involved in EA more broadly, whilst several individuals took on board ideas that are common in EA but felt ambivalently or negatively towards EA as a wider movement.
With an open, non-prescriptive framing, we found that many medical students and doctors (in fact, more than we had anticipated) are interested in maximising their impact. The success of this approach may be due to optics issues with EA as a wider movement as well as stem from the fact that EA’s current recommendations to those from medical backgrounds are suboptimal. Therefore, by not being solely ‘EA’ we can provide a wider set of recommendations, framed positively, as well as a framework for individuals to assess their own priorities—which we believe is both attractive and effective.
2. Fellowships work very well and there should be more career focussed fellowships
There is debate over the value of fellowships in EA - we found that a career focussed fellowship worked very well to introduce many medics to ideas such as the impact of different cause areas, how to do a lot of good as part of your career and effective giving.
Most of our impact so far has been from our fellowship—an 8 week reading group exploring high impact opportunities for medics to pursue alongside and learning from clinical practice—which individuals from first year medical students to junior doctors to senior doctors took part in. This took about ~3 months to write (approximately 250-300 hours of time) but was less challenging than we thought and it has really accelerated our growth.
We intend to write up the feedback from our fellowship with all our data separately, but in general:
We had reasonably strong attendance ( > 80% at all sessions, with 3 out of 57 participants dropping out through the fellowship), and strong engagement
Level of familiarity with effective altruism significantly increased in the group before and after the fellowship
Of the 42 fellows who completed our end of fellowship survey, 90% (n = 38) said the fellowship had a moderate to major effect on their knowledge of high impact opportunities related to medicine, 83% (n= 34) said it had a major to moderate effect on the way they think about their career, 64% (n=27) said it had a moderate to major effect on how their donation habits.
Specifically, a number of individuals are actively pursuing or have changed their current work to be more impactful, or have taken a giving pledge.
We believe it would be valuable if there were more similar fellowships for careers and/or paths to impact from other (vocational) subjects like law, engineering, computer science, journalism as well as other subjects which do not necessarily require university degrees. The impact landscape for professionals can be somewhat specific to vocation with the dynamics and nuances of considering your personal impact, giving and future plans specific to career. For example in medicine, people are already likely to believe that career choice is an ethical issue and they understand the use for potentially uncomfortable impact calculations given the prevalence of unpleasant cost benefit decisions in healthcare - different careers are likely to have features which lead to certain discussions being easier and others more challenging.
We are writing up the lessons from our fellowship in a subsequent article, including our novel design of fellowship reading documents with insights from readings integrated and summarised, as well as the feedback we collected from fellows.
3. Funding speeds things up but you can also do a lot on a little (and avoid most of the ‘free spending’ issues)
In the wake of the recent debates on the EA forum about issues of free-spending in EA, we felt the need to share our 2 cents (pun intended)
We found that we didn’t need much money to have a large impact in community building and therefore could avoid issues with free spending EA organisations others have raised. We received a grant from EAIF (approx $10,000) and this helped give the project significant impetus for which we were very grateful and for which we’d 100% encourage others to apply in similar circumstances. Money helped drive certain projects—such as website design or podcast creation—but in total these expenses were <£200 over our first 6 months. More broadly, people were generous with time—most of the fellowship was created without paying experts for their time, and whilst we offered some financial incentives to fellows (in the form of meal vouchers) we do not think that lack of these financial incentives would have resulted in fewer applicants. Although we are only somewhat confident of this claim, we think it is likely that any further incentives may have been actively negative, as showering food and money on privileged fellows is highly incongruous with discussing some of the world’s biggest issues.
An additional important consideration is that the optics of free spending is important across EA but may represent an even greater risk in certain areas. Many medics feel uncomfortable spending large amounts of money[3], in part stemming from lived experiences of healthcare systems struggling due to chronic lack of funding - so an organisation for medics such as Hi-Med must consider its expenditure seriously, considering both the optics and inherent issues that may exist when spending publicly on e.g. salaries, retreats for medical students.
In general some funding is definitely valuable and having small amounts of funding (a few £100s) can provide significant marginal value in the initial stages of new project start-ups compared to having none. However, diminishing returns may start to occur far sooner than one may think and it is important to think carefully about where there is value in spending money—especially given the potential issues which may arise as a result.
4. Don’t forget about Giving Pledges
With the vastly changed funding landscape in EA, it is important we don’t neglect personal giving. The Giving What We Can Pledge is highly impactful and with the huge influx of money, as well as the shift in many peoples’ values towards longermism and career changes more generally, it should receive more attention than it currently does. We think it should be foregrounded and encouraged for a few reasons.
