EDIT: after reflecting on this comment I think I was too dismissive of the risk of association between EA and Democrats, particularly because I think we’re headed toward a period of Republican domination of US politics and the risk of being associated with Democrats may feasibly outweigh the reward of potential policy. Anyway, below are my original thoughts.
Interesting post. I lean toward disagreeing, for a couple of reasons.
I think you would agree that Congress can, if it adopts EA legislation, be greatly helpful to the EA cause. It just has way more money and influence than EA can dream of at the moment. The questions are then:
Is having a self-proclaimed EA in Congress helpful to getting legislation passed?
Will the potential negative press and association with Democrats be too harmful to the EA movement to be worth it?
On (1), I think the answer is a resounding YES, and you have to overthink really hard to reach a different conclusion. Congresspeople deal with a ton of different interest groups all the time trying to have their preferred policies make their way into legislation. We are competing with all of them, which makes our odds of success quite low, especially since politicians generally consult the interest groups they agree with rather than interest groups actually persuading politicians. Having a congressman tirelessly devoted to the singular cause of getting EA legislation enacted, on the other hand—that can be powerful. I think the Tea Party/Squad comparison is quite a bad one, given that these groups focus on hyper-partisan legislation, which as you say EA is not. A more apt comparison in my eyes is pork. Politicians get funding all the time for projects in their districts so they can report back to their constituents. Hundreds of these things get included in many omnibus bills in order to ensure the vote of every single legislator. If Carrick Flynn is unwilling to vote for legislation without AI safety funding, and Democrats in the House need his vote due to a narrow majority, that boosts our odds significantly (especially since AI safety is pretty non-partisan and unlikely to be a sticking point in the Senate). I think you focus too much on the short-term in this analysis—politicians stick around for a long time, and Flynn could very realistically have a lot of influence in future sessions. The upside here seems massive.
For question 2, first let me comment on negative press. I’m pretty skeptical. Flynn really didn’t advertise EA at all during his campaign, and his opponents did not attack him for it (aside from crypto due to its association with wealth/corruption), and for good reason. It’s really hard to get voters to care about esoteric ideas one way or the other. Voters at large hold pretty authoritarian values and don’t care about Republicans’ attacks on democracy. Very liberal whites are basically the only ones who care about climate change. And these are esoteric issues that are already quite political—the idea that AI safety or animal welfare would become a campaign point is in my view laughable. To add, if it becomes too much of an issue, we (as a movement) can always decide it’s not worth it and stop running candidates. Flynn was a nice trial run that showed us crypto is a weakness, and it’s worth having more tests.
Now onto association with Democrats. EA will always be a left-dominated organization due to the extreme left lean of highly educated people. I do share your concern about Republicans being unwilling to pass EA legislation if it’s associated with Democrats. But I think you vastly exaggerate how much more association EA would have with Democrats if there were a couple of Democratic EA-associated legislators in office, especially if they never really talk about EA in public. And besides, going back to question 1, I still think there is a (much) higher chance of getting legislation passed if we have an EA in Congress.
Your second question “Will the potential negative press and association with Democrats be too harmful to the EA movement to be worth it?” seems to ignore that a major group EAs will be running against will be democrats in primaries.
So it’s not only that you’re creating large incentives for republicans to attack EA, you’re also creating it for e.g. progressive democrats. See: Warren endorsing Flynn’s opponent & somewhat attacking flynn for crypto billionaire sellout stuff
That seems potentially pretty harmful too. It’d be much harder to be an active group on top universities if progressive groups strongly disliked EA.
Which I think they would, if EAs ran against progressives enough that Warren or Bernie or AOC more strongly criticized EA. Which would be in line the incentives we’re creating & general vibe [pretty skeptical of a bunch of white men, crypto billionaires, etc].
Congresspeople deal with a ton of different interest groups all the time trying to have their preferred policies make their way into legislation. We are competing with all of them, which makes our odds of success quite low, especially since politicians generally consult the interest groups they agree with rather than interest groups actually persuading politicians.
