Admittedly, we weren’t factoring in the (ostensible) ripple effects, but our modeling indicates that if we’re interested in robust goodness, we should be spending on chickens.
Also, for the reasons that @Ariel Simnegar already notes, even if there are unappreciated benefits of investing in GHD, there would need to be a lot of those benefits to justify not spending on animals. Could work out that way, but I’d like to see the evidence. (When I investigated this myself, making the case seemed quite difficult.)
That’s all assuming neartermism, right? I agree that neartermism plausibly entails prioritizing non-human animals. But neartermism seems very arbitrary, and should plausibly not receive nearly as much weight as GHD currently receives in OP’s portfolio.*
Rather, my suggestion was that the current explicitly “longtermist” bucket should be thought of as something like “highly speculative longtermism”, and the current GHD bucket should be thought of as something like “improving the long-term via robustly good methods”.
It’s harder to see how helping animals would fit that latter description. (Maybe the best case would be the classic Kantian argument that abusing animals less would help make us morally better.)
*: Insofar as the GHD bucket is really just motivated by something like “sticking close to common sense”, and people started using “neartermism” as a slightly misnamed label for this, the current proposal to shift to very counterintuitive priorities whilst doubling down on the unprincipled rejection of longtermism seems kind of puzzling to me!
We shouldn’t just assume that neartermism is the principled alternative to speculative longtermism, and then stick to that assumption come what may (i.e., even if it leads to the result that we should “mostly” ignore poor people). Rather, I think we should be more open-minded about how we should think of the different “buckets” in a Worldview-Diversified portfolio, and cautious of completely dismissing common-sense priorities (even as we give significant weight and support to a range of theoretically well-supported counterintuitive cause areas).
Nope, not assuming neartermism. The report has the details. Short version: across a range of decision theories, chickens look really good.
That said, I totallyagree that from a purely conceptual perspective, we should “be more open-minded about how we should think of the different ‘buckets’ in a Worldview-Diversified portfolio, and cautious of completely dismissing common-sense priorities (even as we give significant weight and support to a range of theoretically well-supported counterintuitive cause areas).”
Edit to add: I think you might mean something different by “robust goodness” than what I had in mind. From a quick look at your link, you’re considering a range of different decision theories, risk-weightings, etc., and noting that chickens do at least moderately well on a wide range of theoretical assumptions.
I instead meant to be talking about empirical robustness: roughly, “helping the long-term via methods that are especially likely to do some immediate good, and with less risk of proving long-term counterproductive.” Or, more concisely, “longtermism via nearterm goods with positive ripple effects”. And then assessing what does best via this particular theoretical standard (to make up one bucket in our portfolio).
Since your report doesn’t consider ripple effects, it doesn’t address the kind of “robust longtermism” bucket I have in mind.
Admittedly, we weren’t factoring in the (ostensible) ripple effects, but our modeling indicates that if we’re interested in robust goodness, we should be spending on chickens.
Also, for the reasons that @Ariel Simnegar already notes, even if there are unappreciated benefits of investing in GHD, there would need to be a lot of those benefits to justify not spending on animals. Could work out that way, but I’d like to see the evidence. (When I investigated this myself, making the case seemed quite difficult.)
That’s all assuming neartermism, right? I agree that neartermism plausibly entails prioritizing non-human animals. But neartermism seems very arbitrary, and should plausibly not receive nearly as much weight as GHD currently receives in OP’s portfolio.*
Rather, my suggestion was that the current explicitly “longtermist” bucket should be thought of as something like “highly speculative longtermism”, and the current GHD bucket should be thought of as something like “improving the long-term via robustly good methods”.
It’s harder to see how helping animals would fit that latter description. (Maybe the best case would be the classic Kantian argument that abusing animals less would help make us morally better.)
*: Insofar as the GHD bucket is really just motivated by something like “sticking close to common sense”, and people started using “neartermism” as a slightly misnamed label for this, the current proposal to shift to very counterintuitive priorities whilst doubling down on the unprincipled rejection of longtermism seems kind of puzzling to me!
We shouldn’t just assume that neartermism is the principled alternative to speculative longtermism, and then stick to that assumption come what may (i.e., even if it leads to the result that we should “mostly” ignore poor people). Rather, I think we should be more open-minded about how we should think of the different “buckets” in a Worldview-Diversified portfolio, and cautious of completely dismissing common-sense priorities (even as we give significant weight and support to a range of theoretically well-supported counterintuitive cause areas).
Nope, not assuming neartermism. The report has the details. Short version: across a range of decision theories, chickens look really good.
That said, I totally agree that from a purely conceptual perspective, we should “be more open-minded about how we should think of the different ‘buckets’ in a Worldview-Diversified portfolio, and cautious of completely dismissing common-sense priorities (even as we give significant weight and support to a range of theoretically well-supported counterintuitive cause areas).”
Ok, thanks for clarifying!
Edit to add: I think you might mean something different by “robust goodness” than what I had in mind. From a quick look at your link, you’re considering a range of different decision theories, risk-weightings, etc., and noting that chickens do at least moderately well on a wide range of theoretical assumptions.
I instead meant to be talking about empirical robustness: roughly, “helping the long-term via methods that are especially likely to do some immediate good, and with less risk of proving long-term counterproductive.” Or, more concisely, “longtermism via nearterm goods with positive ripple effects”. And then assessing what does best via this particular theoretical standard (to make up one bucket in our portfolio).
Since your report doesn’t consider ripple effects, it doesn’t address the kind of “robust longtermism” bucket I have in mind.