The Case for the EA Hotel

Epistemic Sta­tus: I strongly be­lieve all the things I’m writ­ing here. Th­ese are mostly heuris­tics and men­tal mod­els rather than hard data, which I think is nec­es­sary for a pro­ject so young. I’m try­ing to make a strong case for the EA ho­tel, not a bal­anced one (al­though it will prob­a­bly be bal­anced by the $100 on the line for ar­ti­cles tak­ing the op­po­site view).

The EA Chasm

There’s some­thing bro­ken about the pipeline for both tal­ent and pro­jects in the EA Com­mu­nity. There’s a space in which there’s a lot of tal­ented peo­ple in EA who want to do good, and there’s a lot of peo­ple with ideas about pro­jects that could do good. Fi­nally, pro­jects like Char­ity En­trepreneur­ship seem to in­di­cate that there’s no short­age of ways to do good. What’s miss­ing is a way to go from a tal­ented EA—with no ev­i­dence be­hind your pro­ject, no pre­vi­ous pro­jects un­der your belt, and lit­tle sta­tus within the EA com­mu­nity—to some­one who has enough cap­i­tal to prove that their pro­ject has merit.

This gap ex­ists for a num­ber of rea­sons, in­clud­ing strong risk aver­sion in the EA com­mu­nity, a lack of di­ver­sity in grant de­ci­sion mak­ing pro­cesses, and a lack of man­power to vet hun­dreds of pro­jects for the small amount of money they would need to prove them­selves enough to move up to the “pro­jects with strong ev­i­dence” cat­e­gory. A num­ber of solu­tions have also been pro­posed to fill in this gap, in­clud­ing an EA pro­jects eval­u­a­tion plat­form and a sug­ges­tion for EAs to work on Non-EA pro­jects in or­der to get a good track record and higher sta­tus (and thus be able to be hired or get grants). How­ever, both of these sug­ges­tions miss out on one of the big rea­sons the chasm needs to be filled—strong vet­ting is nice, but there’s no re­place­ment for sim­ply try­ing many things and see­ing what works.

Why The Chasm Matters

This Chasm is a big deal for the com­mu­nity. Or­ga­ni­za­tions like CEA can work to guide the com­mu­nity to­wards a bet­ter fu­ture, and or­ga­ni­za­tions like Char­ity En­trepreneur­ship can slowly work to al­low more or­ga­ni­za­tions that do good work. But by not tap­ping into the cre­ativity and sheer va­ri­ety of thought of the bot­tom two sec­tions of the pic­ture above, the EA com­mu­nity is los­ing out on a large num­ber of utils that come from try­ing a lot of things from a di­ver­sity of per­spec­tives, cre­at­ing tight feed­back loops, and see­ing what works.

Sili­con Valley is great proof of this con­cept. While it’s true that the stan­dards for seed fund­ing have been grow­ing in re­cent years (and this may be an­other fac­tor in the EA model, if they’re try­ing to copy Sili­con Valley), it’s also true that pre­seed ac­cel­er­a­tors with ex­tremely low vet­ting stan­dards have still gen­er­ated tens of billions of dol­lars worth of value. EA, with a sur­plus of ideas that don’t have cap­i­tal to get off the ground, and a sur­plus of tal­ented in­di­vi­d­u­als will­ing to work on these ideas, should view this is a ne­glected op­por­tu­nity to do a lot of good for the world. And they should view the EA Ho­tel as a won­der­ful proof of con­cept for an or­ga­ni­za­tion look­ing to fill in this Chasm.

The EA Ho­tel is More Effec­tive Than Directly Spon­sor­ing In­di­vi­d­u­als or Projects

One way to view the EA ho­tel is as a grant giv­ing or­ga­ni­za­tion that pays for peo­ple’s liv­ing ex­penses for a pe­riod of time, while those peo­ple have op­por­tu­ni­ties to prove that their pro­jects are good enough to get to the next stage of fund­ing. For EAs who are still look­ing for pro­jects, it pro­vides a bridge to fo­cus on gain­ing skills and knowl­edge while get­ting chances to join new pro­jects as they cir­cu­late through the ho­tel.

When the EA ho­tel is looked at in this light, the ques­tion then be­comes “does it make more sense to fund in­di­vi­d­ual pro­jects and EAs, rather than let­ting the EA ho­tel fund them for you?” The EA Ho­tel has sev­eral fea­tures that make it a more effec­tive op­tion.

No Rent

The largest liv­ing ex­pense for most peo­ple (es­pe­cially the large num­ber of EAs in Lon­don, Oxford, and the Bay Area) is rent. When spon­sor­ing some­one your­self, most of your money will be go­ing into that black hole. The EA ho­tel has done the effi­cient thing and bought the ho­tel out­right. This means that rent is not some­thing you as a fun­der have to pay, and the longer the ho­tel lasts and the more res­i­dents it helps, the more effi­cient this mechanism be­comes over pay­ing rent.