Firstly a single giving pledge can result in £10,000s going to highly effective charities, which categorically does good in the world and is valuable in isolation. Secondly, specific to professional groups, it allows professionals to show dedication to EA values who are not able to make career changes in the near term—this is valuable as it allows the EA community to be more inclusive and benefits general optics on what EA ‘is’. Thirdly, and relatedly, a giving pledge may be an effective way of preventing value drift. This is especially important in medicine where people are likely to spend many years in studying/ further training before making significant choices based on impact considerations. Therefore encouraging pledges from individuals keen on doing good early on in their studies is particularly important. For pledging to be able to achieve this, it is important that giving pledges (and those taking them) are appropriately valued—so individuals feel excited about pledging rather than inadequate about their commitment to having a large impact.
We found pledges were the most immediately accessible way for fellows to actualise the content of the previous 8 weeks (in the sense that it could be taken at the end of the fellowship without significant extra commitment, as opposed to applying for a job, which may take several weeks to months) - and even when pledges were very much a secondary focus compared to career changes (being discussed only in week 7), several fellows subsequently pledged. Let’s not lose sight of the value of emphasising and encouraging pledging.
5. Move fast and do stuff
We were initially hesitant to move ahead with projects quickly. For instance, we created public-facing chapters in Oxford, Cambridge and London immediately after launching, and began our fellowship just three months afterwards. Initially, we worried that this wasn’t enough time to have the level of organisation required for a public-facing organisation (and therefore an organisation with higher reputational risk and where making a positive first impression was more important than a more inwardly focussed org). And in the last few months, we have launched in Germany, are running our fellowship for 11 EAs in Iran, and are planning a senior fellowship and further career planning workshops.
Moving fast has helped the Hi-Med community grow, bring in a large number of new members quickly, and maintain the organisation’s momentum. This does increase the propensity for small mistakes or oversights, but learning from mistakes is more time effective than trying to predict and mitigate those same issues. Our key learning is that it is often helpful not to overthink and over-strategize, and to worry less about every google doc, email or publicity post being ‘perfect’. Furthermore, the fact that our approach primarily focussed on providing information to facilitate individuals coming to their own decisions reduced the need to carefully tailor every bit of messaging. As long as the essential framing and top-line messaging is clear, unambiguous and positive we could move quickly and achieve more in a relatively short time frame.
It is important to note that this learning may not be widely applicable—in situations where reputational risks are higher, there are information hazards, or where sensitive messaging is key, slow and steady may indeed win the race. However, there is a huge learning value in trying things, even if they fail—and in incidences where EV is particularly high, and there is a low to moderate downside and optics risk, this strategy is applicable.
6. Medics are a good fit for EA ideas (but are also likely to be put off by some EA messaging)
Medical students and doctors are in some ways a natural fit for EA ideas—as a group of altruistic individuals, who applied for a long degree and training pathway to ‘help people’ and come from a strong scientific background where evidence-based decision making is common and individuals are used to thinking deeply about ethics.[4]
We found that this was the case during our fellowship, as there was less debate than we expected on the key moral and epistemic premises that ground EA. We also found that a number of EA cause areas, in particular global health and development, biosecurity, and AI and technology, resonated strongly with fellows.
However, any medics looking to find out about EA are likely to be put off pretty easily for a few reasons. Perhaps primarily is that initial readings are often based on the idea that being a doctor—a career they have likely invested significant time and effort in - is not particularly valuable owing to the counterfactual and they are advised to consider changing careers if possible.[5] Other factors include
That fact that many medics are motivated by emotional reasons whilst EA narratives, although grounded in compassion, may seem too ‘cold’ in comparison
Medics are more motivated to help people in their immediate surroundings with their hands and find generalising to those far away challenging
Medicine is fairly neartermist (helping patients now), meaning longtermism is far less intuitive
These factors that may put a medic off often do not reflect fundamental differences in values but result from intuitive aversion from the framing of these unfamiliar ideas.
As a result, when medics come across EA for the first time they may be unlikely to want to engage deeply for a combination of reasons. We have heard this colloquially many times whilst community building, leading to a neglect of a large group of potentially aligned and talented people. Providing a space to provide information about and discuss ideas related to impact and medicine openly and positively is therefore potentially important and valuable.
