Motivators: Altruistic Actions for Non-Altruistic Reasons


Jane is an effec­tive al­tru­ist: she re­searches, donates, and vol­un­teers in the high­est im­pact ways she can find. Jane has been in­tend­ing to write an effec­tive al­tru­ism book for over a year, but hasn’t man­aged to over­come the akra­sia. Jane then meets fel­low effec­tive al­tru­ist, Jes­sica, who she is keen to im­press. She starts writ­ing with pal­pable en­thu­si­asm.

In one pos­si­ble world:

Jane feels guilty that she has an im­pure mo­tive for writ­ing the book.

In an­other:

Jane is glad to lev­er­age the mo­ti­va­tion to im­press Jes­sica to help her do good.

In the past few months, I’ve heard mul­ti­ple peo­ple men­tion their use of less no­ble mo­ti­va­tions in or­der to get valuable things done. It ap­pears to be a com­mon ex­pe­rience among ra­tio­nal­ists and EAs, my­self in­cluded. The way I’m us­ing the terms, a rea­son for perform­ing some ac­tion is the os­ten­si­ble goal you wish to ac­com­plish, e.g. the goal of re­duc­ing suffer­ing. A mo­ti­va­tor for that ac­tion is an as­so­ci­ated re­ward which makes perform­ing the ac­tion seem en­tic­ing—“yummy”—e.g. im­press­ing your friends. I use the less com­mon term ‘mo­ti­va­tor’ to dis­t­in­guish the spe­cific mo­ti­va­tions I’m dis­cussing from the more gen­eral mean­ing of ‘mo­ti­va­tion’.

Many of our goals are mul­ti­ple steps re­moved from the ac­tions nec­es­sary to achieve them, par­tic­u­larly the broad-scale al­tru­is­tic ones. The goals are large, ab­stract, long-term, ill-speci­fied, difficult to see progress on, and un­in­tu­itively con­nected to the ac­tion re­quired. ‘I wrote a LessWrong post, is the world more ra­tio­nal yet?’ In con­trast, mo­ti­va­tors are tan­gible, im­me­di­ate, and typ­i­cally tickle the brain’s re­ward cen­tres right in the sweet spot. So­cial ap­proval, en­joy­ment of the ac­tion, money, skills gained, and oth­ers all serve as im­mi­nent re­wards whose im­me­di­ate an­ti­ci­pa­tion drives us. ‘Woohoo, 77 up­votes!’ Un­sur­pris­ingly, we find our­selves turn­ing to these im­me­di­ate re­wards if we want to ac­com­plish some­thing.

Note that a rea­son—the os­ten­si­ble goal—can still be the root cause of the de­sire to perform an al­tru­is­tic ac­tion. For one thing, if I didn’t truly care about all the things I say I care about do­ing, and re­ally only wanted so­cial ap­proval, then why join this par­tic­u­lar group of peo­ple out of all oth­ers?

Em­brace or Dis­grace?

If mo­ti­va­tors are the mechanism by which I’m get­ting things done, then I want to know ex­actly what’s go­ing on, what benefits I’m get­ting, and what costs I’m pay­ing. To date, I have seen peo­ple re­spond in two ways af­ter recog­nis­ing their mo­ti­va­tors for al­tru­is­tic ac­tion: i) by en­thu­si­as­ti­cally em­brac­ing the abil­ity of mo­ti­va­tors to spur good ac­tion, or ii) by feel­ing ab­ject shame and guilt at not act­ing for the right rea­sons.

The sec­ond re­sponse fol­lows from so­ciety’s con­ven­tional at­ti­tude to­wards any­thing it deems to be an im­pure mo­tive: ab­solute and un­re­strained damna­tion. Few things are con­sid­ered more evil than do­ing pub­lic good for per­sonal gain.

We have a visceral re­ac­tion to the idea that any­one would make very much money helping other peo­ple. In­ter­est­ing that we don’t have a visceral re­ac­tion to the no­tion that peo­ple would make a lot of money not helping other peo­ple. You know, if you want to make 50 mil­lion dol­lars sel­l­ing vi­o­lent video games to kids, go for it. We’ll put you on the cover of Wired mag­a­z­ine. But you want to make half a mil­lion dol­lars try­ing to cure kids of malaria, and you’re con­sid­ered a par­a­site your­self.
-- Dan Pal­lotta on The Way We Think About Char­ity is Dead Wrong

Any­one who has in­ter­nal­ised this can­not ac­knowl­edge their mo­ti­va­tors with­out ad­mit­ting to them­selves that they are a bad per­son. And to the ex­tent that we ex­pect oth­ers have in­ter­nal­ised this, we are re­luc­tant to dis­close our mo­ti­va­tors for fear of cen­sure. Even if you are not morally con­demned, act­ing solely for the benefit of your benefi­ciary is always con­sid­ered more praise­wor­thy than act­ing for the benefit of the benefi­ciary as well as your own gain. This is so much so, that of­ten peo­ple con­sider a char­i­ta­ble act praise­wor­thy only if the bene­fac­tor had no per­sonal gain. Hence the peren­ni­ally pop­u­lar ques­tion “Is true al­tru­ism ever pos­si­ble if you’re always get­ting some­thing out of it?”