Cheap Cost of Liv­ing and Lower Stan­dard of Living

One unique thing about the EA ho­tel as a grant-giv­ing mechanism is that it forces the res­i­dents to move to Black­pool. While there are some down­sides to this, I think there are two huge up­sides from a cost-effec­tive­ness per­spec­tive. The first is that the cost of liv­ing is ex­tremely low. Just like with rent, fund­ing the EA ho­tel here con­sis­tently makes your money go fur­ther than fund­ing a ran­dom EA who would choose their own place to live.

Another im­por­tant fact is that stan­dard of liv­ing here is sim­ply lower. While try­ing to be ex­tremely fru­gal in San Fran­cisco, I couldn’t help but no­tice that my stan­dard of liv­ing and hap­piness was im­pacted by those around me. How­ever, as a con­se­quence of liv­ing in Black­pool, and a sec­ondary con­se­quence of only hav­ing my sav­ings and a small liv­ing stipend, I’ve found that I’ve been hap­pier with a much cheaper stan­dard of liv­ing in Black­pool. There’s some data that shows that how stan­dard of liv­ing im­pacts hap­piness is rel­a­tive to oth­ers in your im­me­di­ate en­vi­ron­ment, and is not ab­solute. This means that I can be happy and pro­duc­tive at a much cheaper cost at the EA ho­tel than at a group house in Berkeley, and your dona­tion dol­lars can stretch fur­ther.

Propin­quity and Collaboration

By putting all of the pro­jects to­gether un­der the same roof, the EA ho­tel does an ex­cel­lent job of fos­ter­ing con­nec­tions, en­courag­ing col­lab­o­ra­tions, and cre­at­ing a strong en­vi­ron­ment for serendipity and syn­ergy among pro­jects. In my short time here, I’ve seen a method­ol­o­gist help an or­ga­ni­za­tion with de­sign­ing their RCT, a coder help a differ­ent or­ga­ni­za­tion au­to­mate one of their biggest bot­tle­necks, and an or­ga­ni­za­tion which needed help on mea­sur­ing im­pact get help from some­one who had writ­ten an im­por­tant pa­per on the mat­ter. More im­por­tantly than these in­di­vi­d­ual col­lab­o­ra­tions, I’ve seen peo­ple’s ideas grow and de­velop as they get ex­posed to cri­tiques and new ways of think­ing. This is an effect you sim­ply don’t get if you spon­sor pro­jects sep­a­rately in­stead of as a group.

Su­per­con­nect­ing and Sta­tus Building

The fi­nal thing I’ve seen from the EA ho­tel is that, while be­ing in a cheap, out-of-the-way city, it’s enough of a unique at­trac­tion (and there’s always enough free rooms available) that it has be­come a ‘des­ti­na­tion’ for EAs to check out when they’re in Europe. This is an im­por­tant fact, as nor­mally one of the benefits of be­ing in a more ex­pen­sive city (and one of the rea­sons most startup in­cu­ba­tors are lo­cated there) is that it al­lows you to be­gin build­ing con­nec­tions with the peo­ple you’ll need to know when mov­ing to the next stage of the pyra­mid. How­ever, by hav­ing the “ho­tel” as­pect, and be­com­ing a des­ti­na­tion, the EA ho­tel man­ages to at­tract a steady stream of in­di­vi­d­u­als from all as­pects of the EA com­mu­nity. It has man­aged to be­come an effec­tive net­work­ing hub while be­ing in a city with a cheap cost of liv­ing, and has achieved some­thing for pro­jects that merely fund­ing them to live on their own could not.

The EA Ho­tel Is An Effec­tive Incubator

Thus far, I’ve made the case that there’s a sur­plus of po­ten­tial in the EA com­mu­nity, and a Chasm that needs to be filled to use the sur­plus. I’ve also made the case that some­thing like the EA Ho­tel is an effec­tive way to fill that Chasm. What I haven’t done is make the case that this par­tic­u­lar team and pro­ject have done a good job of re­al­iz­ing that goal.

In the fol­low­ing sec­tion, I’ll at­tempt to give my in­side view of why I be­lieve the pro­ject and team are suited for filling the goal, as a 3-month res­i­dent of the ho­tel, and some­one who has wit­nessed and cre­ated other teams and cul­tures.

Cor­rect Ac­cep­tance Standards

One per­sis­tent crit­i­cism of the ho­tel is that it has too low stan­dards for what pro­jects it ac­cepts. How­ever, the stan­dard that the ho­tel has (ac­cept ev­ery­one when there’s space, and only pri­ori­tize when they’re over ca­pac­ity) is the cor­rect choice for an or­ga­ni­za­tion that’s try­ing to fill the Chasm like the EA ho­tel is.