7. Building and fostering a strong management team and community is important
The more time people in the management team (and in the community more widely) are able to spend getting to know one another the better. Through social events, 1:1s and casual small group conversations you can quickly and easily foster effective management teams and local communities. It helps those in management support one another and know each other’s strengths, motivations and needs. Furthermore, we found having 1:1 calls with individuals who wanted to engage in the community highly productive in the early stages of community building—as it enabled us to support new members proximally and better encourage them to take part in subsequent activities of particular interest to them.
Conclusion
As EA continues to evolve as a movement and a philosophy, it will inevitably engage a more diverse base of people who may have increasingly different cause prioritisations, beliefs and priors. Our experience is that there is an increasing role for organisations that specifically support subgroups; carrying out outreach, providing tailored advice and building welcoming communities. These will form one part of big tent Effective Altruism and provide an important platform to provide ideas and opportunities in a positive way, whilst remaining committed to rigorous thought, worldview diversification and inclusivity.
- ^
We define moderate to highly engaged as at least one of
Participated in fellowship
Contribute to Hi-Med on an organisational level
Attended multiple events
Are involved with other Hi-Med projects
- ^
e.g. Hosting speakers pursuing what we consider impactful work although they are not aware of EA; considering and openly discussing neo-colonial undertones of global health work; focussing on health policy as a cause area beyond biosecurity; thinking about how clinical practice could be more impactful in and of itself, among others.
- ^
We think that this perception is likely to vary somewhat depending on which country the medical professionals are working in. In our experience, especially in the UK, where doctor salaries and funding for the NHS more broadly are a pain point, this seems especially true. This has also held true in other countries where Hi-Med has operated where there are relatively higher salaries for doctors, including in Australia and Germany.
- ^
In the UK’s NHS, there is a strong tradition of cost-benefit analysis which is likely to support utilitarian thinking. However it may more broadly be the case that medics are more deontological than average with strict rules on avoiding harm prevalent within the field
- ^
- Who’s hiring? (May-September 2022) [closed] by (27 May 2022 9:49 UTC; 117 points)
- Monthly Overload of EA—July 2022 by (1 Jul 2022 16:22 UTC; 55 points)
- Apply for the EA Workplace/Professional Group Organizer’s Retreat (Bay Area, 26–28 July) by (8 Jun 2022 9:03 UTC; 25 points)
- Is Medical School High Impact? by (25 Jul 2022 20:21 UTC; 13 points)
I worry a lot of these efforts are strategically misguided. I don’t think noting ‘EA should be a question’, ‘it’s better to be inclusive’ ‘positive approach and framing’ (etc.) are adequate justifications for soft-peddling uncomfortable facts which nonetheless are important for your audience to make wise career decisions. The most important two here are:
‘High-impact medicine’ is, to a first approximation, about as much of a misnomer as ‘High-impact architecture’ (or ‘high-impact [profession X]’). Barring rare edge cases, the opportunities to have the greatest impact are tangential to (or wholly apart from) modal jobs in this profession.
In the typical case, regardless of cause area, Alice who seeks to ‘maximize her impact as a doctor’ will have ~2+ orders of magnitude less impact than Bob who seeks to maximize his impact simpliciter, and is open to leaving clinical practice or medicine as a whole to do it.
To elaborate:
I am sure many medics are put off or dispirited by my writing on medical careers. In precis:
‘Face to face’ clinical practice (i.e. what you get trained for in medical school) is not an effective means of doing good.
Scales badly (i.e. essentially can only treat the patients in front of you in sequence)
Unlikely to be a leading contributor to aggregate population health (especially in low income countries).
Marginal responses in health from medical personnel in high income countries are at best ‘mediocre’ and at worst imperceptible.
Already greatly oversubscribed with appointable candidates.
From this low baseline, the small-n multiples you could get about being smart in how you deploy your clinical practice seldom result in impressive bottom-line impact.
You can earn-to-give by picking particularly lucrative locations/fields for practice. Your earning power is still substantially lower than other typical E2G career options (e.g. banking, entrepreneurship, pharma) plausibly accessible to many medics.
You can practice in lower-income areas, where the expected ‘marginal benefit’ of additional doctors is much greater. Yet even if you extrapolate the marginal analysis I did to these (which surely overestimates impact) you get yearly impact inferior to even unoptimized medical E2G in a country like the UK or US.
Almost by definition, “effective scalable interventions to improve health in resource-constrained contexts” will not require practicioners having several years of post-graduate training and experience as pre-requisites.
