Thoughts on 80,000 Hours’ research that might help with job-search frustrations

I in­tend to start work­ing at 80,000 Hours this Septem­ber, and in the mean­time they’re con­tract­ing me to write some ar­ti­cles about ca­reers and do­ing good, in­clud­ing this one. Nonethe­less, this ar­ti­cle rep­re­sents my per­sonal opinions only, and does not nec­es­sar­ily re­flect the views of the 80,000 Hours team.

An EA Fo­rum post from ear­lier this year demon­strated the difficulty of get­ting a job in effec­tive al­tru­ism or­gani­sa­tions, the frus­tra­tion many peo­ple feel as a re­sult, and the sense that other ways of do­ing good are not as highly val­ued by the EA com­mu­nity as they should be. Peo­ple in the com­mu­nity have pub­lished a num­ber of thought­ful re­sponses.

I am cur­rently do­ing some part-time con­tract work for 80,000 Hours, and I plan to start work­ing there full-time in the fall. As such, I’ve been think­ing a lot about 80,000 Hours’ re­search. And I wanted to add to the dis­cus­sion a few ways I think that 80,000 Hours con­tent might have in­ad­ver­tently con­tributed to these prob­lems, as well as some ideas for how peo­ple can get more out of their ad­vice.

The main take­aways are:

  • Roles out­side ex­plic­itly EA or­ga­ni­za­tions are most peo­ple’s best ca­reer op­tions.

  • Some­times these roles aren’t as visi­ble to the com­mu­nity, in­clud­ing to 80,000 Hours, but that doesn’t mean they aren’t highly im­pact­ful.

  • Many es­pe­cially im­pact­ful roles re­quire spe­cific skills. If none of these roles are cur­rently a great fit for you, but one could be if you de­vel­oped the right skills, it can be worth it to take sub­stan­tial time to do so.

  • You should use 80,000 Hours to figure out what your best ca­reer is and how to get there, not what “the” best ca­reers are.

I haven’t seen peo­ple talk that much about the last point, so I spend the most time on it.

Over-rep­re­sen­ta­tion of EA or­ga­ni­za­tions in 80,000 Hours content

Given un­limited re­sources, 80,000 Hours could cat­a­log ev­ery job op­por­tu­nity that might be some­one’s best op­tion, and then di­rect that per­son to­ward it. But 80,000 Hours is a small team, and has only been around for 7 years. Be­cause of this, their ideas and recom­men­da­tions should be treated as ten­ta­tive and grow­ing over time.

Not only that, they are grow­ing out­ward from the knowl­edge most cen­tral to EA. This means that 80,000 Hours is less likely to know about, and thus less likely to recom­mend, op­por­tu­ni­ties that are less fa­mil­iar within the EA com­mu­nity.

It will be rare for an op­por­tu­nity at an EA or­ga­ni­za­tion to es­cape their no­tice. But many great jobs in the wider world never come to 80,000 Hours’ at­ten­tion, and when they do, there may be no time to look into them.

Thus, op­por­tu­ni­ties at EA or­ga­ni­za­tions are more likely to be fea­tured—in write-ups, in the job-board, in coach­ing ad­vice—than op­por­tu­ni­ties at un­af­fili­ated or­ga­ni­za­tions that are less fa­mil­iar to the EA com­mu­nity. And this will be the case even if the roles at the less fa­mil­iar or­ga­ni­za­tions have higher po­ten­tial im­pact.

The con­trast be­tween ca­reer paths that 80,000 Hours ex­plic­itly recom­mends and those it doesn’t of­ten re­flects differ­ences in those paths’ effec­tive­ness, but some­times it just re­flects differ­ences in how much they’ve been vet­ted. Just as GiveWell’s recom­men­da­tions might be miss­ing an effec­tive non­profit be­cause they haven’t yet looked into it, so might the 80,000 Hours job board be miss­ing many promis­ing roles for high-im­pact work. And this is more likely when the role or the prob­lem it ad­dresses is less fa­mil­iar to the EA com­mu­nity, and so less likely to be re­searched by the 80,000 Hours team.

Talk of tal­ent gaps

One thing the origi­nal Fo­rum poster em­pha­sized is that be­cause they had heard there were “tal­ent gaps” in the EA com­mu­nity, they thought get­ting a job at an EA or­ga­ni­za­tion would be rel­a­tively easy so long as they were gen­er­ally ca­pa­ble.

