Locrian

OK, very interesting! I was not expecting that response. Your view is kind of hard for me to fathom.

I definitely care at a gut level about there being happy more-or-less human people 100,000 years from now, and it seems hard for me to separate this feeling (in my own psyche, not conceptually^[1]) from "I care about my own longevity and happiness, and that of people I know." Maybe one way to try to communicate this feeling is "I expect that 'I' will still still be effectively 10-30% 'alive' as long as any humans exist, for the meanings of "I" and "alive" tha... (read more)

Hmm--when you say "the 'extinction is special' argument," I can imagine this referring to at least two different things:

1. A view that the difference between, say, 99% of currently-alive people dying and 100% is a lot more important than the difference between 98% and 99%, even though they represent the same number of lives, because if anyone survives then there will still be a future for human beings.
2. A view that extinction is literally infinitely bad, in a Pascal's Mugging kind of way where we should pay any price to reduce x-risk a tiny amount.

#1 seems extr... (read more)

This essay seems to be working from the idea that typical longtermists think that there is a chance of civilizational collapse from non-AI causes roughly comparable to x-risk from AI, and inferring that longtermists don't care about civilizational collapse. My impression is that most longtermists do in principle care about civilizational collapse (insofar as it prevents humanity from reaching the ideal long-term future), but just think it's much less likely than extinction from AI.

I think the best attitude towards advice depends a lot the topic. Broad-scale career planning is probably one of the worst areas for advice (with some exceptions), because:

1. Yeah, as you say, the relevant systems are more complex and chaotic than people tend to tell themselves, so it's hard for other people to accurately understand the causes of their success or failure. It could be that if they reran their life, that advice wouldn't work again.
2. Even if they understand exactly what caused their success, different places, times, and fields have very different

Nifty! But man, those are some generous raters. It's interesting that, besides being not very amusing, these jokes feel to me pretty far behind regular model outputs in terms of logical coherence and world-modeling (e.g. are funerals usually given to bad things more than good ones? what is the normal meaning of "changing angles of paintings in my gallery" that is being undercut by "facing the wall"?).

It would be interesting to try this with different genres of humor besides joke-book one-liners. For example (where you could switch out the bracketed stuff f... (read more)

The version of this argument that really bugs me, and that your AI examples demonstrate, is when the "good" end of the comparison is also really bad. I have heard someone say, in earnest, a near paraphrase of "OK, maybe AI could kill everyone, but humans commit genocide too...."

every early state on Earth was a grain state

What does Scott make of tuber-based Andean civilization?

I enjoyed this post! In the spirit of presenting images to ponder, this made me think of The Discarded Image by CS Lewis. It is one of Lewis' scholarly books intended for students of medieval literature, describing the cosmology of that period. One of the points he wants to deliver is that, unlike moderns, medieval people didn't think they were the most intelligent created beings--in church teaching there were nine species of angels arranged hierarchically between us and the absolute intelligence of God. Of course it is one thing to view humankind as the l... (read more)

I don't feel strongly for or against this idea, but I don't see how it makes LW a better place to share your own writing. Is the idea that it would increase the audience for each post, and posters want a larger audience? I would think that the LW frontpage already has a moderate-to-large audience-size advantage over most people's personal Substacks.

(Checking this: Claude estimates 3-7k LW homepage visits per day, and 1-3k readers for a typical post in its first week. This is way below a big Substack's numbers, of course, but in the range of a lot of established, in-the-conversation rationalist Substacks I subscribe to.)

Hi all--I have been a reader of LW and other rationalist writing for the last 15 years. Professionally, I am an ecologist of a mathematical bent, with wide interests. I have created an account mostly to be able to auto-filter the LW frontpage to my liking, but I will occasionally comment and maybe post as the spirit moves me.

If the alternative uncertainty-resolution method is a negotiated settlement, the act of starting negotiations potentially changes the balance of force by the end of the negotiation, since opposing power will learn that you are considering a war and have time to prepare itself. Many modern military conflicts have started with a surprise attack intended to knock out or demoralize the enemy before they could build up their full strength. Once the attack begins, the fact that the defender resists might indicate that you were overconfident, but it could also be... (read more)