Folks with medical training can do a lot of good, but in these roles they either use their medical training ‘not at all’ or ‘very little’. The story is typically not “medical practice has taught these people important skills they now leverage in this new role” but “medicine often selects for able, conscientious, and altruistic people, who can do a lot of good if they turn their hand to something else”.
Although a medical background may offer some capital for many of these roles, it is typically an inefficient use of time versus more direct strategies to get involved (even for something more medically adjecent like research, an MD/PhD +/- junior doctor years versus a PhD alone add ~5 years which could have been spent doing the research instead).
Again on priors, it would be highly surprising if ‘managing hospitals’, ‘conducting scientific research’, ‘working on health policy’ (to say nothing of further afield options) is really greatly complemented by several years of training and practice in a job with little in common with it.
Thus the ‘impact case’ for medicine is typically tightly circumscribed:
Niche cases where medical background (or practice) really is crucial for a particular career path (e.g. roles in translational research/clinical trials)
When more ‘standard’ medical options really are the best one can do, given circumstances (e.g. maybe your highest earning activity really is clinical practice)*
Your shortlist of options for your career all overlap in such a way where some medical qualification/experience is a good buy across the portfolio*. (Or when you plan or pursuing a mixed strategy longer-term)**
If you’re part way towards some medical milestone (e.g. qualification, licensure), although this wouldn’t have been a good buy from Year 0, it is more reasonable now n years have already been sunk, leaving a smaller amount remaining.
(I astericks or double asterick the options I’m typically highly sceptical of as really being good strategies—I often suspect these are rearguard actions fought to preserve an already-cherished career plan or self-image)
Yet I am unrepentant (and instead zealous) in saying these things: even if seldom kind, they are always true, and often necessary. HIM says and implies at various points it disagrees with this assessment (e.g. “We believe medical students and doctors have the potential to effectively tackle some of the world’s leading problems alongside, and learning from, clinical practice.”). Yet whilst my reasons are transparent, yours are opaque. I doubt they would be credible if made clear, but you are warmly invited to prove me wrong.
But perhaps I misunderstand, and the main disagreements are more matters of style and messaging than substance. Although the answer to high impact medicine is ‘the highest impact options usually amount to leaving the profession sooner rather than later’, most people don’t (or can’t) always take the highest impact option (cf. my own mediocre charitable donation record). It is senseless to antagonise these people, and much better to provide intermediate options: perhaps with time they come around, and regardless their contributions (e.g. in donations to effective charities) are worth celebrating in their own right.
Yet there is a fine line between tact and insincerity; candour is often valuable even (or especially) when it makes the recipient uncomfortable. More importantly, it is unclear whether ‘foot in the door’ or ‘door in the face’ messaging is superior: maybe you have to pick which sorts of people you are least unhappy ‘putting off’. Most importantly, not saying the most important thing risks people who could have done more good doing a lot less good. I think someone like Lucia Coulter (or Jan Brauner, or Ryan Carey, or Bridget Williams, or Brenton Mayer, or Peter MacIntryre, or Cassidy Nelson, or Lee Sharkey, or Jonas Vollmer, or Sashika Coxhead, or...) is plainly having many times more impact than their counterpart who stayed principally within in clinical practice modulo a giving pledge or similar. Thus if HIM messaging leads someone to mistake these options as equivalently good, this plausibly outweighs its positive impact so far.
Hi Greg,
Thank you for your comment.
Big picture, I wanted to clarify two specific points where you have misunderstood the aims of the organisation (we take full responsibility for these issues however as if you have got this impression it is possible others have too).
1. We do not necessarily encourage people to apply for and study medicine. We are not giving any advice to high school level students about degree choices and paths to impact. To quote what you wrote, “medicine often selects for able, conscientious, and altruistic people, who can do a lot of good if they turn their hand to something else.” We think this is likely true, which is why we think specifically performing outreach to, and helping people who are already studying and practicing medicine find highly impactful careers, is likely to be an appealing option.
2. We foreground transparency in all our communications with our members—and am sorry this is not something you agree with when you write that our reasons are opaque, but I think this is a strong claim to make . We aim to talk honestly about the impact of clinical medicine; discussing the articles that you have written throughout the fellowship as well as other pieces on the issue e.g. Dr Launer’s article referencing yours in the BMJ. We do not have a specific ideology we want to sell anyone on—we want to present people with important information and allow them to form their own conclusions through discussions. Disagreement with EA or existing articles is not something to be feared, but to seek out in order to improve our understanding of impact. We think this is reasonably important for community building
Specifically, I think our difference in opinion is the way we approach the idea of impact in medicine—this involves some specific disagreements (below) but also general framing and outlook.
