We do not and frankly that is a governance problem on a scale so staggering I staggered just writing this comment. This is exactly the kind of ‘fire alarm’ near-miss that I would’ve expected lawmakers to have specified as an immediate ‘stop all progress for now’-trigger or a ‘every time this happens you need to immediately publish everything you know internally about this incident’. Instead, we got a blog post.
I’m sure those working in technical alignment too share these frustrations to an extent, but over in policy alignment, this knee-deep opacity is a serious, constant hurdle for any unaffiliated, third-party researchers. I do hope though that this incident inspires policy overhauls internally even more than motivating new regulations.
Making someone stop littering by giving them larger fines every time they do certainly works, but it’s much more preferable if they stop by becoming the kind of person who wants to stop littering.
Do you agree with Parfit’s view on this re: normative truths though? Because I wholly agree with him here, but specifically I think because I think the usual cop-out response is not really a response at all. Imagine for example someone says “Well, but maybe there’s this alien species where for them, suffering feels really good.” My response would be to say well then that’s not suffering, or this thought experiment requires some special version of suffering that contradicts the plain English reading of the word. But then also the next part of the paragraph reads:
What are your own personal thoughts on this? Because I think I agree with this, but I am trying to understand Parfit’s view, which you will have to be a proxy for unfortunately, given that the man is dead. But in this case, does rational intuition require any kind of ability to replicate or imagine phenomenal experience? Like if someone tells me suffering is bad, and I’ve been coddled all my life and I’ve never experienced suffering outside of boredom—would Parfit say I’m well equipped enough to grasp the normative reason that suffering is bad?
Do you see this as closer to being a trivially true tautological statement, or an axiomatic one? Because I’m almost struggling to see the tautology here when we expand the term “being right” into “the right thing to do”. Like if I heard someone say “the right thing to do is that which minimizes suffering” to the question “what is the right thing to do”, I would feel weird criticizing their response as harbouring a tautology.
At the same time, if this does count as tautological, are there then any kinds of definitions that don’t contain them?
What would it mean for properties to be individuated coarsely vs finely? This is quite interesting. Is there some kind of simple motivating example that might help illustrate what the differences between these two views could be, but also why nonetheless in their perspective, those differences on this matter (or perhaps all matters) end up resolving into ‘differences of dialect’?
I’m also trying to understand, on this point, does evolution merely mean genetic evolution? Because I feel like the question of whether some property is causally inert wrt selection may well depend on whether we mean selection on the level of genes, or on the level of memes.
Because in my head claims about normative truths are important and useful precisely because they are subject to memetic selection. And if how we believe we should act affects how we act, and how we act, due to motivated reasoning and typical ego protection mechanisms in humans, affects how we rationalize the way we’ve acted, and hence what we think others should do, in what sense does Parfit mean that normative truths are causally inert? (At least on your first impression)
But also then in your opinion are claims about should-nots as much of moral truths, to the extent that realists argue there are moral truths, as claims about ‘shoulds’? Because from one perspective any claim about a thing we should not do, is also a claim about what we should do.
What about the fact of wanting ‘[at least someone] to be right in debates over morality’? Because naively in my head at least if I try to slot in a different fundamental truth into the sentence, then there is at least one coherent reading of it (but I don’t know if this merely seems coherent to me, or if you would agree). But I would say that “the mere fact that I would like to be right in debates over the speed of light implies something about the existence of a truth about the speed of light”, is a totally respectable thing to claim, but is this a misapplication?
Thank you for this excellent write-up by the way. I probably would not have ever fully read volume 3 end to end, so I’m glad someone else in our community did the heavy lifting for us. My only request for the future is more pictures and graphs.