...
A more apt comparison in my eyes is pork. Politicians get funding all the time for projects in their districts so they can report back to their constituents. Hundreds of these things get included in many omnibus bills in order to ensure the vote of every single legislator.
I agree that pork is good comparison, since it has highly concentrated support but nonexistent opposition. But the second section I quoted above directly contradicts the first section: pork is an example of how interest groups can be effective at getting their preferred policies into legislation, even if no politician themself benefits from the pork.
It’s very rare for the groups that benefit from pork to try to steer it by themselves holding office.
I don’t think we are really competing with all other interest groups; there’s not a limited amount of legislation. There is a limited-ish budget, but it’s so high that there’s room for many different priorities already (including pork).
But I think you vastly exaggerate how much having more association EA would have with Democrats if there were a couple of Democratic EA-associated legislators in office, especially if they never really talk about EA in public.
Well, maybe. It’s really tough to evaluate hypotheticals like this; but the actually existing political campaign that I’m commenting on lead to multiple articles about EA in the national political press just from the primary, let alone the general or somebody holding office.
I don’t think this is an accurate view of pork. Pork is pushed for by legislators, not by interest groups. These projects have some support from within the district, sure, but it’s really the legislators that want them to happen so they can advertise to their constituents. Similarly EA would be much more likely to make its way into legislation if it were pushed for by a devoted legislator than by an outside interest group.
As for the articles in the press, I think Yglesias makes a pretty convincing case that these can do quite a bit of good as well; in my mind they’re probably net good, but I understand the concern.
Pork is pushed for by legislators, not by interest groups.
I don’t really understand the distinction you’re drawing. Interest groups definitely lobby to receive pork.
These projects have some support from within the district, sure, but it’s really the legislators that want them to happen so they can advertise to their constituents.
Is the claim here that the actual spending itself doesn’t matter, and the reason it occurs is solely that the politician likes to be able to talk about the spending?
That runs counter to my understanding; this paper claims otherwise for Brazil, and this paper suggests otherwise for the US.
I’m repeating myself, which I guess is a sign I’m not writing clearly. I think the way you’re looking at it is this:
Interest group lobbies → politician pushes for the pork- > pork gets included.
My claim is that the true mechanism is this:
Politician wants some funding to show constituents → may turn to interest groups in district to see what projects need funding (or may just know about projects in their home district) → politician pushes for pork → pork gets included
EDIT: after reflecting on this comment I think I was too dismissive of the risk of association between EA and Democrats, particularly because I think we’re headed toward a period of Republican domination of US politics and the risk of being associated with Democrats may feasibly outweigh the reward of potential policy. Anyway, below are my original thoughts.
Interesting post. I lean toward disagreeing, for a couple of reasons.
I think you would agree that Congress can, if it adopts EA legislation, be greatly helpful to the EA cause. It just has way more money and influence than EA can dream of at the moment. The questions are then:
Is having a self-proclaimed EA in Congress helpful to getting legislation passed?
Will the potential negative press and association with Democrats be too harmful to the EA movement to be worth it?
On (1), I think the answer is a resounding YES, and you have to overthink really hard to reach a different conclusion. Congresspeople deal with a ton of different interest groups all the time trying to have their preferred policies make their way into legislation. We are competing with all of them, which makes our odds of success quite low, especially since politicians generally consult the interest groups they agree with rather than interest groups actually persuading politicians. Having a congressman tirelessly devoted to the singular cause of getting EA legislation enacted, on the other hand—that can be powerful. I think the Tea Party/Squad comparison is quite a bad one, given that these groups focus on hyper-partisan legislation, which as you say EA is not. A more apt comparison in my eyes is pork. Politicians get funding all the time for projects in their districts so they can report back to their constituents. Hundreds of these things get included in many omnibus bills in order to ensure the vote of every single legislator. If Carrick Flynn is unwilling to vote for legislation without AI safety funding, and Democrats in the House need his vote due to a narrow majority, that boosts our odds significantly (especially since AI safety is pretty non-partisan and unlikely to be a sticking point in the Senate). I think you focus too much on the short-term in this analysis—politicians stick around for a long time, and Flynn could very realistically have a lot of influence in future sessions. The upside here seems massive.