On some of the specific points:
1. High Impact Medicine not a misnomer- I disagree with your assessment of the situation; yourself, and all the people who you mention in the article, were or are doctors, and fit under what we would define to be High Impact Medics. When you are conducting outreach, I think it makes sense to describe the profession that someone currently has, even if you were to recommend jobs ‘tangential to or wholly apart from’ jobs in that profession.
2. On the career capital and skills of medical degrees for opportunities outside medicine- I think I have addressed the above somewhat in reframing the purpose of our organisation away from pushing students into medical degrees. However, I would push back further on your claim that several years of training and practice as a medic is not good experience for several highly impactful careers. I think that it is quite difficult to attribute how much highly impactful people who are medics can attribute their career success to medicine. For someone like yourself, it might be reasonably easy to say very little.; however, I think there are very likely other individuals for which this is not the case (In our experience, and in having interviews with lots of folk doing highly impactful things, this is the case. In fact, it rings true for several of the medics you mention).
3. For the case of Alice who seeks to maximise her impact as a doctor - I think Alice would disagree with you here and there are a bunch of high impact opportunities within medicine that we foreground:
Health policy work, which is often: 1. taken alongside clinical work 2. in some instances can be done well from a bottoms up approach while working in a clinical setting, 3. Often earmarks and wants the perspective of clinicians.
Effective giving and earning to give- I think doctors in a lot of countries, especially with growth of private medicine, have quite a high earning capacity. Especially for those medics who are further along or who really love their clinical work, earning to give within medicine seems like it might be the best EtG option
Working in a LIC, particularly if you are also involved in training local healthcare workforce, likely has 2-3 orders of magnitude more impact than being a junior doctor in a big metropolitan city (happy to send the studies that support this). I think you and I may disagree on the object level about whether this reaches the bar of “other EA jobs”- I think it might. In addition, if you take into consideration that not everyone has the capacity or capability to do other EA jobs, in a lot of cases, doing some of the above might be the highest impact opportunities for that person.
4. To address your concern about whether our messaging might lead someone to think that clinical practice in a LIC or a giving pledge are equivalently good to the other individuals you cite. Firstly I think this is highly unlikely given the content of our fellowship (with case studies from some of these individuals) and podcast (a number of these individuals are interviewed) and we make clear that there is a range of impact possible from a medical background. Secondly, in circumstances for which that is not feasible or realistic for an individual, or in which that is unlikely to be the case for them, we still think it is robustly good for them to increase their impact 2-3 orders of magnitude in the ways described above.
This is part of ‘big tent’ effective Altruism as we discuss above and may not be something you value as highly as we do. In general though I think this approach is likely to engage a whole lot more people initially than a ‘door in the face’ approach and that encouraging initial steps e.g. pledges or working in a LMIC will increase the likelihood of subsequent change, rather than feeling you have achieved your impact quota. We are reasonably confident that we portray the relative impact of different career options and decisions with high transparency, and enable people to maximise their impact as much they can—though perhaps we emphasise personal fit more than you would.
Speaking to all fellows at a 1:1 level after the fellowship and from the resources they have read over the fellowship, we think the risk we have lost ‘potential impact’ through individuals conflating different career options as equivalently good is very low—and rather the impact of individuals choosing to make impactful career decisions downstream predominates (hence this article).
Happy to chat more about all of this and thanks for your thoughts!
Howdy, and belatedly:
0) I am confident I understand; I just think it’s wrong. My impression is HIM’s activity is less ‘using reason and evidence to work out what does the most good’, but rather ‘using reason and evidence to best reconcile prior career commitments with EA principles’.
By analogy, if I was passionate about (e.g.) HIV/AIDS, education, or cancer treatment in LICs, the EA recommendation would not (/should not) be I presume I maintain this committment, but rather soberly evaluate how interventions within these areas stack up versus all others (with the expectation I would be very unlikely to discover the best interventions which emerge from this analysis will line up with what my passions previously aligted upon). Instead setting up a ‘Givewell for education interventions’ largely misses the point (and most of the EV ‘on the table’).
