Oct 10, 2018
Extension of Give praise.
This is part analysis, part a heartfelt story of my engagement with the LW/EA community. I usually don't like to kick up dust, but I've decided to write an honest representation of my feelings, here and there possibly sacrificing accuracy. Adjust your interpretation accordingly.
Starting out and reading about LW and EA, I was psyched to jump in with the movement and get stuff to happen. There was a romantic vista of joining the ranks of *reasonable* people, that *would* understand my unconventional ideas, and thus actually *get real shit done*. The prospect was liberating.
I subscribe to the idea of human needs, in the sense that our utility function is an addition of sigmoids that project amounts of resources to allocate to the fulfillment of each need. Each sigmoid would have a different offset, meaning that we only start caring about the second need when the first one is mostly satisfied. Happiness would be a function of satisfaction, with 0 happiness corresponding to a state where a special subset of "deficiency" needs are satisfied, and nothing else. In that sense, I think Maslow's hierarchy was mostly right. Not in the sense of the specific needs he proposed, or their order, but in the sense of the underlying logic.
One of these deficiency needs is status, which I define as having some level of influence on social reality. I hold this need to approach fulfillment as it approaches a level of influence that is the same as the highest status person around. Complete fulfillment happens if you're equal or higher status than everyone else around you. As a corollary, everyone is fulfilled if and only if everyone is equal.
Additionally, these needs are all represented by subagents that are 'activated' (i.e. run a process in the background) if the need is not fulfilled. We have limited processing resources, and so a lack of status lowers one's IQ.
EA is NOT status balanced, and it's been eating me up. I feel threatened, have burned myself out multiple times, and expect this to be the major source of the mental health epidemic that is plaguing us.
I've worked with volunteers. No problem from an impact perspective. We all want to save the world, right?
Of the approximately 50 people that signed up, only 2 have stayed around until now, and only about 4 stayed around for longer than a week. Why are people so flaky?
Well let's say you find yourself at the bottom of this hierarchy, but at least these people are reasonable so you're willing to jump on the bandwagon. You look around, join some meetups, sign up to some volunteering jobs, but none of it is sufficient to be taken seriously, so you keep trying new things.
I call these low status convulsions. You find yourself in the darkness trying to fix yourself, trying to level up, so that one day you might impress people enough to be listened to, to be seen, talked about. To be given a say in the makeup of your social environment, up there in the green with Yudkowsky and Hanson and Soares and whatnot.
So you find yourself in this volunteering opportunity with some EA's and they tell you some stuff you can do, and you do it, and you're left in the dark again. Is this going to steer you into safe waters? Should you do more? Impress more? Maybe spend more time on that Master's degree to get grades that set you apart, maybe that'll get you invited with the cool kids? Maybe write some cool posts on LessWrong in the hopes of getting lots of upvotes? Let's just do them all at once because there doesn't seem to be any other way out...
...But then reality hits you and you find yourself overburdened, so you flake on some of your bandaid promises to make some room for the low-risk low-hope stuff, and suck up your tears because no one cares anyway. They're all way too busy with their own scrambles for recognition.
It's fucking grim.
I'd be a bad rationalist if I didn't examine the other side of the story.
Some people are really more intelligent than others, and the differences are even larger in the tail of the distribution. From a meritocratic perspective, a small portion of the community should really make most of the decisions.
There is a consensus among some EA institutions that the expected value of a random EA project is approximately 0, because there are as many that might harm, as there are projects that help.
I can't judge whether this is true, for I haven't examined the thinking behind it, but can you believe that the mean member of a movement of hugely talented people which identifies with *measuring* impact, can't do better than chance? I'm going to go with the charitable assumption that this is true. Can you believe how hard it is to do the right thing, and how bad things would be if we didn't have a hierarchy?
In short, tearing down the status hierarchy might make some people happy, but it interferes with impact. The thing we're doing it for in the first place.
There is a way out.
Hope is an anticipation that things will work out if you do X, plus an anticipation that X is doable, plus a resolve to do it. It effectively shuts up the background process that is activated by the anticipation of a thing not working out, much like when the thing actually works out.
For example, being 5 goals ahead in a football match feels the same as having already won. Seeing an oasis in the desert feels the same as already drinking the water. Learning of a positive weather prediction may curb your doubt of going on that holiday. The relief happens not when the need is satisfied, but when the need is going to be satisfied beyond a reasonable doubt. That's when the suffering ends.
As it stands, there are no guarantees for most people that they will ever reach the level of status they need in EA to be relieved from their suffering and the foggy cognition that comes with it.
Making a promise helps a little, for a short while. That's why we're flaky. Signing up as a convulsion.
This is contrast with many other places. Companies have clear hierarchies and performance ratings. Healthy social groups have mechanisms for keeping people at roughly the same level of status. Academia has tenure tracks (and those that aren't on it are struggling). Even Buddhist temples, having been optimized for promoting mental health for millennia (and in my experience being way more effective than the fledgling western tradition of psychotherapy), have a clear explicit hierarchy.
With heroic effort I'm no longer suffering as much from an uncertain future, but my former self would have instantly be imbued with the hope, motivation, optimism that he came for in the first place, had he been put on a predictable path to full personhood right at the start. It would still be a relief right now. Halfway isn't as good as all the way, and I have barely any indications that anything in the crushing pile of work I'm doing is any good.
On the other side of the coin, my former self would have been happy to move on if it was clear that he couldn't possibly climb the hierarchy even if he tried. The struggle is in the uncertainty.
So what I want to propose is that we define much more clearly what it takes to be taken seriously around here.
One way is to define a precise membership test that successfully puts one in a narrative of inclusion. This is what I was trying to convey in The league of rationalists. That's what initiation rituals are for.
Another way might be to simply carve out more slots. There is not one status hierarchy, but an intertwined set of hierarchies with at the top positions that guarantee membership. Membership is guaranteed when you have a convincing narrative that you are a necessary part of the group. When you have leverage.
This might be done by specializing in a skill that makes you a worthy asset. Who is the best EA software developer? Who is the best ops person?
Another way is to properly define what we value, and affirm when someone has indeed reached a point of being respectable. For example: shoutout to those people that have sticked with volunteering for RAISE after committing to it, and showing proper consideration when they couldn't continue. That's Veerle de Goederen, Remmelt Ellen, Lewis Hammond, Roland Pihlakas and myself. Well done.
Edit: Two more excellent ideas from a comment by ricraz:
there's another sense in which "we can all be high-status": within our respective local communities. I'm curious how you feel about that, because that was quite adequate for me for a long time, especially as a student.
On a broader level, one actionable idea I've been thinking about is to talk less about existential risk being "talent constrained", so that people who can't get full-time jobs in the field don't feel like they're not talented. A more accurate term in my eyes is "field-building constrained".
Whatever we do, something needs to be done. Right now a large portion of the movement lives with an underdefined identity, a lack of status, feeling like they have to work so much harder to be seen, and it's never enough. It's not just hurting them, it's making them unproductive and prone to corruption. How well do you think if you're thirsty? That's the reality for most of us.