80,000 Hours tried to ad­dress this is­sue in a Novem­ber ar­ti­cle, Think Twice Be­fore Talk­ing About Ta­lent Gaps: Clar­ify­ing Nine Mis­con­cep­tions. Ben Todd says that talk of “tal­ent gaps” at 80,000 Hours and el­se­where might have in­ad­ver­tently con­tributed to some mi­s­un­der­stand­ings. Here are three points from Ben’s post that help ex­plain why the ex­is­tence of a “tal­ent gap” doesn’t mean that work­ing at an EA or­ga­ni­za­tion is the right fit for most smart and ca­pa­ble peo­ple who are en­thu­si­as­tic about effec­tive al­tru­ism:

  • The ex­is­tence of “tal­ent gaps” doesn’t re­ally mean that there is a lack of tal­ent available. The term refers in­stead to a lack of spe­cific skills needed in the com­mu­nity. Hence, a bet­ter term is “skill bot­tle­necks.”

  • Skill bot­tle­necks in the com­mu­nity don’t mean that it will be easy to get hired at an EA or­ga­ni­za­tion. The rea­son is the point above: one can be hugely tal­ented in var­i­ous ways, but not have the spe­cific skills needed.

  • Im­por­tant skill bot­tle­necks also ex­ist in policy, academia, and other fields with­out ex­plicit EA or­gani­sa­tions. Th­ese are very im­por­tant and of­ten re­quire spe­cific skil­ling up out­side EA.

Th­ese clar­ifi­ca­tions were made only af­ter the con­fu­sion over tal­ent gaps be­came wide­spread, and Ben’s post didn’t help as much as 80,000 Hours might have hoped. The full post is worth check­ing out, both for jus­tifi­ca­tion of these po­si­tions and other points about how ‘skill bot­tle­necks’ seem to work in the EA com­mu­nity.

Less fo­cus on earn­ing to give

One rea­son 80,000 Hours started talk­ing about “tal­ent gaps” in the first place was to com­bat the mis­con­cep­tion that it, or EA in gen­eral, are just about earn­ing to give, and to high­light the need for peo­ple to do di­rect work as well. 80,000 Hours has been ag­gres­sive in its efforts here be­cause many peo­ple out­side the core EA com­mu­nity still per­ceive EA as pri­mar­ily about, or even syn­ony­mous with, earn­ing to give. And 80,000 Hours knows that its past em­pha­sis on earn­ing to give played a sub­stan­tial role in mak­ing this the case.

80,000 Hours’ push­back seems to have failed to con­vey nu­ance. Some peo­ple—es­pe­cially those who never had the pre­con­cep­tion that EA is pri­mar­ily about earn­ing to give—have con­cluded that it thinks earn­ing to give is a bad op­tion for al­most ev­ery­one, or that more fund­ing is not use­ful for the EA com­mu­nity. How­ever, that would be tak­ing things too far.

80,000 Hours thinks earn­ing to give is the best op­tion for a sub­stan­tial num­ber of peo­ple—those for whom it’s their com­par­a­tive ad­van­tage. They are keen, how­ever, to make sure that peo­ple fully con­sider di­rect work op­tions, in­stead of de­fault­ing to earn­ing to give be­cause they’ve heard it is the best way to do good with one’s ca­reer.

This points to a gen­eral difficulty: be­cause 80,000 Hours has a large and varied au­di­ence, with a wide range of pre­con­cep­tions about what makes for an effec­tive ca­reer, it is hard for it to com­mu­ni­cate equally well with ev­ery kind of reader. In try­ing to dis­abuse one group of peo­ple of a tena­cious mis­con­cep­tion, 80,000 Hours prob­a­bly in­ad­ver­tently cre­ated an­other mis­con­cep­tion among a differ­ent group of peo­ple.

Re­peat­ing “EA is not all about earn­ing to give” over and over again has a differ­ent effect on peo­ple who never thought that was the case, or who fol­low 80,000 Hours closely enough to hear the mes­sage many times, than it does on peo­ple who did think that was the case and only read some­thing about effec­tive al­tru­ism once a year.

The fo­cus on more se­nior roles

As the EA com­mu­nity has ma­tured, 80,000 Hours has shifted its main fo­cus to­ward filling as many mid-ca­reer (and some­times even se­nior) roles as it can. This is be­cause when the best pos­si­ble can­di­date fills one of these roles, not only can they have an out­sized di­rect im­pact, they can help other peo­ple have a greater im­pact too, through men­tor­ship, man­age­ment, bet­ter defin­ing prob­lems that oth­ers can then work on, and draw­ing ex­pe­rienced peo­ple into their field.

It used to be much harder to get mem­bers of the effec­tive al­tru­ism com­mu­nity into more ad­vanced po­si­tions, be­cause the com­mu­nity was mostly young and in­ex­pe­rienced. At one point, most of 80,000 Hours’ read­ers were in col­lege or had re­cently grad­u­ated. But that has changed slowly over time. 80,000 Hours’ au­di­ence is now a lot more di­verse, in­clud­ing not only young peo­ple but also plenty in grad­u­ate school or the mid­dle of im­pres­sive ca­reers.