So too here. It would be surprising to discover medical careers—typically selected before acquaintance with EA principles—would be optimal or near-optimal by their lights (I’d be surprised if m/any EAs who weren’t already doctors thought it was). The face-value analysis is pessimistic on the ‘is this best’ question, notwithstanding (e.g.) there is a lot of variance within field to optimise: HIV/AIDS interventions vary in effectiveness by orders of magnitude, yet that doesn’t make them priorities on the current margin. As, to a first approximation, reality works in first-order terms, we’d want some very good reasons for second order considerations nonetheless carrying the day: sentiments like ‘big tent’, ‘EA is a question’ (etc.) can support anything (would it apply to PlayPumps?), so we should attempt to weigh these things up.
Your first point of clarification illustrates the ‘opacity’ I have in mind. “Not necessarily encouraging” folks to apply to medical school implies a lot of epistemic wiggle room: “Should I enter medicine?” and “Should I leave medicine?” are different but closely related questions (consider a 17 year old applying to medicine versus an 18 year old first year student), and answers to the former sense-check answers to the latter. If you really think having impact as a doctor is for many people some of the best things they can do, this suggests for similar people you would encourage them entering the profession (this doesn’t imply HIM should start doing this, but I think most in EA-land would find this result surprising and worth exploration—not least, it suggests a re-write of the 80k profile.) In contrast, if the answer is “even for those initially minded to enter medicine, we’d usually recommend against it as an EA career choice”, then there should be a story why this usual recommendation is greatly attenuated (or reversed?) for those already in the profession—particularly at an early stage like medical school. Again, this doesn’t govern HIM strategy—but it is informative, and knowing what you yourself think is the answer is important for transparent communication with your audience (even if they find this uncomfortable).
1) Regardless of the semantics of whether one should call someone like myself a ‘medic’ or not now, the substantive issue seems to be around whether medicine (generally speaking) is a high impact activity or not. Suppose (i.e. I’m not claiming this is the story for either of these professions I use as examples) (a):
‘High Impact law’: where the folks in the profession find their highest impact options often involve the practice of law in their ‘day job’, or ‘not strictly legal’ roles where their legal training is an important-to-crucial piece of career capital.
Contrast (b):
‘High Impact accountancy’: where folks in this profession find their highest impact options very rarely involve the practice of accountancy, and their best career options are typically those where their accounting background is only tangentially relevant (e.g. acquaintance with business operations, a ‘head for figures’).
In the latter case, ‘high impact accountancy’ looks like an odd term if the real message is to provide accountants with better career options which typically involve leaving the profession. If medicine was like (a), all seems well; but I think it is like (b), thus we disagree.
2) I’d be surprised if most of the folks I mentioned would find several years of medical experience valauble—especially (for the key question of career choice) whether this was a leading opportunity versus alternative ways of spending 10-20% of their working lives. I can ask around, but if you have testimony to hand you’re welcome to correct me. I’d guess medical experience is much more relevant for much more medically adjacent (or simply medical) careers—but, per grandparent, these careers tend to be a lot less impactful in the first place.
3) Our hypothetical Alice may be right about the options you note being ‘higher impact’ than typical practice. Yet effectiveness is multiplier stacking (cf.), so Bob (who doesn’t labour under the ‘having impact as a doctor’ constraint) can still expect 10-100x more impact. The latter two examples you give (re. earning to give and working in a LIC) allow direct estimation:
Re. E2G, US and UK doctors are in the top ~5% of their respective populations in earnings. Many other careers plausibly accessible to doctors (e.g. technical start-ups, quant trading, SWE, consulting) have income distributions which have either dramatically higher expected earnings, higher median earnings (e.g. friends of mine in some of these fields had higher starting salary than my expected peak medical salary), or both. This all sets aside that marginal returns to further money where there is a lot of aligned money looking for interventions to fund may be much lower now (cf. ‘earning to give’ careers typically finding themselves a long way down 80k recommendations; forum discourse ad nauseum about ‘talent constraint’, unease about all the lucre sloshing around, etc. etc.).
Re. LIC practice, if we take the 2-3 omag multiplier at face value (this looks implausible at the upper end), then combining that with 2ish DALYs/year of practice in a high income countries (taking my figures at face value, which are likely too high, you get 2*300 = 600 DALYs. In Givewell donations, with some conversion of (say) 40 DALYs = one ‘life saved’ (not wildly unreasonable as the lives saved are typically <5 year olds), this is ~~ 70 000 dollars/year. This is in the reach of E2G doctors (leave alone careers E2G more broadly), and the real number is almost surely lower (probably by an integer factor): the ‘medical practice’ side of the equation is much less rigorous than the givewell CEE, and should be anticipated to regress down.
