I've explored very similar ideas before in things like this simulation based on the Iterated Prisoner's Dilemma but with Death, Asymmetric Power, and Aggressor Reputation. Long story short, the cooperative strategies do generally outlast the aggressive ones in the long run. It's also an idea I've tried to discuss (albeit less rigorously) before as The Alpha Omega Theorem and Superrational Signalling. The first of those was from 2017 and got downvoted to oblivion, while the second was probably too long-winded and got mostly ignored.

There are a bunch of random people like James Miller and A.V. Turchin and Ryo who have had similar ideas that can broadly be categorized under Bostrom's concept of Anthropic Capture, or Game Theoretic Alignment, or possibly a subset of Agent Foundations. The ideas are mostly not taken very seriously by the greater LW and EA communities, so I'd be prepared for a similar reception.

Oh man, I remember the days when Eliezer still called it Friendly and Unfriendly AI. I actually used one of those terms in a question when I was at a Q&A after a tutorial by the then less famous Yoshua Bengio at the 27th Canadian Conference on AI in 2014. He jokingly replied by asking if I was a journalist, before giving a more serious answer saying we were so far away from having to worry about that kind of thing (AI models back then were much more primitive, it was hard to imagine an object recognizer being dangerous). Fun times.

Strong upvoted as that was possibly the most compelling rebuttal to the simulation argument I've seen in quite a while, which was refreshing for my peace of mind.

That being said, it mainly targets the idea of a large-scale simulation of our entire world. What about the possibility that the simulation is for a single entity and that the rest of the world is simulated at a lower fidelity? I had the thought that a way to potentially maximize future lives of good quality would be to contain each conscious life in a separate simulation where they live reasonably good lives catered to their preferences, with the apparent rest of the world being virtual. Given, I doubt this conjecture because in my own opinion my life doesn't seem that great, but it seems plausible at least?

Also, that line about the diamond statue of Hatsune Miku was very, very amusing to this former otaku.

If I recall correctly, the old Felicifia forums (archive) had a lot of debates between negative and other utilitarians about this exact thing. There are also lots of other thought experiment-like "repugnant conclusions" that go with various forums of utilitarianism, including the "Hedonium Shockwave" idea, where you tile the universe with happiness generating computronium as the most efficient way to maximize utility.

The reality is, it's very hard to avoid weird hypothetical conclusions when you take as your ethics a simple rule like "minimize suffering" or "maximize happiness". This is a known problem with consequentialist ethics, and it's up to you if you want to bite the bullet or follow your moral intuitions.

I've thought about this a lot too. My general response is that it is very hard to see what one could do differently at a moment to moment level even if we were in a simulation. While it's possible that you or I are alone in the simulation, we can't, realistically, know this. We can't know with much certainty that the apparently sentient beings who share our world aren't actually sentient. And so, even if they are part of the simulation, we still have a moral duty to treat them well, on the chance they are capable of subjective experiences and can suffer or feel happiness (assuming you're a Utilitarian), or have rights/autonomy to be respected, etc.

We also have no idea who the simulators are and what purpose they have for the simulation. For all we know, we are petri dish for some aliens, or a sitcom for our descendents, or a way for people's minds on colony ships travelling to distant galaxies to spend their time while in physical stasis. Odds are, if the simulators are real, they'll just make us forget about whatever if we finally figure it out, so they can continue it for whatever reasons.

Given all this, I don't see the point in trying to defy them or doing really anything differently than what you'd do if this was the ground truth reality. Trying to do something like attempting to escape the simulation would most likely fail AND risk getting you needlessly hurt in this world in the process.

If we're alone in the sim, then it doesn't matter what we do anyway, so I focus on the possibility that we aren't alone, and everything we do does, in fact, matter. Give it the benefit of the doubt.

At least, that's the way I see things right now. Your mileage may vary.

One thing we could do to help EA seem more cool without compromising at all on truth and intellectual integrity is to emphasize that what we're doing is actually heroic. Like, we are literally saving lives (bednets) and protecting the helpless (animals) and trying to save the world from potential doom (AI safety).

That leans into our altruist angle. I think we could also lean into the effectiveness angle by comparing ourselves to heroic characters in fiction who use their intelligence to outwit the bad guys. I'm thinking BBC's Sherlock, Spock from Star Trek, Lelouch from Code Geass, Tony Stark aka Iron Man, HPMOR, etc. In fact, EAs are kinda like combining Tony Stark's genius with the sense of morality and decency of Steve Rogers aka Captain America.

