I think a name change might be good, but am not very excited about the “Global Priorities” name. I expect it would attract mostly people interested in seeking power and “having lots of influence” and I would generally expect a community with that name to be very focused on achieving political aims, which I think would be quite catastrophic for the community.
I actually considered this specific name in 2015 while I was working at CEA US as a potential alternative name for the community, but we decided against it at the time for reasons in this space (and because changing names seems hard).
While I’m not sure we’re using terms like “political” and “power” in the same way, as far as I can tell this worry makes a lot of sense to me.
However, I think there is an opposite failure mode: mistakenly believing that because of one’s noble goals and attitudes one is immune to the vices of power, and can safely ignore the art of how to navigate a world that contains conflicting interests.
A key assumption from my perspective is that political and power dynamics aren’t something one can just opt out of. There is a reason why thinkers from Plato over Macchiavelli to Carl Schmitt have insisted that politics is a separate domain that merits special attention (and I’m not saying this as someone who is not particularly sympathetic to any of these three on the object level). [ETA: Actually I’m not sure if Plato says that, and I’m confused why I included him originally. In a sense he may suggest the opposite view since he sometimes compares the state to the individual.]
Internally, community members with influence over more financial or social capital have power over those whose projects depend on such capital. There certainly are different views with respect to how this capital is best allocated, and at least for practical purposes I don’t think these are purely empirical disagreements and instead involve ‘brute differences in interests’.
Externally, EAs have power over beneficiaries when they choose to help some but not others. And a lot of EA projects are relevant to the interests of EA-external actors that form a complex network of partly different and partly aligned interests and different amounts of power over each other. Perhaps most drastically, a lot of EA thought around AI risk is about how to best influence how essentially the whole world will be reshaped (if not an outright plan for how to essentially take over the world).
Therefore, I think we will need to deal with ‘politics’ anyway, and we will attract people who are motivated by seeking power anyway. Non-EA political structures and practice contain a lot of accumulated wisdom on how to navigate conflicting interests while limiting damage from negative-sum interactions, on how to keep the power of individual actors in check, and on how to shape incentives in such a way that power-seeking individuals make prosocial contributions in their pursuit of power. (E.g. my prior is that any head of government in a democracy is at least partly motivated by pursuing power.)
To be clear, I think there are significant problems with these non-EA practices. (Perhaps most notably negative x-risk externalities from international competition.) And if EA can contribute technological or other innovations that help with reducing these problems, I’m all for it.
Yet overall I feel like I more often see EAs make the mistake of naively thinking they can ignore their externally imposed entanglement in political and power dynamics, and that there is nothing to be learned from established ways for how to reign in and shape these dynamics (perhaps because they view established practice and institutions largely as a morass of corruption and incompetence one better steers clear of). E.g. some significant problems I’ve seen at EA orgs could have been avoided by sticking more closely to standard advice of having e.g. a functional board that provides accountability to org leadership.
My best guess is that, on the margin, it would be good to attract more people with a more common-sense perspective on politics and power-seeking as opposed to people who lack the ability or willingness to understand how power operates in the world, and how to best navigate this. If rebranding to “Global Priorites” would have that effect (which I think I’m less confident in than you), then I’d count that as a reason for rebranding (though I doubt it would be among the top 5 most important pro or con reasons).
I’m noticing I don’t fully understand the way in which you think “Global Priorities” would attract power-seekers, or what you mean by that. Like, I have a vague sense that you’re probably right, but I don’t see the direct connection yet. Would be very interested in more elaboration on this.
I mean, I just imagine what kind of person would be interested, and it would mostly be the kind of person who is ambitious, though not necessarily competent, and would seek out whatever opportunities or clubs there are that are associated with the biggest influence over the world, or sound the highest status, have the most prestige, or sound like would be filled with the most powerful people. I have met many of those people, and a large fraction of high-status opportunities that don’t also strongly select for merit seem filled with them.
