Yes, but this means that you know something additional about the structure of the representor, not just its range. I'm asking whether we can do better than maximality even for intervals alone, without adding any further details.

(By the way, pictures are broken.)

UPD: Though, you probably mean that we do know some additional structure for EV if we look at how it is constructed from probabilities and utilities, for which we have just intervals without structure.

Thanks.

No, I don't think so. I don't remember him mentioning such an implication, and he has sometimes explicitly mentioned the risk that humanity will end up happy and unconcerned about wild animal suffering, here or elsewhere (example).

Good work! It's useful to have such a collection of arguments on space colonization all in one place.

I personally disagree with Pearce on AI risks and digital suffering (which he denies) and on metaethics (he's a moral realist), for instance.

I also worried that happier people don't necessarily imply wider moral circle, more compassion and so on, so it's not clear what to expect from scenarios where, for example, there is widespread selection against suffering at the genetic level.

(Strategically I agree more with Magnus Vinding and Brian Tomasik on this topic.)

In particular, it would be interesting to know how likely such selection pressures (as described by Pearce) are to be significant, and how they will affect human values.

He seems to be pretty optimistic about his Abolitionist Project (I'm not). If you're familiar with his work, where do you think he's making wrong assumptions when predicting the future?

What do you think about "The Reproductive Revolution" by David Pearce (or his other works, such as "The Hedonistic Imperative") in this context?

It seems that your view is close to some kind of person-affecting view. I'm not very familiar with person-affecting views. From what I've read, it seems to me that there are options like this:

1. A theory that requires us to care only about the present generation (because only present humans exist, future humans are merely possible).
2. A theory that additionally requires us to care about future generations, but only from some downside-focused perspective (in particular, a suffering-focused one).

Then (1) would escape much of cluelessness because we don't need to aggregate over the whole cosmos, but (2) would be largely equivalent to some regular theory (I mean a theory that does not rest on a person-affecting condition), such as regular negative utilitarianism, and so would not automatically escape DiGiovanni's arguments but would face the same problem.

Is your theory different from (1-2)? How do you determine which beings are "merely possible" and which ones "(will) exist for sure"? Which beings count as claim-holders? Would you classify your theory as downside-focused? Could we understand the relationship between "claims" and "impersonal value" as something like lexical priority? If we restrict our consideration to claims only, how does your theory differ, at the axiological level, from negative preference utilitarianism? Do you agree that we should maximize something like expected value, where the relevant value is the aggregate strength of all claims? (In other words, how different is your approach at a decision-theoretical level? Is it scope- and probability-sensitive?)

The maximality rule is too demanding. Under maximality, we can't even say that [1, 10 000] is better than [-10 000, 1.0001]. Are there any good reasons to avoid more permissive rules even in cases like this?

Of course, if your UEV intervals overlap and you choose one of them, you may make a mistake if your idealized version would choose the other one. But choosing at random is no better in this regard.

What seems important is how costly such mistakes are, i.e., how the overall performance of your choice rule compares with that of other rules—such as choosing at random.

Can these performance measures be defined without additional assumptions about how values are distributed within UEV intervals?

Why complete cluelessness is counterintuitive to me

I mean, specifically, this kind of pervasive cluelessness, where you can't justifiably decide between any two actions. It seems that such pervasive cluelessness is in tension with the idea of instrumental convergence. I'm personally sympathetic to imprecise probabilities, but intuitively I'm not convinced that cluelessness is that bad that we can't justify even basic learning and other (supposedly convergent) instrumental strategies, such as (at least) those in the low-footprint capacity building category.