For question 2, first let me comment on negative press. I’m pretty skeptical. Flynn really didn’t advertise EA at all during his campaign, and his opponents did not attack him for it (aside from crypto due to its association with wealth/corruption), and for good reason. It’s really hard to get voters to care about esoteric ideas one way or the other. Voters at large hold pretty authoritarian values and don’t care about Republicans’ attacks on democracy. Very liberal whites are basically the only ones who care about climate change. And these are esoteric issues that are already quite political—the idea that AI safety or animal welfare would become a campaign point is in my view laughable. To add, if it becomes too much of an issue, we (as a movement) can always decide it’s not worth it and stop running candidates. Flynn was a nice trial run that showed us crypto is a weakness, and it’s worth having more tests.
Now onto association with Democrats. EA will always be a left-dominated organization due to the extreme left lean of highly educated people. I do share your concern about Republicans being unwilling to pass EA legislation if it’s associated with Democrats. But I think you vastly exaggerate how much more association EA would have with Democrats if there were a couple of Democratic EA-associated legislators in office, especially if they never really talk about EA in public. And besides, going back to question 1, I still think there is a (much) higher chance of getting legislation passed if we have an EA in Congress.
Your second question “Will the potential negative press and association with Democrats be too harmful to the EA movement to be worth it?” seems to ignore that a major group EAs will be running against will be democrats in primaries.
So it’s not only that you’re creating large incentives for republicans to attack EA, you’re also creating it for e.g. progressive democrats. See: Warren endorsing Flynn’s opponent & somewhat attacking flynn for crypto billionaire sellout stuff
That seems potentially pretty harmful too. It’d be much harder to be an active group on top universities if progressive groups strongly disliked EA.
Which I think they would, if EAs ran against progressives enough that Warren or Bernie or AOC more strongly criticized EA. Which would be in line the incentives we’re creating & general vibe [pretty skeptical of a bunch of white men, crypto billionaires, etc].
I agree that pork is good comparison, since it has highly concentrated support but nonexistent opposition. But the second section I quoted above directly contradicts the first section: pork is an example of how interest groups can be effective at getting their preferred policies into legislation, even if no politician themself benefits from the pork.
It’s very rare for the groups that benefit from pork to try to steer it by themselves holding office.
I don’t think we are really competing with all other interest groups; there’s not a limited amount of legislation. There is a limited-ish budget, but it’s so high that there’s room for many different priorities already (including pork).
Well, maybe. It’s really tough to evaluate hypotheticals like this; but the actually existing political campaign that I’m commenting on lead to multiple articles about EA in the national political press just from the primary, let alone the general or somebody holding office.
I don’t think this is an accurate view of pork. Pork is pushed for by legislators, not by interest groups. These projects have some support from within the district, sure, but it’s really the legislators that want them to happen so they can advertise to their constituents. Similarly EA would be much more likely to make its way into legislation if it were pushed for by a devoted legislator than by an outside interest group.
As for the articles in the press, I think Yglesias makes a pretty convincing case that these can do quite a bit of good as well; in my mind they’re probably net good, but I understand the concern.
I don’t really understand the distinction you’re drawing. Interest groups definitely lobby to receive pork.
Is the claim here that the actual spending itself doesn’t matter, and the reason it occurs is solely that the politician likes to be able to talk about the spending?
That runs counter to my understanding; this paper claims otherwise for Brazil, and this paper suggests otherwise for the US.
I’m repeating myself, which I guess is a sign I’m not writing clearly. I think the way you’re looking at it is this:
Interest group lobbies → politician pushes for the pork- > pork gets included.
My claim is that the true mechanism is this:
Politician wants some funding to show constituents → may turn to interest groups in district to see what projects need funding (or may just know about projects in their home district) → politician pushes for pork → pork gets included