Currently both EA and Rationality are weird in a way that is not immediately interesting to people who follow that algorithm, which strikes me as quite good. In universities, when I’ve gone to things that sounded like “Global Priorities” seminars, I mostly met lots of people with a political science degree, or MBA’s, really focusing on how they can acquire more power and the whole conversation being very status oriented.
The Defense Professor’s fingers idly spun the button, turning it over and over. “Then again, only a very few folk ever do anything interesting with their lives. What does it matter to you if they are mostly witches or mostly wizards, so long as you are not among them? And I suspect you will not be among them, Miss Davis; for although you are ambitious, you have no ambition.”
“That’s not true!” said Tracey indignantly. “And what’s it mean?”
Professor Quirrell straightened from where he had been leaning against the wall. “You were Sorted into Slytherin, Miss Davis, and I expect that you will grasp at any opportunity for advancement which falls into your hands. But there is no great ambition that you are driven to accomplish, and you will not make your opportunities. At best you will grasp your way upward into Minister of Magic, or some other high position of unimportance, never breaking the bounds of your existence.”
—HPMOR, Chapter 70, Self-Actualization (part 5)
Added: The following is DEFINITELY NOT a strong argument, but just kind of an associative point. I think that Voldemort (both the real one from JK Rowling and also the one in HPMOR) would be much more likely to decide that he and his Death Eaters should have “Global Priorities” meetings than “Effective Altruist” meetings. (“We’re too focus on taking over the British Ministry for Magic, we need to also focus on our Global Priorities.“) In that way I think the former phrase has a more general connotation of ”taking power and changing the world” in a way the latter does not.
I think this is a good point. That said, I imagine it’s quite hard to really tell.
Empirical data could be really useful to get here. Online experimentation in simple cases, or maybe we could even have some University chapters try out different names and see if we can infer any substantial differences.
1) I’m convinced that a “GP” community would attract somewhat more power-seeking people. But they might be more likely to follow (good) social norms than the current consequentialist crowd. Moreover, we would be heading toward a more action-oriented and less communal group, which could reduce the attraction to manipulative people. And today’s community is older and more BS-resistant with some legibly-trustworthy leaders. But you seem to think there would be a big and harmful net effect—can you explain?
2) assuming that “GP” is too intrinsically political, can you think of any alternatives that have some of its advantages of “GP” without that disadvantage?
we would be heading toward a more action-oriented and less communal group, which could reduce the attraction to manipulative people
I don’t expect a brand change to “Global Priorities” to bring in more action-oriented people. I expect fewer people would donate money themselves, for instance, they would see it as cute but obviously not having any “global” impact, and therefore below them.
(I think it was my inner Quirrell / inner cynic that wrote some of this comment, but I stand by it as honestly describing a real effect that I anticipate.)
But we would also be heading toward a more action-oriented and less communal group, which could reduce the attraction to manipulative people
I don’t understand this. We would be trending towards seeking more power, which would further attract power-seekers. We have already substantially gone down this path. You might have different models of what attracts manipulative people. My model is doing visibly power-seeking and high-status work is one of the most common attractors.
Moreover, today’s community is older and more BS-resident with some legibly-trustworthy leaders.
I think we have overall become substantially less BS-resistant as we have grown and have drastically increased the surface area of the community, though it depends a bit on the details.
But you seem to think there would be a big and harmful net effect—can you explain?
Yep, I would be up for doing that, but alas won’t have time for it this week. It seemed better to leave a comment voicing my concerns at all, even if I don’t have time to explain them in-depth, though I do apologize for not having the time to explain them in full.
I don’t understand this. We would be trending towards seeking more power, which would further attract power-seekers. We have already substantially gone down this path. You might have different models of what attracts manipulative people. My model is doing visibly power-seeking and high-status work is one of the most common attractors.
I’m concerned about people seeking power in order to mistreat, mislead, or manipulate others (cult-like stuff), as seems more likely in a social community, and less likely in a group of people who share interests in actually doing things in the world. I’m in favour of people gaining influence, all things equal!
