Typical sphere of responsibility, yes, but I don’t think offloading most of those to the EV Ops team diminishes independence, in terms of each org head deciding what their org does?
It’s also not that clear what the divisions look like: for example, I’m pretty sure the orgs make their own decisions on who to hire, so my guess is the “hiring” under EV Ops is more the mechanics of implementing hiring decisions?
I think that offloading these sorts of things does meaningfully diminish independence, even if it doesn’t entirely abrogate it. IME the kinds of decisions & policies included in the list Larks quotes have a quite profound effect on the character of an organisation.
I would consider all of these things to be ultimately the responsibility of the CEO in an ordinary organisation.
I think these are typically ultimately the responsibility of the sponsee org CEOs. I see them as generally entering this agreement, in a way where they are effectively deciding to use the sponsor for operations support.
Sponsees typically have the option of leaving the sponsored org, and I think that makes a big difference.
It’s similar to an org that’s outsourced much of it’s product to a white-label organization.
All that said, there’s definitely a spectrum here. It’s definitely possible that a big org could control and manage sub-orgs within a sponsee/sponsored relationship. I think much of the devil is in the details.
I think it would be useful for people at CEA to help clarify this. Who ultimately is in charge of each project? I assume it’s typically that project’s CEO, but perhaps not in all cases.
A related question is who is in charge of firing the CEO. If it is the case that the CEF board is the one responsible, then it could be seen as much more of the CEF’s direct responsibility.
That’s a serious red flag to me. I feel like EA orgs should be seeking to have less coupling rather than more. The fact that five people—one of whom has seemingly no interaction with the community, and who we know almost nothing about - are the de facto rulers of almost the whole EA movement seems extremely unhealthy for its future.
Interestingly, I have roughly the opposite concern. I’m worried that the current structure of EVF results in too little accountability for individual projects under EV’s umbrella.
Formally, the board is accountable for the actions of an org it “runs”. But the great majority of actual decisions are made by the executive officers, especially the CEO/ED. In the case when the board is responsible for a dozen or more autonomous suborgs, their ultimate responsibility/accountability seems likely to be even more nominal than usual.
I expect that, in practice, EVF suborg leaders have significantly less oversight & accountability than they would if they all had separate boards. (I also have concerns about how this structure might reduce accountability to the community/public, but these are currently too ill-formed to express very clearly.)
In practice though, your concerns seem to point to the same conclusion, don’t they? That different sub-orgs should be separate from each other and have mostly-disjoint boards?
I’m not sure how much we should be relying on the legal institution of the nonprofit board here, since these are often pretty toothless. (Honestly, CEA’s board has shown more teeth than many ever do, though they were still very slow to act in this case IMO.)
But I’m also not sure what the alternative is.
(My comment on this other thread is also somewhat relevant here.)
If the organisations are effectively run individually, then they answer to their funders.
If they all get funding directed from the same group who just de facto fund whatever falls under the EV umbrella, then we’re back to the EV trustees having all the power.
If they get substantial funding from we, the movement, that seems healthier, but only if there’s some kind of meaningful competition. That only seems possible if they limit their scope to allow for meaningful comparisons between orgs—which I argued for here, though didn’t get much engagement :\
I’m not actually sure these are different concerns—it depends how the logistics work. If the rulership method of the EV trustees is ‘appoint someone, then direct funding to them’, then that’s still a lot of influence. If it’s ‘appoint someone, then let them find their own funding’, it’s quite a lot less so (though those waters could be muddied by funders being implicitly influenced to give funding to anyone under the EV umbrella).
It’s my loose impression that different EVF orgs do need to seek their own funding, at least for stuff that doesn’t fall under EV Ops’ purview (though the latter is not inconsiderable!). I’m not sure I’ve ever seen this laid out explicitly, though, and agree that would help.
though those waters could be muddied by funders being implicitly influenced to give funding to anyone under the EV umbrella
This comment seems crazy to me. These people have legal accountability for the organization EVF. They are not “de facto rules of almost the whole EA movement”. Can you actually propose a scenario in which these people exert much if any influence, apart from getting to fire the leadership of EVF, which is at most a modestly big deal anyway?
They also get to hire the leadership of EVF. Those people then work in the knowledge that they can theoretically be fired by the trustees, giving the trustees all the soft power that that implies. They may not currently wield it, but wielding and having power are two different things. Similarly, it may not be a net positive to their lives (I suspect it isn’t for most of them), but someone not benefiting from power doesn’t mean they don’t have it.
