First off, thank you for this research and for sharing it with the community. My overall feeling on this work is extremely positive, and the below is one (maybe my only?) critical nitpick, but I think it is important to voice.
Causes classified as longtermist were Biosecurity, Nuclear risk, AI risk, X-risk other and Other longtermist.
Causes classified as neartermist were Mental health, Global poverty and Neartermist other.
Causes classified as Other were Animal Welfare, Cause Prioritization, EA movement building and Climate change.
I have to object to this. I donāt think longtermism is best understood as a cause, or set of causes, but more as a justification for working on certain causes over others. e.g.:
Working on Nuclear Risk could be seen as near-termist. You can have a person-affecting view of morality and think that, given the track record of nuclear near-miss incidents, that itās a high priority for the wellbeing of people alive today
We just lived through a global pandemic, there is active concern about H5N1 outbreaks right now, so it doesnāt seem obvious to me that many people (EA or not) would count biosecurity in the ālongtermistā bucket
Similarly, many working on AI risk have short timelines that have only gotten shorter over the past few years.[1]
Climate Change could easily be seen through a ālongtermistā lens, and is often framed in the media as being an x-risk or affecting the lives of future generations
Approaching Global Poverty from a āgrowth > randomistaā perspective could easily be justified from a longtermist lens given the effects of compounding returns to economic growth for future generations
(Iām sure everyone can think of their own counter-examples)
I know the groupings came out of some previous factor analysis you did, and you mention the cause/ājustification difference in the footnotes, and I know that there are differences in community cause prioritisation, but I fear that leading with this categorisation helps to reify and entrench those divisions instead of actually reflecting an underlying reality of the EA movement. I think itās important enough not to hide the details in footnotes because otherwise people will look at the ālongtermistā and āneartermistā labels (like here) and make claims/āinferences that might not correspond to what the numbers are really saying.
I think part of this is downstream of ālongtermismā being poorly defined/āunderstood (as I said, it is a theory about justifications for causes rather than specific causes themselves), and the ālongtermist turnā having some negative effects on the community, so isnāt a result of your survey. But yeah, I think we need to be really careful about labelling and reifying concepts beyond the empirical warrant we have, because that will in turn have causal effects of the community.
In fact, I wonder if AI was separated ut from the other 3 ālongtermistā causes, what the others might look like. I think a lot of objections to ālongtermismā are actually objections to prioritising āAI x-riskā work.
I have to object to this. I donāt think longtermism is best understood as a cause, or set of causes, but more as a justification for working on certain causes over others. e.g.:
Working on Nuclear Risk could be seen as near-termist.
We agree and say as much:
As we have noted previously, people may of course prioritise these causes for reasons other than longtermism or neartermism per se. Likewise, people might support the āOtherā causes here for neartermist or longtermist reasons.
And hereās what we say at the linked previous post:
For simplicity, we label support for these, non-longtermist, causes āneartermistā, as we have in previous years. However, itās worth noting explicitly that there is little reason to suppose that neartermism specifically, (e.g., attitudes or beliefs related to helping present vs. future generations or different time preference) explain support for these causes, rather than different epistemic beliefs (e.g., about appropriate kinds of evidence) or support for more traditional causes etc.
We also explicitly discuss your example of Climate Change making much the same point:
The classifications we used here were informed by our prior analyses of the factor structure of the cause prioritisation items and a priori theoretical considerations (namely Climate Change is clearly associated with the neartermist causes not the longtermist causes, but theoretically some might think of it as a longtermist cause, and we count it as āOtherā in this analysis)
I acknowledge though that your concern is that āthis categorisation helps to reify and entrench those divisionsā. I think this is possible, but I think that:
It is important to talk about that (I think weād be more confused/āless informed if we just considered all the causes as separate/āindependent).
Referring to these clusters of causes and ideas in terms of ālongtermismā and āneartermismā is established terminology. Crucially, I donāt think thereās an obviously better set of term, because āexistential riskā wouldnāt capture some causes within this bucket (e.g. in previous years we had a ābroad longtermismā item which was also part of this cluster)
I think itās important enough not to hide the details in footnotes
I think itās an entirely reasonable view to think discussion of this should be in the text, not a footnote. Though we had a lot of information in both the footnotes and appendix, so itās tricky.