Alas, I think that isn’t actually what tends to attract the most competent manipulative people. Random social communities might attract incompetent or average-competence manipulative people, but those are much less of a risk than the competent ones. In general, professional communities, in particular ones aiming for relatively unconditional power, strike me as having a much higher density of manipulative people than random social communities.
I also think when I go into my models here, the term “manipulative” feels somewhat misleading, but it would take me a while longer to explain alternative phrasings.
TBC, this feels like a bit of a straw man of my actual view, which is that power and communality jointly contribute to risks of cultishness and manipulativeness.
I think a name change might be good, but am not very excited about the “Global Priorities” name. I expect it would attract mostly people interested in seeking power and “having lots of influence” and I would generally expect a community with that name to be very focused on achieving political aims, which I think would be quite catastrophic for the community.
I actually considered this specific name in 2015 while I was working at CEA US as a potential alternative name for the community, but we decided against it at the time for reasons in this space (and because changing names seems hard).
While I’m not sure we’re using terms like “political” and “power” in the same way, as far as I can tell this worry makes a lot of sense to me.
However, I think there is an opposite failure mode: mistakenly believing that because of one’s noble goals and attitudes one is immune to the vices of power, and can safely ignore the art of how to navigate a world that contains conflicting interests.
A key assumption from my perspective is that political and power dynamics aren’t something one can just opt out of. There is a reason why thinkers from Plato over Macchiavelli to Carl Schmitt have insisted that politics is a separate domain that merits special attention (and I’m not saying this as someone who is not particularly sympathetic to any of these three on the object level). [ETA: Actually I’m not sure if Plato says that, and I’m confused why I included him originally. In a sense he may suggest the opposite view since he sometimes compares the state to the individual.]
Internally, community members with influence over more financial or social capital have power over those whose projects depend on such capital. There certainly are different views with respect to how this capital is best allocated, and at least for practical purposes I don’t think these are purely empirical disagreements and instead involve ‘brute differences in interests’.
Externally, EAs have power over beneficiaries when they choose to help some but not others. And a lot of EA projects are relevant to the interests of EA-external actors that form a complex network of partly different and partly aligned interests and different amounts of power over each other. Perhaps most drastically, a lot of EA thought around AI risk is about how to best influence how essentially the whole world will be reshaped (if not an outright plan for how to essentially take over the world).
Therefore, I think we will need to deal with ‘politics’ anyway, and we will attract people who are motivated by seeking power anyway. Non-EA political structures and practice contain a lot of accumulated wisdom on how to navigate conflicting interests while limiting damage from negative-sum interactions, on how to keep the power of individual actors in check, and on how to shape incentives in such a way that power-seeking individuals make prosocial contributions in their pursuit of power. (E.g. my prior is that any head of government in a democracy is at least partly motivated by pursuing power.)
To be clear, I think there are significant problems with these non-EA practices. (Perhaps most notably negative x-risk externalities from international competition.) And if EA can contribute technological or other innovations that help with reducing these problems, I’m all for it.
Yet overall I feel like I more often see EAs make the mistake of naively thinking they can ignore their externally imposed entanglement in political and power dynamics, and that there is nothing to be learned from established ways for how to reign in and shape these dynamics (perhaps because they view established practice and institutions largely as a morass of corruption and incompetence one better steers clear of). E.g. some significant problems I’ve seen at EA orgs could have been avoided by sticking more closely to standard advice of having e.g. a functional board that provides accountability to org leadership.
My best guess is that, on the margin, it would be good to attract more people with a more common-sense perspective on politics and power-seeking as opposed to people who lack the ability or willingness to understand how power operates in the world, and how to best navigate this. If rebranding to “Global Priorites” would have that effect (which I think I’m less confident in than you), then I’d count that as a reason for rebranding (though I doubt it would be among the top 5 most important pro or con reasons).
I agree that changing names is hard and costly (you can’t do it often), something that definitely should be taken into account.