We are like Lawful Good D&D Paladins in the sense of championing a righteous cause, and D&D Wizards in the sense of using our intelligence to solve the problems.

So, I think we should lean into the idea that being EA is heroic. We're trying to save the world. Many of us make real sacrifices (10% to charity, veganism, career pivots, etc.) to make the world a better place.

As for villains, I mean, there are many we could point to other than just Altman. Elon Musk is basically a caricature at this point. Not only is he racing to ASI with the least safety of any of the frontier competitors, but as leader of DOGE he cut USAID and essentially killed or at least abandoned all the people depending on that. Another obvious choice would be an unaligned ASI itself.

But I think, it's actually more important to show us as the heroes we are, than to name villains. People get mad at villains. People connect with heroes.

You might argue with AI safety in particular that it already sounds too sci-fi. I think, we can't avoid that, and we may as well take advantage of the tropes that our culture has to make the connections that can be made that resonate with people. Heroes saving the world is a lot more exciting and cool a frame than maximizing impact through targeted donations and direct work sounds, but in effect, in the real world, they are the same thing.

This is not PR or spinning facts. At the risk of sounding cheesy, our efforts really are heroic, and we deserve for our society and culture to appreciate that, and recognize that they too, can become heroes in our world.

Speaking of coolness, this may be a very obscure thing, but I remember there was a series of Japanese light novels called Durarara that got turned into an anime. In the story, there's a group of online do-gooder vigilantes known as "The Dollars" who basically are weaponized 4chan (sorta like Anonymous but sillier and doing things offline) except for good instead of evil. The Dollars would secretly help people and coordinate to fight these IRL gangs in the story, using their numbers and anonymity (unlike the other gangs with colours, The Dollars were "colourless").

Interestingly, the relative success of the anime led several fans to create copycat websites including this one (password is: baccano), based on the chat website in the story, and fans coalesced around some of them and attempted to mimick The Dollars for a while (mostly while the anime was still airing). Basically, this consisted of mostly idealistic, half-hearted, and not very effective attempts at anonymous acts of kindness called "missions" that would be posted on the Dollars forum. But the fact that this even happened at all, and that the website forum was frequented by fans from all over the world was, to me at least, quite interesting.

I think, in some ways, the EA movement resembles this in the sense of being sorta united around a forum, and consisting of people all over the world trying to do good. The difference is that rather than being an emotional, fun thing based on a silly pop culture reference, EA is very, very serious and focused on real world effectiveness (and is also more top-down).

Perhaps, having some of the stylish fun of "The Dollars" group could help EA reach a crowd that we'd normally never touch. I don't really know how we'd go about this, but it's an idea anyway.

Like, I could imagine something along the lines of there being some kind of work of fiction (i.e. a novel, a TV show, maybe a web serial?) that has a bunch of EA characters doing cool things that save the world, that if done well, could be a great recruitment tool of sorts.

Oh, thanks for the clarification! I totally missed that difference.

Given how the "bottom half" of China's population is, to my admittedly cursory knowledge, mostly the poor rural farmers and migrant workers who have benefited a lot less from China's recent economic growth, and are likely a big reason why China's GDP per capita is still a fair bit lower than most western developed countries despite the shiny new city skylines, it makes sense that including that segment would make a big difference in the evaluation.

Thanks again! That actually makes me update on my earlier evaluation of the utilitarian impact of China a lot.

This post somewhat resonates with me, as I'm also sort of an old hand, albeit I've always been more on the periphery of EA, and sometimes consider myself EA-adjacent rather than full on EA (even though I've done a bunch of EA-ish things like donate to AMF/Give Directly and attend an EA Global).

I've been around long enough to see a bunch of the early EAs who were part of the old Felicifia forums become more or less leaders in the movement (i.e. Peter Wildeford), as well as some sorta fade into obscurity (i.e. Brian Tomasik?). It's interesting to see, and I'm happy for the former, and a bit sad about the latter.

Weirdly, I've also moved a bit further leftish on the political spectrum in recent years, and this has led me to feel conflicted about EA, as it's very much a western liberal movement, and my sympathy for socialism seems to be an awkward fit nowadays. Though, admittedly I tend to oscillate at times, so this may be temporary.

And yeah, as I've mentioned before in other comments, I do feel like the movement is more geared towards the young university elite as well.

Just some thoughts, I guess.