Asserting it’s ‘at most a modestly big deal’ in such an incredulous tone seems counterproductive when that’s a large part of the point in discussion.
I think it’s a bit weird to call this “soft” power? It’s quite explicit power: the trustees/boards have final responsibility for their organizations, including all the projects under them.
That seems plausible, but raises the question of why someone who needs to maintain such a high level of privacy is a suitable leader of a movement which should be as committed to transparency as possible.
Even among EAs I’m atypically pro-transparency, but while I would worry if there was little public information or community engagement from most of the board members that’s far from the case. If you overly constrain the board members you’re willing to have you can miss out on people with skills or experience that would be really valuable for the organization.
I’m not suggesting she do a five-page interview in Time Magazine—just some kind of engagement with the community she’s somehow, virtually without its knowledge, become custodian of. Like, an occasional forum post, monthly office hours, or similar.
Part of the problem with the current structure is precisely that no-one gets to ask the question ‘how qualified are the Watchmen?’ - we just have to take it on the trust of the other four, which compounds the problem. By comparison, how would you feel if two, three, or four of the remaining five were replaced by someone equally as mysterious? I think each unknown makes the concerns about concentration of influence and lack of accountability look proportionately worse.
BTW I agree with the sentiment that the EA movement would benefit from knowing a little more about McCauley. Even just some very basic questions like:
What cause areas are most important to you?
What do you see as the role of EVF?
How did you get into EA? How have your views evolved over time?
I think it’s a relatively small issue, but not knowing much about the EVF board seems like it somewhat reduces the abilities of EAs to properly calibrate their views on how much to trust/rely on EVF and its subsidiaries.
I will also say that it doesn’t feel super representative of the EA community that all 5 of EVF’s board members appear to primarily be longtermists. This isn’t necessarily an issue if all the board members are doing a good job of listening to and representing the views of the EA community, rather than just their own. But it’s sort of odd.
According to the website, EV provides these services:
I would consider all of these things to be ultimately the responsibility of the CEO in an ordinary organisation.
Typical sphere of responsibility, yes, but I don’t think offloading most of those to the EV Ops team diminishes independence, in terms of each org head deciding what their org does?
It’s also not that clear what the divisions look like: for example, I’m pretty sure the orgs make their own decisions on who to hire, so my guess is the “hiring” under EV Ops is more the mechanics of implementing hiring decisions?
I think that offloading these sorts of things does meaningfully diminish independence, even if it doesn’t entirely abrogate it. IME the kinds of decisions & policies included in the list Larks quotes have a quite profound effect on the character of an organisation.
I think these are typically ultimately the responsibility of the sponsee org CEOs. I see them as generally entering this agreement, in a way where they are effectively deciding to use the sponsor for operations support.
Sponsees typically have the option of leaving the sponsored org, and I think that makes a big difference.
It’s similar to an org that’s outsourced much of it’s product to a white-label organization.
All that said, there’s definitely a spectrum here. It’s definitely possible that a big org could control and manage sub-orgs within a sponsee/sponsored relationship. I think much of the devil is in the details.
I think it would be useful for people at CEA to help clarify this. Who ultimately is in charge of each project? I assume it’s typically that project’s CEO, but perhaps not in all cases.
A related question is who is in charge of firing the CEO. If it is the case that the CEF board is the one responsible, then it could be seen as much more of the CEF’s direct responsibility.
On some more thinking about it, I’m actually curious who has authority to remove/replace the CEOs. I think that’s the key question to me.
The EVF trustees, no? Ex: https://www.centreforeffectivealtruism.org/blog/announcing-a-change-of-leadership-at-cea
That’s a serious red flag to me. I feel like EA orgs should be seeking to have less coupling rather than more. The fact that five people—one of whom has seemingly no interaction with the community, and who we know almost nothing about - are the de facto rulers of almost the whole EA movement seems extremely unhealthy for its future.
Interestingly, I have roughly the opposite concern. I’m worried that the current structure of EVF results in too little accountability for individual projects under EV’s umbrella.
Formally, the board is accountable for the actions of an org it “runs”. But the great majority of actual decisions are made by the executive officers, especially the CEO/ED. In the case when the board is responsible for a dozen or more autonomous suborgs, their ultimate responsibility/accountability seems likely to be even more nominal than usual.
I expect that, in practice, EVF suborg leaders have significantly less oversight & accountability than they would if they all had separate boards. (I also have concerns about how this structure might reduce accountability to the community/public, but these are currently too ill-formed to express very clearly.)
