Lovely post and I agree with most of your philosophical points, but I think this is very much a case of you defending the principal at the cost of weakening and diluting it so much that it’s not clear whether it actually matters in practice because after all, just because our uncertainty over whether say more economic growth in the developing world is good given effects on wild animals doesn’t meet your strict and demanding definition of cluelessness doesn’t mean that in practice, we are not still in a somewhat comparable position where while in theory, an agent might be able to resolve the question to their satisfaction in practice. It’s very hard to see what we should do about it. Basically, for a day-to-day actions, it may not matter that much whether we should be strictly clueless as opposed to merely having no idea what we should do in the more mundane sense because the both possibilities have somewhat similar outcomes. Even if for example, we can conclude that perhaps more investigation into for example, the consciousness of insects or the ecological impacts of more growth in the developing World would be helpful in resolving our question. Basically, the points at which you touch on principles that we could apply in our actual day-to-day decision making in the real world are precisely those points where I think your post is weakest so I’m not sure how much you manage to defend the idea that we should behave much more differently than if we were genuinely clueless, even if you mount a strong case against actual true cluelessness.
re: "it may not matter that much whether we should be strictly clueless as opposed to merely having no idea what we should do in the more mundane sense" - a key issue is whether the reference to "having no idea" is merely descriptive (as a matter of fact, we often feel this way: we haven't yet reached a conclusion) or normative (positively claiming that we *should* have no idea and *shouldn't* reach any more determinate conclusion).
My overarching point is that, in practice, people should feel free to make judgment calls, even if they're fallible. The cluelessness advocate, by contrast, prohibits people from doing that and *mandates* suspension of judgment.
Okay, that’s fair and certainly something that you manage to establish, and that’s eminently defensible, even though is much less ambitious than a more substantive defence against the cluelessness objection. Although to be fair, I perhaps mistakenly have generally understood the cluelessness objection to be less about arguing that people should not exercise their judgement and more arguing that there is no particular reason to favour this or that intervention compare to other alternatives.
I think part of what’s going on is that when you talk about cluelessness you have a very specific set of technical philosophical positions in mind, which makes sense since you are yourself a philosopher. Where as I when I hear that word think of a much more different critique, which I think of by the same name, even though it has much less extreme theoretical ambitions compare to the philosophical project you’re talking about so I basically misunderstood, what you were responding to because I use the word to mean something else.
Good post! I agree with most of it. However that last principle seems very undermotivated. Why not expect that any action will regress to the mean in terms of long run effects absent any specific reason in the other direction? It's not like actions that are good in the short run have some categorically different feature that makes their long term causal powers meaningfully different--at least then we should get an explanation of what that is.
It seems to me that we should just treat long run EV as neutral unless we have specific evidence otherwise, and then simply count up the EV we can know.
I was thinking of long-run EV as *including* the short-run (so I should have written "invisible-inclusive value" rather than just "invisible value" in the official statement of the principle). But also thinking that there are often positive ripple-effects from empowering people -- slightly greater economic growth, for example. All we need for positive expectation is a greater expectation of positive than negative ripple effects: you can still think that the *most* likely outcome is a washout.
Right, I guess it's a matter of how long term you're thinking. It seems right that there short term good actions are often good in the predictable near long term, if you will.
Lovely post and I agree with most of your philosophical points, but I think this is very much a case of you defending the principal at the cost of weakening and diluting it so much that it’s not clear whether it actually matters in practice because after all, just because our uncertainty over whether say more economic growth in the developing world is good given effects on wild animals doesn’t meet your strict and demanding definition of cluelessness doesn’t mean that in practice, we are not still in a somewhat comparable position where while in theory, an agent might be able to resolve the question to their satisfaction in practice. It’s very hard to see what we should do about it. Basically, for a day-to-day actions, it may not matter that much whether we should be strictly clueless as opposed to merely having no idea what we should do in the more mundane sense because the both possibilities have somewhat similar outcomes. Even if for example, we can conclude that perhaps more investigation into for example, the consciousness of insects or the ecological impacts of more growth in the developing World would be helpful in resolving our question. Basically, the points at which you touch on principles that we could apply in our actual day-to-day decision making in the real world are precisely those points where I think your post is weakest so I’m not sure how much you manage to defend the idea that we should behave much more differently than if we were genuinely clueless, even if you mount a strong case against actual true cluelessness.
re: "it may not matter that much whether we should be strictly clueless as opposed to merely having no idea what we should do in the more mundane sense" - a key issue is whether the reference to "having no idea" is merely descriptive (as a matter of fact, we often feel this way: we haven't yet reached a conclusion) or normative (positively claiming that we *should* have no idea and *shouldn't* reach any more determinate conclusion).
My overarching point is that, in practice, people should feel free to make judgment calls, even if they're fallible. The cluelessness advocate, by contrast, prohibits people from doing that and *mandates* suspension of judgment.
Okay, that’s fair and certainly something that you manage to establish, and that’s eminently defensible, even though is much less ambitious than a more substantive defence against the cluelessness objection. Although to be fair, I perhaps mistakenly have generally understood the cluelessness objection to be less about arguing that people should not exercise their judgement and more arguing that there is no particular reason to favour this or that intervention compare to other alternatives.
I think part of what’s going on is that when you talk about cluelessness you have a very specific set of technical philosophical positions in mind, which makes sense since you are yourself a philosopher. Where as I when I hear that word think of a much more different critique, which I think of by the same name, even though it has much less extreme theoretical ambitions compare to the philosophical project you’re talking about so I basically misunderstood, what you were responding to because I use the word to mean something else.
(For readers: I've replied to this on Richard's cross-post on the EA Forum: https://forum.effectivealtruism.org/posts/CyNmdcweSb7jKWeZR/hard-questions-aren-t-hopeless.)
Good post! I agree with most of it. However that last principle seems very undermotivated. Why not expect that any action will regress to the mean in terms of long run effects absent any specific reason in the other direction? It's not like actions that are good in the short run have some categorically different feature that makes their long term causal powers meaningfully different--at least then we should get an explanation of what that is.
It seems to me that we should just treat long run EV as neutral unless we have specific evidence otherwise, and then simply count up the EV we can know.
I was thinking of long-run EV as *including* the short-run (so I should have written "invisible-inclusive value" rather than just "invisible value" in the official statement of the principle). But also thinking that there are often positive ripple-effects from empowering people -- slightly greater economic growth, for example. All we need for positive expectation is a greater expectation of positive than negative ripple effects: you can still think that the *most* likely outcome is a washout.
Right, I guess it's a matter of how long term you're thinking. It seems right that there short term good actions are often good in the predictable near long term, if you will.