Though I donāt claim itās a single thing, rather than a cluster of correlated things. And empirically, our longtermist-neartermist cause scores measure is strongly correlated with peopleās stated abstract beliefs. The single abstract item explicitly about longtermism, is correlated with LT-NT at r=0.457, which is appreciably strong for a necessarily noisy cause prioritisation score and single item, in a social science context.
Thanks for responding David, and again I think that the survey work youāve done is great :) We have many points of agreement:
Agreed that you basically note my points in the previous works (both in footnotes and in the main text)
Agreed that itās always a hard tradeoff when compressing detailed research findings into digestible summaries of researchāI know from professional experience how hard that is!
Agreed that there is some structure which your previous factor analysis and general community discussions picked up on, which is worth highlighting and examining
I still think that the terminology is somewhat misguided. Perhaps the key part I disagree is that āReferring to these clusters of causes and ideas in terms of ālongtermismā and āneartermismā is established terminologyāāeven if it has been established I want to push back and un-establish because I think itās more unhelpful and even harmful for community discussion and progress. Iām not sure what terms are better, though some alternatives Iāve seen have been:[1]
I guess, to state my point as clearly as possible, I donāt think the current cluster names ācarve nature at its jointsā, and that the potential confusion/āambiguity in use could lead to negative perceptions that arenāt accurate became entrenched
Thanks JWS, it certainly sounds like we agree more than we disagree.
even if it has been established I want to push back and un-establish
Thatās definitely fair!
For what itās worth I think that the explanation for differences in support for these two different clusters of causes is more epistemic and than it is to do with attitudes towards the longterm or near term per se.[1] Ideally, Iād like the terms we use to not (be seen to) refer to the explanation for supporting the causes at all, since I think the reasons are heterogeneous.
In any case, we definitely agree that none of these terms are perfect, and I suspect no terms are going to be completely satisfactory, but Iām open to continued discussion about what better terms would be.
Although, in terms of predicting āLT minus NTā cause prioritisation from our cause-related idea items, the ālong term futureā item and ālow probability, high impactā items were about equally predictive.
Interestingly, this also holds true in unpublished work we have looking at the general public, for whom objections that influencing the far future is impractical or impossible are more consequential than their lack of concern for future generations.
First off, thank you for this research and for sharing it with the community. My overall feeling on this work is extremely positive, and the below is one (maybe my only?) critical nitpick, but I think it is important to voice.
I have to object to this. I donāt think longtermism is best understood as a cause, or set of causes, but more as a justification for working on certain causes over others. e.g.:
Working on Nuclear Risk could be seen as near-termist. You can have a person-affecting view of morality and think that, given the track record of nuclear near-miss incidents, that itās a high priority for the wellbeing of people alive today
We just lived through a global pandemic, there is active concern about H5N1 outbreaks right now, so it doesnāt seem obvious to me that many people (EA or not) would count biosecurity in the ālongtermistā bucket
Similarly, many working on AI risk have short timelines that have only gotten shorter over the past few years.[1]
Climate Change could easily be seen through a ālongtermistā lens, and is often framed in the media as being an x-risk or affecting the lives of future generations
Approaching Global Poverty from a āgrowth > randomistaā perspective could easily be justified from a longtermist lens given the effects of compounding returns to economic growth for future generations
EA movement building has often been criticised as focusing on ālongtermistā causes above others, and that does seem to be where the money is focused
Those concerned about Animal Welfare also have concerns about how humanity might treat animals in the future, and if we might lock-in our poor moral treatment of other beings
(Iām sure everyone can think of their own counter-examples)
I know the groupings came out of some previous factor analysis you did, and you mention the cause/ājustification difference in the footnotes, and I know that there are differences in community cause prioritisation, but I fear that leading with this categorisation helps to reify and entrench those divisions instead of actually reflecting an underlying reality of the EA movement. I think itās important enough not to hide the details in footnotes because otherwise people will look at the ālongtermistā and āneartermistā labels (like here) and make claims/āinferences that might not correspond to what the numbers are really saying.