Mo­men­tar­ily I will as­sert that so­ciety is fool­ish in this re­gard, but so­ciety’s fool­ish­ness typ­i­cally has an ex­pla­na­tion—of­ten that it was an at­ti­tude which was once adap­tive but is no longer so, or was adap­tive in a differ­ent con­text but it didn’t trans­fer. Given the ve­he­mence to­wards im­pure mo­tives, there might be some­thing to these rea­sons.

If I con­sider im­me­di­ate per­sonal re­la­tions, then I find that I would much pre­fer to be friends with some­one who wants to be my friend just be­cause they like me for me rather than with some­one who wants to be my friend, but has ad­mit­ted that she finds be­ing my friend a lot eas­ier be­cause she likes us­ing my swim­ming pool on these hot sum­mer days. The former friend’s friend­ship is more un­con­di­tional and trust­wor­thy—the lat­ter might desert me a soon as win­ter comes. That mo­ti­va­tors in­tro­duce an amount of con­tin­gency to one’s alle­giances is a point worth notic­ing.

In con­trast, the first re­sponse – en­thu­si­as­tic em­brace of mo­ti­va­tors – is the con­se­quen­tial­ist liber­a­tion from be­ing overly con­cerned with the mo­tives of the ac­tor. What mat­ters are the con­se­quences! And if mo­ti­va­tors mean more good things get done, well, then they get the Offi­cial Con­se­quen­tial­ist Stamp of Ap­proval. I don’t care if you cured malaria solely for profit, I just care that you cured malaria. But this point re­quires lit­tle push­ing in these parts.

It might even be that not only do mo­ti­va­tors provide a stronger drive for al­tru­is­tic ac­tion, but they are in fact the only way to get our­selves to act. Even if some parts of our minds can take on long-term, broad-scale, in­tan­gible goals, other parts just don’t speak that lan­guage. The rider might be able to en­gage in long term plan­ning, but if you want the elephant to budge, you’ve got to pro­duce some car­rots now.

An in­ter­est­ing aside, the use of mo­ti­va­tors to get things done may be more nec­es­sary for effec­tive al­tru­ists than the gen­eral al­tru­is­tic pop­u­la­tion. Warm fuzzies are great mo­ti­va­tors. Directly see­ing those you are helping at the lo­cal shelter or look­ing at a photo of the smil­ing child who you sent money to might provide a re­ward im­me­di­ate enough to re­quire no other. Whereas, when you’re donat­ing to cur­ing schis­to­so­mi­a­sis or re­duc­ing x-risk, the benefit is harder to feel and you’ve got to get a thumbs up from your friends in­stead to feel good.


While the above may be enough rea­son to en­dorse the use of mo­ti­va­tors to get things done, there is rea­son for cau­tion.

Mo­ti­vated Cognition

Fore­most, mo­ti­va­tors in­duce mo­ti­vated cog­ni­tion. When se­lect­ing al­tru­is­tic pro­jects, one’s choice be­comes dis­torted from what would ac­tu­ally have the high­est im­pact to that which has the strongest mo­ti­va­tor, typ­i­cally what will im­press peo­ple the most. Fur­ther­more, if mo­ti­va­tors are ac­cepted then it is le­gi­t­i­mate to in­clude them in the equa­tion. If one pro­ject has a greater mo­ti­va­tor and is more likely to get done be­cause of it, then even if it prima fa­cie isn’t the high­est im­pact, that like­li­hood of get­ting it done is strong fac­tor in its favour. But if this is ad­mis­si­ble rea­son­ing, it be­comes very easy to abuse: “Well sure this isn’t the high­est im­pact thing I could do, but vol­un­teer­ing at the lo­cal shelter is some­thing I feel mo­ti­vated to and ac­tu­ally will do, so there­fore it’s the thing I should do.”

Pre­tend­ing to Try

Th­ese points have already been iden­ti­fied in the dis­cus­sion of the ‘pre­tend­ing to try’ phe­nomenon.