Let’s re­turn to our Sili­con Valley metaphor, and the pre-seed in­cu­ba­tor I al­luded to ear­lier, The Founders In­sti­tute. The Founders In­sti­tute, while I don’t think they ad­mit it pub­li­cly, has a similar policy of ac­cept­ing as many can­di­dates as there are slots, and try­ing to max­i­mize the amount of pro­jects rather than hav­ing some per­ceived qual­ity cut­off. The Founders In­sti­tute knows two things.

  1. At this stage in their ca­reer, it’s very hard to vet first time founders. Without a track record, all they have to go on is charisma and clar­ity of thought, which is ac­tu­ally some­thing that many first time founders will only learn only through the pro­cess of cre­at­ing their first startup.

  2. Some­times the best ideas look com­pletely ridicu­lous. Con­sider the idea of cre­at­ing a web­site where strangers can rent out their homes to other strangers.

So in­stead, the Founders In­sti­tute does some­thing else—it im­ple­ments a se­ries of tight feed­back loops and stan­dards, caus­ing founders to have to prove both them­selves and their pro­jects to grad­u­ate the pro­gram. While the ac­cep­tance rate for the Founders in­sti­tute is very high, the grad­u­a­tion rate is only around 30%. The hope is that most of those 30% have achieved enough in their pro­ject to get them a more tra­di­tional seed round.

Similarly, the EA ho­tel has weekly check-ins to gauge the progress of their par­ti­ci­pants, and is work­ing on im­ple­ment­ing more stringent feed­back loops for the peo­ple who en­ter the ho­tel. The goal, in­stead of try­ing to vet the peo­ple and pro­jects up front, is to use the pro­cess it­self to vet the pro­ject and the in­di­vi­d­ual. As they pass in­creas­ingly high bars, they even­tu­ally cross the bar where they achieve good ev­i­dence for their pro­ject, and can then move on to the next stage of the pyra­mid. If it turns out they can’t meet that bar, they go back down to the pre­vi­ous stage of the pyra­mid, work on lev­el­ing up, and try again when they think they’re ready.

Re­mov­ing Triv­ial Inconveniences

As a cre­ator or early par­ti­ci­pant in a new pro­ject, fo­cus is ev­ery­thing. Time and at­ten­tion are wasted when put to­ward things other than those that di­rectly work to im­pact your biggest met­rics, or val­i­date your biggest as­sump­tions. Fur­ther­more, the type of work you have to do to val­i­date or in­val­i­date these as­sump­tions is scary, hard, and of­ten emo­tion­ally drain­ing. Every lit­tle bit of en­ergy that you can save by not hav­ing to deal with triv­ial in­con­ve­niences is a bless­ing.

At the EA Ho­tel, my gro­cery shop­ping is taken care of for me. Din­ners are cooked for me. Grab and go food for break­fast and lunch is re­stocked with­out my hav­ing to think about it. My dishes are done for me. My sheets are changed for me. All of this al­lows me to avoid an in­cred­ible amount of con­text switch­ing that sim­ply doesn’t have to hap­pen be­cause the EA Ho­tel rec­og­nizes the im­por­tance of fo­cus. Fur­ther­more, they’re always im­prov­ing. A big por­tion of the man­agers’ job is find­ing triv­ial in­con­ve­niences and re­mov­ing them. Areas get more or­ga­nized over time, sys­tems get re­fined over time, busy work gets re­moved over time. This is ex­actly the en­vi­ron­ment that I ex­pect to be able to more effec­tively cre­ate valuable pro­jects over time.

A Pro­duc­tivity Culture

I’ve been a part of sev­eral group houses in which a large por­tion of the peo­ple who lived there worked from home. I’ve been a part of at least one at­tempt to in­still a strong cul­ture of work­ing hard when in the pres­ence of your peers. There’s only one cul­ture that I’ve been a part of that I think has more of a cul­ture of pro­duc­tivity than the EA Ho­tel, and I’d say it’s in the top 5-10% of cre­at­ing and sus­tain­ing strong or­ga­ni­za­tional cul­tures, in­clud­ing both for-prof­its and non-prof­its (so much so that it has been called cult-like).

I think that the at­ti­tude to­wards triv­ial in­con­ve­niences is a big part of this. The idea that the man­age­ment is clear­ing so much space for work cre­ates a cul­ture where ev­ery­one is sim­ply work­ing. I should note that when bring­ing this to the EA ho­tel, there was at least one per­son who said he doesn’t be­lieve the EA Ho­tel has enough of a pro­duc­tivity cul­ture. When pol­led, ev­ery­one else pre­sent agreed that they’re more pro­duc­tive here than they have been in any other con­text. This is a huge boon for an or­ga­ni­za­tion try­ing to vet pro­jects as quickly as pos­si­ble, and some­how the EA ho­tel does this bet­ter than any other co-liv­ing situ­a­tion I’ve been a part of.

