Do EAs underestimate opportunities to create many small benefits?

Dis­claimer: I origi­nally wrote this for the EA Face­book group since it’s a quick and un­pol­ished thought, but I wanted to post it here in­stead be­cause it got long, I think the dis­cus­sion is bet­ter here, and I like the easy way to link to the post af­ter­wards. Let me know if you think it is bet­ter suited for the EA Face­book group.


Right now, EAs fo­cus pri­mar­ily on global poverty, an­i­mal welfare, re­duc­ing ex­is­ten­tial risk, and meta ac­tivi­ties to im­prove the first three . How­ever, some EAs are in­ter­ested in look­ing for more. What other im­por­tant causes might we be over­look­ing?

One idea I’ve been re­cently think­ing about is the op­por­tu­nity to pro­duce many small benefits may be very ne­glected and pro­duce op­por­tu­ni­ties for a “fifth fo­cus area”.


EA has Many Small Benefits vs. One Large Benefit

I’m cu­ri­ous if peo­ple think EAs sys­tem­at­i­cally un­der­es­ti­mate the im­pact of cre­at­ing a tiny benefit for a nu­mer­ous amount of peo­ple rather than cre­at­ing large benefits.

An ex­am­ple of cre­at­ing a large benefit for a rel­a­tively small group of peo­ple is to give a $1K cash trans­fer to a sin­gle Kenyan. An ex­am­ple of cre­at­ing a small benefit for a large group of peo­ple would be to im­prove the speed of the in­ter­net across the en­tire United States or to spend time im­prov­ing an im­por­tant open source pro­ject like Ruby on Rails. It’s pos­si­ble that re­duc­ing ex­is­ten­tial risk by a tiny amount could be con­sid­ered mak­ing a small benefit for many peo­ple, though I don’t re­ally think of it as such.

Of course, it’s also cer­tainly pos­si­ble to cre­ate large benefits for a large group of peo­ple, such as by end­ing fac­tory farm­ing.


Why might it be un­der­es­ti­mated?

When you’re at a cross­walk and a bus ap­proaches, you can ei­ther wait for the bus to pass and then cross, or cross and make the bus wait for you. If you wait, you save one minute of time. If you make the bus wait, each per­son on that bus waits for one minute of time. If the bus has sixty peo­ple in it, that’s an hour that was just spent wait­ing. That sounds like a lot when framed in those terms, but it’s not some­thing we ever think about when cross­ing the street. Is mak­ing the bus wait for you re­ally that self­ish?

Per­haps not. Per­haps there are still good rea­sons to cross the street in­stead of mak­ing the bus wait—mainly, that’s usu­ally what the bus ex­pects. How­ever it ad­dresses the most difficult part of util­i­tar­i­anism—un­der­stand­ing that giv­ing +1 unit of util­ity to 10,000 peo­ple is more im­por­tant than giv­ing 9,000 units of util­ity to one per­son (see “Tor­ture vs. Dust Specks”). The “many small benefits” ap­proach is very un­in­tu­itive. Even EAs have bi­ases and find this un­in­tu­itive, so I would ex­pect many peo­ple (in­clud­ing my­self) to un­der­es­ti­mate op­por­tu­ni­ties to cre­ate many small benefits.

Fur­ther­more, im­pact­ing many peo­ple a small amount doesn’t fit well into the non-profit frame­work to de­liver clear and un­der­stand­able value to a clearly defined and un­der­stood pop­u­la­tion of peo­ple.

It is difficult to mea­sure the im­pact of many small benefits. How much benefit does an hour of time spent de­vel­op­ing Ruby on Rails provide to the Ruby on Rails pro­ject? How much value does Ruby on Rails provide pro­gram­mers and peo­ple mak­ing tech­nol­ogy com­pa­nies? How much value do those tech­nol­ogy com­pa­nies provide peo­ple around the world? How does the sum of all that value com­pare to giv­ing $100 (an hour of a typ­i­cal pro­gram­mer’s salary) as a cash trans­fer to a Kenyan?

Lastly, the “many small benefits” ap­proach seems to be the jus­tifi­ca­tion for a large amount of non-EA ac­tivi­ties , such as those who think we should im­prove “arts and cul­ture”. EAs may pat­tern match these bad ar­gu­ments for non-EA ac­tivi­ties to peo­ple ar­gu­ing for “many small benefits” from a EA per­spec­tive, and it could be hard to figure out which “many small benefits” ap­proaches fit into the best op­por­tu­ni­ties to help the world.

Mea­sure­ment difficulty and it not fit­ting into a tra­di­tional char­ity frame­work may be gen­uine rea­sons to fa­vor the “large benefits” ap­proach, but I don’t think they’re enough to make the “many small benefits” ap­proach not worth think­ing about. We should aim to study how some “many small benefits” ap­proaches may im­pact the world over­all more pos­i­tively than our cur­rent EA ac­tivi­ties.


What are some good ex­am­ples of “many small benefits” that EAs could pur­sue?

Some ex­am­ples I could think of that seem worth re­search­ing, say as part of the Open Philan­thropy Pro­ject:

* Con­tribut­ing to good open source pro­jects could make it much eas­ier to cre­ate and im­prove com­pa­nies and/​or save hun­dreds of thou­sands of hours in de­vel­oper time. The en­tire se­cu­rity of the in­ter­net could be at stake and I’m not the only one who thinks open source may be uniquely ne­glected in our cur­rent for-profit and non-profit fund­ing en­vi­ron­ment .

* Re­duc­ing the amount of traf­fic through bet­ter trans­porta­tion in­fras­truc­ture could save hun­dreds of thou­sands of hours in lost com­mut­ing time.

* Some changes to how VC fund­ing works could im­prove the ac­cu­racy of VC cap­i­tal, im­prov­ing the al­lo­ca­tion of trillions of dol­lars. Other high-lev­er­age changes may be pos­si­ble in other ar­eas of fi­nance as well.

...Of course, some of these may already re­ceive sig­nifi­cant fund­ing and many of these may not, on fur­ther re­flec­tion, ac­tu­ally be worth it when com­pared to our ex­ist­ing fo­cus ar­eas. How­ever, I think some of these are worth think­ing about more when they could be im­por­tant, tractable, and ne­glected.