As you say, various constraints (professional or personal) may rule out these other options: perhaps I aim at earning to give, but it happens that medical practice is my most lucrative employment (obviously much more plausible if one is later in one’s career); perhaps even if in general the sort of person drawn to medicine can make better contributions outside of the profession, this is not true for me in particular. Yet candour seems to oblige foregrounding such constraints often cut 90%+ of potential impact (and thus the importance of testing whether these constraints are strict).
4) Although comparators are tricky (e.g. if my writing on medical careers was vastly less effective it would be hard to tell), the content of the career plan changes noted in the OP would be more or less reassuring. re. what high impact med is accomplishing. Per above, as getting the last multipliers are important, HIM’s impact is largely determined by the tail of highest impact plan changes.
I think your two comments here are well-argued, internally consistent, and strong. However, I think I disagree with
in the context of EA career choice writ large, which I think may be enough to flip the bottom-line conclusion.
I think the crux for me is that I think if the differences in object-level impact across people/projects is high enough, then for anybody whose career or project is not in the small subset of the most impactful careers/projects, their object-level impacts will likely be dwarfed by the meta-level impact.
On the object-level for your examples, I think for “high-impact architecture,” having people with nontrivial background in architecture is likely useful for building civilizational refuges. More directly, I’ve talked to people who think that having 1-3 EA concierge doctors in the community (who can do things like understand our cultural contexts and weird problems and prescribe medicine in jurisdictions like the US and the UK) can be extremely helpful in increasing the impact of top talent in EA. This is analogous to the impact of e.g. existing community health or mental health workers in the community.
Potentially relevant subquestions:
To what extent does work in EA require EA alignment and acculturation?
The more you think EA orgs can hire well outside of EA for projects outside of EA natural core competencies, the more it matters that EAs target a relatively small subset of high-impact careers and skillsets to specialize in.
Conversely, if you think (as I do) that alignment and acculturation is just really important for excelling in EA jobs, it matters that we have people acquiring a wider scope of jobs and skillsets.
Do we live in a “big world” or a “small world” of EA things to do?
If we think there’s a narrow set of the best actions and causes, and a small number of people working in any of them, it matters more that individuals optimize for selecting the best things to do, on a birds’ eye view.
If, conversely, we think the range of really good actions and causes is relatively wide, then it matters more that individuals weigh factors like personal fit heavily.
An potential argument here is that the profile you wrote on doctoring was in the context of back when EA was much smaller. We may expect conditions “on the ground” to have changed a lot, and while “concierge EA doctor” would be a dumb career to aspire to five years ago, perhaps it is less so now.
(I personally think we likely still live in a relatively small world, which I think undercuts my counterarguments significantly).
Relatedly, how important is EA exploration vs exploitation?
How damning is the danger of introducing people with worse epistemics into the EA movement? And is worsening epistemics the most important/salient downside risk?
What are the best ways to prevent the above from happening?
Is it having really good first-order reasoning and arguments?
Is it having really good all-things-considered views that try to track all the important considerations, including rather estoric ones?
???
It seems bizarre that, without my strong upvote, this comment is at minus 3 karma.
Karma polarization seems to have become much worse recently. I think a revision of the karma system is urgently needed.
I have some hope that splitting out votes into two dimensions (approval and agreement) might help with situations like this. At least it seems to have helped with some recent AI-adjacent threads on LW that were also pretty divisive.
Yes, that is also my hope. Thanks for developing this.
It might this might just be the work one or two people. Maybe the mods can take a look?
We could create a script (using a sprinkling of NLP or classifier) to identify unreasonably downvoted comments and show how prevalent this is.
Thanks so much for writing this up! I really appreciate hearing more about what’s going on inside different groups. I’m especially excited to see the work being done in professional groups.
I think this is a really great approach. You get to ‘have your cake and eat it’ by connecting people interested in EA to it and nudging many more people to higher-impact option in a way that encourages a lot of autonomy and agency. I also quite like the approach mentioned in footnote #2.
Thanks Luke, I definitely think that autonomy and agency, particularly for professionals who are already established in a career, is a good approach to take, and might be a slight difference between community building in university/for professional groups (at least anecdotally, this is our experience).
And on footnote (2), I think this is actually something reasonably important I want to write more about-for instance, in our fellowship, we noticed that people responded well to information that was from well-known sources like high impact journals or news sources, and we found that if we had too many docs from sources they were unfamilar with (EA Forum), it bred some hesitation and skepticism. Although I think there is a wealth of absolutely fantastic material on ‘EA sources’, I think this is an important thing to be aware of in doing outreach to people unfamiliar with EA!