A Growth Culture

One of the big pushes the EA ho­tel has made in the last few months is to foster a cul­ture of growth. There are weekly talks by mem­bers of the ho­tel that are highly at­tended. There are weekly op­por­tu­ni­ties for de­bug­ging bot­tle­necks in your life, and learn­ing new skills to make that de­bug­ging more effec­tive. A siz­able por­tion of the ho­tel at­tends the lo­cal gym, and there is some­one to go with al­most ev­ery day of the week, at var­i­ous times that suit you.

As im­por­tant as these in­di­vi­d­ual ac­tivi­ties are, the most im­por­tant thing is the cul­ture that de­vel­ops around them. Growth is ac­cepted and ex­pected at the EA ho­tel, and that’s im­por­tant for peo­ple cre­at­ing new pro­jects and learn­ing the skills as they go.

A Sup­port Culture

Another big shift I’ve seen the EA ho­tel make in the past few months is to­wards a cul­ture of sup­port. In con­crete terms, you can see this in the siz­able pop­u­la­tion of peo­ple who par­ti­ci­pate in morn­ing and nightly hugs, greet­ing each other with a strong hug the first and last time they see each other ev­ery day. It’s also strongly visi­ble in the ex­is­tence of a Ho­tel Guest Rep­re­sen­ta­tive, whose main job is to listen when ho­tel res­i­dents are hav­ing a hard time and look out for their in­ter­ests. It’s visi­ble in the nightly group din­ner, and the easy dis­cus­sion that usu­ally ac­com­pa­nies it. How­ever, it’s more visi­ble in the day-to-day in­ter­ac­tions you have with guests, such as when some­one offered to bring food up to my room when I was sick, or see­ing the cel­e­bra­tion when a guest got their pa­per pub­lished in a jour­nal.

Creat­ing a pro­ject from scratch is hard, first time startup founders of­ten find them­selves fal­ling into de­pres­sion and loneli­ness, failing sim­ply be­cause they don’t have the sup­port to take on the de­mands of the job. The sup­port cul­ture of the EA Ho­tel goes a long way to­wards mak­ing it more bear­able.

Con­sis­tent Improvement

A fi­nal thing that has im­pressed me about the EA ho­tel is the abil­ity of Greg and the trustees to take feed­back and im­prove the con­cept over time. I’ve already men­tioned the changes I’ve seen in the past few months, but an even bet­ter sign to me is the way Greg listens to crit­i­cism and re­sponds to feed­back. When­ever he hears a good idea, it’s im­me­di­ately writ­ten down, and the best ideas are tested and im­ple­mented over time. This gives me cause to be­lieve that the EA ho­tel hasn’t just lucked onto the above as­pects of its cul­ture, but is likely to con­tinue to de­velop into an even more effec­tive or­ga­ni­za­tion over time.


I’ve made three ma­jor claims in this post. First, I’ve made the claim that EA as a move­ment could be do­ing a lot more good if it filled the Chasm in its pipeline. Se­cond, I’ve ar­gued that some­thing like the EA Ho­tel is a good way to fill this Chasm. Fi­nally, I’ve ar­gued that the EA Ho­tel has func­tioned well as an or­ga­ni­za­tion ded­i­cated to filling this Chasm, and that there’s rea­son to be­lieve it will con­tinue to do well in the fu­ture. While I don’t think it’s im­pos­si­ble the EA Ho­tel could fail at this goal, my in­side view gives me con­fi­dence that it’s very likely to suc­ceed. How­ever, even with the out­side view, I be­lieve the mod­els given in this post make the case that the EA Ho­tel, if suc­cess­ful, would be highly use­ful in ex­pec­ta­tion, and make a strong cost-effec­tive­ness case for an EA Ho­tel like pro­ject if you think those mod­els are ac­cu­rate. I look for­ward to your feed­back and com­ments.

About Me

I’m Matt Gold­en­berg. I’ve been liv­ing at the EA Ho­tel for about three months, and I’m due to leave in an­other three. I’ve been a res­i­dent at a num­ber of EA and ra­tio­nal­ity group houses in the Bay Area, in­clud­ing a brief stay at Event Hori­zon, a stint at Milvia House, and as a cofounder of Gen­tle Mesa. I’ve pre­vi­ously run a num­ber of small busi­nesses, and one startup. While at the EA ho­tel, I’ve been work­ing on my main pro­ject Pro­ject Metis, as well as writ­ing a num­ber of ar­ti­cles on the side for LessWrong.