In practice though, your concerns seem to point to the same conclusion, don’t they? That different sub-orgs should be separate from each other and have mostly-disjoint boards?
Plausibly.
I’m not sure how much we should be relying on the legal institution of the nonprofit board here, since these are often pretty toothless. (Honestly, CEA’s board has shown more teeth than many ever do, though they were still very slow to act in this case IMO.)
But I’m also not sure what the alternative is.
(My comment on this other thread is also somewhat relevant here.)
If the organisations are effectively run individually, then they answer to their funders.
If they all get funding directed from the same group who just de facto fund whatever falls under the EV umbrella, then we’re back to the EV trustees having all the power.
If they get substantial funding from we, the movement, that seems healthier, but only if there’s some kind of meaningful competition. That only seems possible if they limit their scope to allow for meaningful comparisons between orgs—which I argued for here, though didn’t get much engagement :\
I’m not actually sure these are different concerns—it depends how the logistics work. If the rulership method of the EV trustees is ‘appoint someone, then direct funding to them’, then that’s still a lot of influence. If it’s ‘appoint someone, then let them find their own funding’, it’s quite a lot less so (though those waters could be muddied by funders being implicitly influenced to give funding to anyone under the EV umbrella).
It’s my loose impression that different EVF orgs do need to seek their own funding, at least for stuff that doesn’t fall under EV Ops’ purview (though the latter is not inconsiderable!). I’m not sure I’ve ever seen this laid out explicitly, though, and agree that would help.
Seems right to me, at least directionally.
This comment seems crazy to me. These people have legal accountability for the organization EVF. They are not “de facto rules of almost the whole EA movement”. Can you actually propose a scenario in which these people exert much if any influence, apart from getting to fire the leadership of EVF, which is at most a modestly big deal anyway?
They also get to hire the leadership of EVF. Those people then work in the knowledge that they can theoretically be fired by the trustees, giving the trustees all the soft power that that implies. They may not currently wield it, but wielding and having power are two different things. Similarly, it may not be a net positive to their lives (I suspect it isn’t for most of them), but someone not benefiting from power doesn’t mean they don’t have it.
Asserting it’s ‘at most a modestly big deal’ in such an incredulous tone seems counterproductive when that’s a large part of the point in discussion.
I think it’s a bit weird to call this “soft” power? It’s quite explicit power: the trustees/boards have final responsibility for their organizations, including all the projects under them.
‘Soft influence’ then, if they have a hands-off approach? Either way, it seems concerning.
I’m assuming you’re talking about Tasha McCauley? I’m also confused by this: I had thought this was CEA’s former (2017-2018) CEO with a similar name.
Looks like she’s CEO of GeoSim and an OpenAI board member. I also found a recording of a panel appearance at EAG 2017.
Yeah. I got that far, but her otherwise total nonengagement with/visibility to a community she has so much control over seems worrying, no?
I wonder if the low visibility is that she’s married to a celebrity and so has pretty different privacy tradeoffs than most of us?
That seems plausible, but raises the question of why someone who needs to maintain such a high level of privacy is a suitable leader of a movement which should be as committed to transparency as possible.
Even among EAs I’m atypically pro-transparency, but while I would worry if there was little public information or community engagement from most of the board members that’s far from the case. If you overly constrain the board members you’re willing to have you can miss out on people with skills or experience that would be really valuable for the organization.
I’m not suggesting she do a five-page interview in Time Magazine—just some kind of engagement with the community she’s somehow, virtually without its knowledge, become custodian of. Like, an occasional forum post, monthly office hours, or similar.
Part of the problem with the current structure is precisely that no-one gets to ask the question ‘how qualified are the Watchmen?’ - we just have to take it on the trust of the other four, which compounds the problem. By comparison, how would you feel if two, three, or four of the remaining five were replaced by someone equally as mysterious? I think each unknown makes the concerns about concentration of influence and lack of accountability look proportionately worse.
BTW I agree with the sentiment that the EA movement would benefit from knowing a little more about McCauley. Even just some very basic questions like:
What cause areas are most important to you?
What do you see as the role of EVF?
How did you get into EA? How have your views evolved over time?
I think it’s a relatively small issue, but not knowing much about the EVF board seems like it somewhat reduces the abilities of EAs to properly calibrate their views on how much to trust/rely on EVF and its subsidiaries.
I will also say that it doesn’t feel super representative of the EA community that all 5 of EVF’s board members appear to primarily be longtermists. This isn’t necessarily an issue if all the board members are doing a good job of listening to and representing the views of the EA community, rather than just their own. But it’s sort of odd.