80,000 Hours has re­cently fo­cused more on mak­ing ma­te­rial for the lat­ter groups, partly be­cause that con­tent doesn’t ex­ist el­se­where, and partly be­cause filling se­nior roles is es­pe­cially im­pact­ful. But this may have caused read­ers ear­lier in their ca­reers to be dis­cour­aged by con­tent that isn’t chiefly aimed at them. Again, the fact that the 80,000 Hours au­di­ence is di­verse re­sults in the same con­tent be­ing very helpful for some, but un­helpful or even coun­ter­pro­duc­tive for oth­ers.

Ad­vice about build­ing ca­reer cap­i­tal by start­ing at the bot­tom of an or­ga­ni­za­tion should have com­bated this is­sue, by show­ing peo­ple who are less ad­vanced in their ca­reers how they should work to­ward more se­nior roles. But this ad­vice wasn’t always em­pha­sized enough. And be­cause of the mis­con­cep­tions about “tal­ent gaps,” es­pe­cially within EA or­ga­ni­za­tions, some might have got­ten the im­pres­sion that start­ing in a more en­try-level role wasn’t go­ing to be nec­es­sary.

Most read­ers who are still early in their ca­reers must spend con­sid­er­able time build­ing tar­geted ca­reer cap­i­tal be­fore they can en­ter the roles 80,000 Hours pro­motes most. This might be frus­trat­ing, or make peo­ple feel like they’re “not do­ing enough.” But build­ing ca­reer cap­i­tal that’s rele­vant to where you want to be in 5 or 10 years is of­ten ex­actly what you should be do­ing.

The “big list” view of do­ing the most good with your ca­reer (and why it doesn’t make sense)

One way to think about do­ing good with our ca­reers is to pic­ture a big list of all the ca­reer paths in the world any­one could pur­sue, or­dered from most to least im­pact­ful. The higher our own ca­reer ranks, the more good we’re do­ing, and the bet­ter we should feel about our­selves, the more peo­ple should re­spect us, and so on.

This is a bit of a car­i­ca­ture, but it doesn’t seem that far from the way peo­ple some­times think. I know I of­ten think in this way. And it can some­times feel like 80,000 Hours’ ca­reer ad­vice is ar­tic­u­lat­ing a big list like this. More­over, once we’re think­ing in this way, it’s easy to feel like where our path ranks on such a list should have im­por­tant con­se­quences for our so­cial sta­tus or self-es­teem.

Still, I think this “big list” pic­ture is wrong, for three rea­sons:

1. It imag­ines a sin­gle, unified rank­ing of ca­reer paths or­dered from most to least im­pact­ful.

But the com­pli­ca­tion of in­ter­ac­tion effects be­tween differ­ent peo­ple’s ac­tions makes me skep­ti­cal that there is such a unified rank­ing, even in the­ory.

The im­pacts of differ­ent peo­ple’s paths and ac­tions of­ten de­pend on one an­other, such as in cases of co­op­er­a­tion or trade. Some peo­ple’s ac­tions might de­ter­mine the pos­si­bil­ities for other peo­ple’s ac­tions or make them more or less effec­tive. Th­ese fac­tors makes it very hard to mean­ingfully rank them.

For ex­am­ple, how would you rank ac­tions by mul­ti­ple peo­ple that are all nec­es­sary con­di­tions for any of them to have an im­pact? Whose ac­tions were more im­pact­ful, Nor­man Bor­laug’s or Nor­man Bor­laug’s mother’s?

2. For ev­ery reader, such a list would in­clude many paths that they can’t take.

Most of the ca­reer paths on a big list of ev­ery­one’s pos­si­ble paths will be ir­rele­vant to you, be­cause ob­vi­ously your op­tions are limited by your cir­cum­stances. More on this be­low.

3. The “big list” pic­ture makes an un­war­ranted as­sump­tion about how we should feel about our­selves and oth­ers.

Many peo­ple in the EA com­mu­nity think that the best ac­tions to take in any given situ­a­tion are those with the most pos­i­tive im­pact. But it doesn’t fol­low that the peo­ple we should hold in high­est es­teem, or who should be most proud, are those whose ac­tions have the most pos­i­tive im­pact.

I don’t claim to know ex­actly how we should think about the crite­ria for things like es­teem and pride. But I don’t think these should be pro­por­tional only to how much good a per­son has done.