And I think, if we have to show that cluelessness is at least not that pervasive, it seems useful to look at the most seemingly absurd cases. Being clueless about whether epistemic improvement is worthwhile is one of them. If we (or other aligned agents that we create) can in principle be non-clueless about some specific class of strategies, for instance, why wouldn't we at least sometimes be justified in taking actions that would make us non-clueless about them?^[1] (hence, being additionally non-clueless about whether this transition is better than doing nothing or pursuing some other alternative with a small opportunity cost) And if we are indeed justified in doing so, why can't we derive some broader system of instrumental strategies from this fact?^[2] (hence, potentially, having even more non-clueless states) It sounds strange that even strategies aimed at becoming non-clueless wouldn't be better justified than doing something random.

I have to admit, the situation is much less obvious than I first thought. Interestingly, some imprecise probabilists may have to pay to avoid free knowledge (unlike precise ones). And there is a thing called dilation: new information may make your intervals wider than before. But most importantly, knowledge is not actually free. And it's yet unclear to me how we should account for long-term changes in the structure of our entire decision trees in general. This seems to involve the topic of sequential rationality.

Anyway, why is impartial altruism so different from other value systems in this respect?^[3] DiGiovanni writes:

...suppose that instead of spending some of your free time reflecting on virtue ethics, you reflect on whether some intervention could be robust to unawareness. Maybe this shapes the salience of different strategies to your future self, and you end up attempting an intervention that’s worse than what you would’ve done by default.

In principle, we can come up with examples where learning goes wrong, but it's not yet clear why we should treat them as severely undermining the whole idea of epistemic improvement.

Intuitively, these side effects may seem like hand-wringing nitpicks. But I think this intuition comes from mistakenly privileging intended consequences, and thinking of our future selves as perfectly coherent extensions of our current selves.

But we don't have to think that our future selves are such perfectly coherent extensions. They are indeed different, but they may differ to a degree small enough for us to make a justified decision.

So far, it is not clear why impartiality is special. Learning seems to have broadly similar structures across many domains. Impartial altruism doesn't seem to be an exceptional domain such that some magical demons suddenly appear from nowhere to make you go mad if you know too much.^[4] Then why aren't there some learning strategies available to us that are more justified than doing nothing?^[5]

We’re trying to weigh up speculative upsides and downsides to a degree of precision beyond our reach. In the face of this much epistemic fog, unintended effects could quite easily toggle the large-scale levers on the future in either direction.

This seems to be relevant when deciding between some direct interventions (such as donations), but how does the associated cluelessness infect learning itself—which is supposed to help us compare them? Cluelessness may emerge here if we don't know how to compare learning with its alternative, which is possibly higher-EV. But what if the alternative we consider is something which has low direct impact on your environment, such as resting? Then by choosing to learn, you may miss something valuable if it lies further down your decision tree and resting makes that opportunity more accessible. This is certainly normal, but isn't unique to impartial altruism.

We shouldn't always prefer learning to doing nothing—it's even possible that we should do it less often than some other things. This is because successful approaches to epistemic improvement should mix learning with other things, such as resting when tired, eating when hungry, not dying in the process, etc. But isn't it possible for us to identify some combination of such low-impact actions that is better than, for instance, indefinitely pursuing just one of them? Empirically, this seems possible in many domains, and we may be able to identify some similarities across them. So again, it seems unclear, what is it about impartial altruism that makes things more clueless here—to the degree that we can't even decide whether we better be alive and learn things.

It will be helpful to have greater clarity about the potential sources and mechanisms of cluelessness (and their relative strength) in such specific most-absurdly-looking cases.^[6]

So the idea is, as I understand it, that the maximality rule is too harsh when it comes to incomparability, while also being too sensitive to small changes, and that we may have much more choice-relevant information about UEV than maximality suggests. This information is better preserved if we take a graded approach to comparison. We can then use these graded verdicts, such as "A is 0.98-better than B", to favor A over B even when maximality says A and B are simply incomparable. We are thus able to see much more, although incomparability may still be too problematic when the degree of betterness is small.

So it seems that this approach allows us to make more informative, fine-grained judgements about cluelessness.

Very interesting! Curious to see where it takes us.

P. S. I agree that low-footprint capacity building seems to be one of the most promising strategies to overcome cluelessness.