I think part of this is downstream of ālongtermismā being poorly defined/āunderstood (as I said, it is a theory about justifications for causes rather than specific causes themselves), and the ālongtermist turnā having some negative effects on the community, so isnāt a result of your survey. But yeah, I think we need to be really careful about labelling and reifying concepts beyond the empirical warrant we have, because that will in turn have causal effects of the community.
In fact, I wonder if AI was separated ut from the other 3 ālongtermistā causes, what the others might look like. I think a lot of objections to ālongtermismā are actually objections to prioritising āAI x-riskā work.
Thanks for the detailed comment!
We agree and say as much:
And hereās what we say at the linked previous post:
We also explicitly discuss your example of Climate Change making much the same point:
I acknowledge though that your concern is that āthis categorisation helps to reify and entrench those divisionsā. I think this is possible, but I think that:
There is something[1] unifying these causes
It is important to talk about that (I think weād be more confused/āless informed if we just considered all the causes as separate/āindependent).
Referring to these clusters of causes and ideas in terms of ālongtermismā and āneartermismā is established terminology. Crucially, I donāt think thereās an obviously better set of term, because āexistential riskā wouldnāt capture some causes within this bucket (e.g. in previous years we had a ābroad longtermismā item which was also part of this cluster)
I think itās an entirely reasonable view to think discussion of this should be in the text, not a footnote. Though we had a lot of information in both the footnotes and appendix, so itās tricky.
Though I donāt claim itās a single thing, rather than a cluster of correlated things. And empirically, our longtermist-neartermist cause scores measure is strongly correlated with peopleās stated abstract beliefs. The single abstract item explicitly about longtermism, is correlated with LT-NT at r=0.457, which is appreciably strong for a necessarily noisy cause prioritisation score and single item, in a social science context.
Thanks for responding David, and again I think that the survey work youāve done is great :) We have many points of agreement:
Agreed that you basically note my points in the previous works (both in footnotes and in the main text)
Agreed that itās always a hard tradeoff when compressing detailed research findings into digestible summaries of researchāI know from professional experience how hard that is!
Agreed that there is some structure which your previous factor analysis and general community discussions picked up on, which is worth highlighting and examining
I still think that the terminology is somewhat misguided. Perhaps the key part I disagree is that āReferring to these clusters of causes and ideas in terms of ālongtermismā and āneartermismā is established terminologyāāeven if it has been established I want to push back and un-establish because I think itās more unhelpful and even harmful for community discussion and progress. Iām not sure what terms are better, though some alternatives Iāve seen have been:[1]
Richard Chappellās āPure suffering reduction vs Reliable global capacity growth vs High-impact long-shotsā
Laura Duffyās āEmpirical EA vs Reason-driven EAā
Ryan Briggsās āBed-nets vs Light-coneā
I guess, to state my point as clearly as possible, I donāt think the current cluster names ācarve nature at its jointsā, and that the potential confusion/āambiguity in use could lead to negative perceptions that arenāt accurate became entrenched
Though I donāt think any of them are perfect distillations
Thanks JWS, it certainly sounds like we agree more than we disagree.
Thatās definitely fair!
For what itās worth I think that the explanation for differences in support for these two different clusters of causes is more epistemic and than it is to do with attitudes towards the longterm or near term per se.[1] Ideally, Iād like the terms we use to not (be seen to) refer to the explanation for supporting the causes at all, since I think the reasons are heterogeneous.
In any case, we definitely agree that none of these terms are perfect, and I suspect no terms are going to be completely satisfactory, but Iām open to continued discussion about what better terms would be.
Although, in terms of predicting āLT minus NTā cause prioritisation from our cause-related idea items, the ālong term futureā item and ālow probability, high impactā items were about equally predictive.
Interestingly, this also holds true in unpublished work we have looking at the general public, for whom objections that influencing the far future is impractical or impossible are more consequential than their lack of concern for future generations.
FWIW I made my comment with that in mind, I think those causes involve more psychological distance even if theyāre not explicitly longtermist