Maybe not directly relevant to what you wrote, but you can say what sort of things people have suggested for what high-impact medics might do and how this would be different from ‘regular’ medics? I’m just curious as I don’t know what this might be (although I hear leaving medicine to work at 80k is a pretty snazzy option ;) )
Yeah so our fellowship kind of dives into ways that medics can increase their impact, and EA cause areas where they might be particularly suited/have some comparative advantage. Some specific and non-exhaustive examples
For those interested in global health and development, medics might have a comparative advantage in working in the space due to clinical knowledge of area, respect and career capital
For those interested in health security/ biosecurity, there seems like there is an increasing role for medics who might act as a conduit between researchers and policymakers, as well as doctors and other public health professionals who are implementers on the ground.
Within clinical medicine, there might be robust ways to improve the amount of good people do e.g. through where they choose to work—i.e. LIC vs HIC, earning to give.
There are lots of skills that medics have (triage and prioritisation, systematic approach to problems, generally strong people skills) that make them a good fit for things like operations roles or careers advising e.g. at 80K ;)
If you are interested, happy to send you across our course curriculum!
I think we should move away from messaging like “Action X only saves 100 lives. Spending money on malaria nets instead would save 10000 lives. Therefore action X sucks.” Not everyone trusts the GiveWell numbers, and it really is valuable to save 100 lives in any absolute way you look at it.
I understand why doctors might come to EA with a bad first impression given the anti-doctor sentiment. But we need doctors! We need doctors to help develop high-impact medical interventions, design new vaccines, work on anti-pandemic plans, and so many other things. We should have an answer for doctors who are asking, what is the most good I can do with my work, that is not merely asking them to donate money.
I absolutely think we should stick to that messaging. Trying to do the the most good, rather than some good is the core of our movement. I would point out that there are also many doctors who were not discouraged and chose to change their career entirely as a result of EA. I personally know a few who ended up working on the very things you encourage!
That said we should of course be careful when discouraging interventions if we haven’t looked into the details of each cost-effectiveness analysis, as it’s easy to arrive at a lower looking impact simply due to methodological differences between Givewell’s cost-effectiveness analysis and yours.
Let’s separate this out
There are some medics who completely buy EA and have changed their entire career directly in line with EA philosophy
There are some medics who are looking to increase and maximise the impact of their careers, but who aren’t sold on all or aspects of EA. They also may have a particular cause area preference e.g. global medical education, that isn’t thought of as a high impact cause area by EAs
I think our philosophy is to work with both of these groups, rather than just (1).[1] I think the way we do that is by acknowledging that EA is fundamentally a question; we talk through EA ideology and frameworks without being prescriptive about the ‘answers’ and conclusions of what people should work on.
I think that this recent summary on a post on the forum is quite helpful here
We do fundamentally serve (1) and think this is a great group of people we shouldnt miss either!
Thanks for your comment and completely agree with you! I think the framing of what is the most I can do with my work is a great one that is underappreciated.
I really like framings which acknowledge how hard (emotionally) it can be to choose malaria nets.
I’d like to push back a bit on that—it’s so common in the EA world to say, if you don’t believe in malaria nets, you must have an emotional problem. But there are many rational critiques of malaria nets. Malaria nets should not be this symbol where believing in them is a core part of the EA faith.
I’m not saying that.
The point I was trying to make was actually the opposite—that even for the “cold and calculating” EAs it can be emotionally difficult to choose the intervention (in this case malaria nets) which doesn’t give you the “fuzzies” or feeling of doing good that something else might.
I was trying to say that it’s normal to feel like some decisions are emotionally harder than others, and framings which focus on that may be likely to come across as dismissive of other people’s actions. (Of course, i didn’t elaborate this in the original comment)
I don’t make this claim in my comment—I am just using malaria nets as an example since you used it earlier, and it’s an accepted shorthand for “commonly recommended effective intervention” (but maybe we should just say that—maybe we shouldn’t use the shorthand).
I think I sit somewhere between you both- broadly we think that there shouldn’t be “one” road to impact ; whether that be bed nets or something else Our explicit purpose is to use EA frameworks and thinking to help people reach their own conclusions. We think that common EA causes are very promising and Very likely to be highly impactful, but we err on the side of caution in being overly prescriptive.