Differ­ent peo­ple have differ­ent op­tions available to them. Maybe some ca­reer path would be re­ally high-im­pact, but it’s im­pos­si­ble for some­one to take, for what­ever rea­son. From a util­i­tar­ian per­spec­tive, it clearly doesn’t make sense for them to feel bad about not tak­ing that path, even if it would have done a lot of good had they been able to do so. Feel­ing bad won’t make them any more able to take the path, nor does it seem to make things over­all bet­ter in other ways.

From a non-util­i­tar­ian per­spec­tive, it doesn’t seem like how much good I do fully de­ter­mines how I should feel about my­self ei­ther. I hope that in 100 years most peo­ple will be do­ing much more good than I can to­day be­cause they have greater wis­dom, new tech­nol­ogy, and bet­ter co­or­di­na­tion. Should this com­par­i­son with the big pos­si­ble im­pacts of fu­ture gen­er­a­tions make us feel worse about our­selves and one an­other? I don’t think so. How I should feel about my­self and how oth­ers should feel about me are not just func­tions of how much good I do.

De­spite these flaws, the “big list” pic­ture seems like it might be play­ing a part in the frus­tra­tion and dis­ap­point­ment that many peo­ple feel in the highly com­pet­i­tive EA job mar­ket. It has definitely played a part in my own ex­pe­rience.

Of course, get­ting re­jec­tions is never easy. But if you read 80,000 Hours as offer­ing a “big list” of the best ca­reers, and you feel like your sta­tus with your­self or your com­mu­nity is tied up with do­ing the top things on that list, that makes get­ting re­jected from one of those top things feel even worse.

The “per­sonal list” view

It seems to me that the way to think about do­ing the most good you can do with your ca­reer is to put the em­pha­sis on the you. Each per­son has their own list of pos­si­ble ca­reers available to them, ranked from best to worst. What each per­son’s list looks like de­pends on:

  • The op­tions available to them.

  • How other peo­ple will act given what they do.

Our am­bi­tion should be to do the best things on our per­sonal lists.

Be­cause each per­son’s list only in­cludes ac­tions available to them, and be­cause the rank­ing is de­ter­mined in part by what other peo­ple will do in re­sponse to each ac­tion, this ap­proach doesn’t have the prob­lems that the “big list” pic­ture does.

But it’s re­ally hard to figure out what is on your per­sonal list, or how the differ­ent op­tions com­pare to each other.

We should read 80,000 Hours as try­ing to help peo­ple figure out what their per­sonal lists look like. Some­times they do this by giv­ing gen­eral ad­vice about how to figure out what op­tions are on your list and how they rank. This of­ten in­volves tools for as­sess­ing your skills and in­ter­ests, or for un­der­stand­ing how your ac­tions could af­fect other peo­ple.

Some­times 80,000 Hours tries to help peo­ple with their lists by pro­mot­ing par­tic­u­lar ca­reer paths that they think do be­long high on the lists of some peo­ple who haven’t yet re­al­ised it. But if, af­ter some in­ves­ti­ga­tion, you find some path they talk about isn’t on your per­sonal list, that shouldn’t make you feel bad.

Some­one might ob­ject: if we don’t make it feel bad to not do the most im­pact­ful things, how will we be mo­ti­vated to do the most good?

But as I said above, it is im­pos­si­ble to mo­ti­vate some­one to do some­thing that sim­ply isn’t an op­tion for them. At most, peo­ple should feel bad for not do­ing the top things on their per­sonal lists, or not try­ing to figure out what those things are.

A differ­ent ob­jec­tion is that maybe what my list looks like is it­self grounds for pride or es­teem.

This has some in­tu­itive ap­peal, but seems un­jus­tified to me. It can’t be jus­tified in terms of mo­ti­va­tion, for the rea­son just given. And I don’t see any other available ar­gu­ment. Usu­ally we think highly of peo­ple for do­ing the right ac­tion among a set of op­tions available to them, or for fol­low­ing the rea­sons they have to think or act in a cer­tain way. But I don’t know what would jus­tify think­ing highly of some­one just for their list be­ing what it is.

I ac­tu­ally think the EA so­cial com­mu­nity is gen­er­ally on board with these points. Peo­ple don’t go around try­ing to make peo­ple feel bad be­cause they aren’t as im­pact­ful as some­one else, and we rec­og­nize that im­pact­ful work is of­ten in­ter­de­pen­dent in com­plex ways. But it’s im­por­tant to be vigilant be­cause it’s pretty easy to fall into the “big list” way of think­ing.

In sum: 80,000 Hours’ re­search does not and can­not yield a “big list” of the best ca­reer paths, be­cause no such thing ex­ists. In­stead, we should use 80,000 Hours con­tent to map out our own per­sonal lists and figure out how to do the top things on them.

Thanks to Robert Wiblin and Howie Lem­pel for feed­back on a draft of this post.