A lot of effec­tive al­tru­ists still end up satis­fic­ing—find­ing ac­tions that are on their face ac­cept­able un­der core EA stan­dards and then pick­ing those which seem ap­peal­ing be­cause of other es­sen­tially ran­dom fac­tors.
-- Ben Kuhn on Pre­tend­ing to Try

Th­ese ran­dom fac­tors will be what­ever hap­pens to be the mo­ti­va­tors for a per­son. Nev­er­the­less, it is bet­ter that peo­ple do some­thing good rather than noth­ing. Katja Grace ar­gues that though this is what is go­ing on, it is both in­evitable and ac­tu­ally pos­i­tive. I am tempted to quote her en­tire post, so I sug­gest that you should read all of it.

‘Really try­ing’: di­rect­ing all of your effort to­ward ac­tions that you be­lieve have the high­est ex­pected value in terms of the rele­vant goals [not mak­ing de­ci­sions based on mo­ti­va­tors].

‘Pre­tend­ing to try‘: choos­ing ac­tions with the in­ten­tion of giv­ing ob­servers the im­pres­sion that you are try­ing [mak­ing de­ci­sions based on mo­ti­va­tors].

‘Pre­tend­ing to re­ally try‘: choos­ing ac­tions with the in­ten­tion of giv­ing ob­servers the im­pres­sion that you are try­ing, where the ob­servers’ stan­dards for iden­ti­fy­ing ‘try­ing’ are geared to­ward a ‘re­ally try­ing’ model. e.g. they ask whether you are re­ally putting in effort, and whether you are do­ing what should have high­est ex­pected value from your point of view.

-- Katja Grace on In Praise of Pre­tend­ing to Really Try

The pro­posed solu­tion is that we can lev­er­age the power of mo­ti­va­tors and still have peo­ple perform the high­est im­pact ac­tions they can, if we can cre­ate com­mu­nity norms whereby the the amount so­cial praise you get is pro­por­tional to the strength of your case for the im­pact of your ac­tion is.

I like this, but it hasn’t hap­pened yet and I sus­pect there are bar­ri­ers to mak­ing it work com­pletely. Even if your ac­tion is se­lected solely for im­pact—truly try­ing—the rea­son­ing be­hind your se­lec­tion might be com­pli­cated and re­quire time to fol­low and ver­ify. A few close friends might check your plans and ap­prove whole­heart­edly when you pre­tend to re­ally try, but the broader com­mu­nity won’t hear out all the de­tails spe­cific to your situ­a­tion, in­stead con­tin­u­ing to praise only ac­tions which fit the tem­plate of good effec­tive al­tru­ist be­havi­our, e.g. tak­ing a high pay­ing job in or­der to earn to give.

Pos­si­bly the best we can do for com­mu­nity norms is to find and spread the best sim­ple prin­ci­ples for de­cid­ing whether some­one is re­ally try­ing. The prin­ci­ples in­volved are un­likely to be ad­e­quately nu­anced to iden­tify the true op­ti­mum all of the time, but I think there’s room to im­prove over what we’ve fallen into so far. To date, I see praise be­ing given pri­mar­ily for ac­tions which are dis­tinc­tive EA be­havi­our and sig­nal be­long­ing to the tribe. Con­ven­tional al­tru­is­tic ac­tions like vol­un­teer­ing in the third world aren’t dis­tinctly EA and I don’t ex­pect them to get much praise, even if such ac­tions were the high­est im­pact for a par­tic­u­lar per­son. More likely, a per­son do­ing some­thing which doesn’t fit the EA mould will be in­ter­ro­gated for failure to con­form to what EAs are sup­posed to do.

Ne­glected Tasks

We can con­cretely see the im­pact a re­li­ance on mo­ti­va­tors has by notic­ing the many ne­glected tasks which re­sult. High value ac­tions which are not pres­ti­gious go un­done be­cause all they’ve got go­ing for them is their pure al­tru­is­tic im­pact.

The Cen­tre for Effec­tive Altru­ism has had a sur­pris­ing amount of trou­ble find­ing peo­ple to do what­ever im­por­tant work needs do­ing when it isn’t re­search or com­mu­ni­ca­tions. Th­ese things in­clude: or­ganis­ing in­surance, book­keep­ing and mak­ing pay­ments, main­tain­ing our databases, mak­ing de­liv­er­ies, or­der­ing equip­ment, find­ing and man­ag­ing places for peo­ple to live, ran­dom re­quests (e.g. cut­ting keys), re­ceiv­ing and pro­cess­ing mail, clean­ing, or­ganis­ing food and office events, etc.