I’m noticing I don’t fully understand the way in which you think “Global Priorities” would attract power-seekers, or what you mean by that. Like, I have a vague sense that you’re probably right, but I don’t see the direct connection yet. Would be very interested in more elaboration on this.
I mean, I just imagine what kind of person would be interested, and it would mostly be the kind of person who is ambitious, though not necessarily competent, and would seek out whatever opportunities or clubs there are that are associated with the biggest influence over the world, or sound the highest status, have the most prestige, or sound like would be filled with the most powerful people. I have met many of those people, and a large fraction of high-status opportunities that don’t also strongly select for merit seem filled with them.
Currently both EA and Rationality are weird in a way that is not immediately interesting to people who follow that algorithm, which strikes me as quite good. In universities, when I’ve gone to things that sounded like “Global Priorities” seminars, I mostly met lots of people with a political science degree, or MBA’s, really focusing on how they can acquire more power and the whole conversation being very status oriented.
Thanks, I find that helpful, and agree that’s a dangerous dynamic, and could be exacerbated by such a name change.
—HPMOR, Chapter 70, Self-Actualization (part 5)
Added: The following is DEFINITELY NOT a strong argument, but just kind of an associative point. I think that Voldemort (both the real one from JK Rowling and also the one in HPMOR) would be much more likely to decide that he and his Death Eaters should have “Global Priorities” meetings than “Effective Altruist” meetings. (“We’re too focus on taking over the British Ministry for Magic, we need to also focus on our Global Priorities.“) In that way I think the former phrase has a more general connotation of ”taking power and changing the world” in a way the latter does not.
I think this is a good point. That said, I imagine it’s quite hard to really tell.
Empirical data could be really useful to get here. Online experimentation in simple cases, or maybe we could even have some University chapters try out different names and see if we can infer any substantial differences.
Interesting.
1) I’m convinced that a “GP” community would attract somewhat more power-seeking people. But they might be more likely to follow (good) social norms than the current consequentialist crowd. Moreover, we would be heading toward a more action-oriented and less communal group, which could reduce the attraction to manipulative people. And today’s community is older and more BS-resistant with some legibly-trustworthy leaders. But you seem to think there would be a big and harmful net effect—can you explain?
2) assuming that “GP” is too intrinsically political, can you think of any alternatives that have some of its advantages of “GP” without that disadvantage?
I don’t expect a brand change to “Global Priorities” to bring in more action-oriented people. I expect fewer people would donate money themselves, for instance, they would see it as cute but obviously not having any “global” impact, and therefore below them.
(I think it was my inner Quirrell / inner cynic that wrote some of this comment, but I stand by it as honestly describing a real effect that I anticipate.)
I don’t understand this. We would be trending towards seeking more power, which would further attract power-seekers. We have already substantially gone down this path. You might have different models of what attracts manipulative people. My model is doing visibly power-seeking and high-status work is one of the most common attractors.
I think we have overall become substantially less BS-resistant as we have grown and have drastically increased the surface area of the community, though it depends a bit on the details.
Yep, I would be up for doing that, but alas won’t have time for it this week. It seemed better to leave a comment voicing my concerns at all, even if I don’t have time to explain them in-depth, though I do apologize for not having the time to explain them in full.
I’m concerned about people seeking power in order to mistreat, mislead, or manipulate others (cult-like stuff), as seems more likely in a social community, and less likely in a group of people who share interests in actually doing things in the world. I’m in favour of people gaining influence, all things equal!
Alas, I think that isn’t actually what tends to attract the most competent manipulative people. Random social communities might attract incompetent or average-competence manipulative people, but those are much less of a risk than the competent ones. In general, professional communities, in particular ones aiming for relatively unconditional power, strike me as having a much higher density of manipulative people than random social communities.
I also think when I go into my models here, the term “manipulative” feels somewhat misleading, but it would take me a while longer to explain alternative phrasings.
TBC, this feels like a bit of a straw man of my actual view, which is that power and communality jointly contribute to risks of cultishness and manipulativeness.
nods My concerns have very little to do with cultishness, so my guess is we are talking about very different concerns here.