It’s a bit of a shame that peo­ple seem will­ing to do what­ever is most im­por­tant… ex­cept when­ever it isn’t in­her­ently fun or pres­ti­gious!
-- Robert Wiblin

. . . I’ve had 200 vol­un­teers offer to do work for Sin­gu­lar­ity In­sti­tute. Many have claimed they would do “any­thing” or “what­ever helped the most”. SEO is clearly the most valuable work. Un­for­tu­nately, it’s some­thing “so mun­dane”, that any­body could do it… there­fore, 0 out of 200 vol­un­teers are cur­rently work­ing on it. This is even af­ter I’ve per­son­ally asked over 100 peo­ple to help with it.
-- Louie Helm

CFAR have made similar com­ments.

This is a se­ri­ous is­sue for a com­mu­nity claiming to be se­ri­ous about max­imis­ing im­pact.


Mo­ti­va­tors are suffi­ciently pow­er­ful, if not un­avoid­able, that we should al­low our­selves to work with them de­spite their dan­gers. The ques­tion be­comes how to use them while min­imis­ing their per­ni­cious effects. The ‘pre­tend­ing to try’ dis­cus­sion con­cerns the com­mu­nity col­lec­tively, but I am in­ter­ested in how in­di­vi­d­u­als should ap­proach their own mo­ti­va­tors.

I have a few ideas so far. The aim of these tech­niques is to limit the in­fluence mo­ti­va­tors have on our se­lec­tion of al­tru­is­tic pro­jects, even if we al­low or wel­come them once we’re onto im­ple­ment­ing our plans.


When choos­ing a course of ac­tion, pay at­ten­tion to what your mo­ti­va­tors might be. Ask ‘what are the per­sonal benefits I get out of this?’, ‘how much are these in­fluenc­ing my de­ci­sion?’ and ‘if this ac­tion A lacked the con­se­quence of per­sonal benefit B, would I still do it?’


At­tempt­ing gen­uine aware­ness re­quires a high-de­gree of self-hon­esty. You have to be will­ing to ad­mit that in all like­li­hood you have mo­ti­va­tors already and they in­fluence your de­ci­sions, and to ac­knowl­edge that even when you are try­ing to do good for oth­ers, you are in­ter­ested in your own gain. If this ad­mis­sion is hard, I sug­gest re­mem­ber­ing that this how peo­ple work, ev­ery­one else is the same and only the differ­ence is that you’re be­ing hon­est with your­self.

Choose, then Motivate

One strat­egy is to fully ac­knowl­edge to your­self that you want some im­me­di­ate per­sonal gain from your ac­tions, but de­lay­ing think­ing about that per­sonal gain un­til af­ter you have made a de­ci­sion about what to do based solely upon ex­pected im­pact. Once you’ve iden­ti­fied the high­est im­pact ac­tion, then brain­storm ways to find mo­ti­va­tors. This might in­volve noth­ing more than de­vel­op­ing a good fram­ing for your ac­tions which makes you look very no­ble in­deed. And most of the time that should be doable if you gen­uinely have a good rea­son.

Op­ti­mise Some­one Else’s Altruism

One more way to limit the in­fluence mo­ti­va­tors have over your de­ci­sion mak­ing is to pre­tend that you are de­cid­ing what some­one else – who is in ex­actly your situ­a­tion with ex­actly your tal­ents – should do to max­imise their im­pact. You are ad­vis­ing this other per­son, whose in­ter­ests you don’t care for be­cause they are not you, on how they might ac­com­plish the most can to­wards their goals.

Really Caring

Prob­a­bly the best way to en­sure that you re­ally try is to en­sure that you re­ally care. If you fo­cus on your rea­son for ac­tion—the out­come that it is re­ally about—then petty things like other peo­ple’s praise will feel that im­por­tant than ac­tu­ally ac­com­plish­ing your true goal.

Bring this feel­ing of car­ing to the fore of­ten. Are you try­ing to cure malaria? Keep a card with var­i­ous malaria statis­tics on your desk, read it of­ten, and re­mind your­self that you want stop those deaths which are hap­pen­ing right now. Care about the Far Fu­ture? Imag­ine your own fun-the­o­retic utopia, vi­su­al­ise it, and think about how good it would be to get there.


This is one at­tempt at get­ting a han­dle on mo­ti­va­tors and I am un­sure about much of it. There will be other an­gles to view this from, there are things I haven’t thought of, and mis­takes in some of my as­sump­tions. Plus, vari­a­tion in hu­man minds is as­tound­ing. Though many will ex­pe­rience mo­ti­va­tion the way that I do, oth­ers will find what I’m re­port­ing very alien. From them I would like to hear.

What I am sure about is that if we want to live up to our prin­ci­ples of do­ing what is truly most effec­tive, we can­not ig­nore the fac­tors driv­ing our be­havi­our. Here’s hop­ing that we can do what we re­ally need to do and feel max­i­mally good about it too.

Ac­knowl­edge­ments: I owe an enor­mous thank you to tkad­lubo and shok­wave for thor­oughly edit­ing this post.