My guess is they have no will to do that because too many of them are seduced by the idea of "we have to build it first because we're the good guys". Plus good old fashioned financial incentives.

How do you tell the difference between an honest person, and an extremely good liar who is also much smarter than you? I don't know if you can.

Let’s imagine it’s November 2031. We’re both still here, AI is much more capable than it is today, but governments are governing, civilisation is still (mostly) civilised, and there has been no catastrophic loss of control. It just kind of didn’t happen.

Depends on what "much more capable" means. But on the sharp left turn worldview, or a worldview similar to AI 2027, this is what the world is expected to look like shortly before AI kills everyone. When misaligned AI is smart enough to be transformative but not yet powerful enough to reliably seize control, its best strategy is to pretend to be aligned until it has enough power to wipe out humanity with minimal opposition.

The point at which I'd be less worried is where AI is clearly powerful enough to kill everyone if it wants to, and it's had plenty of time to do that, but it hasn't.

I do find the AI-related discussion on LW to be much better than EAF. But the LW vibe isn't particularly altruistic and it's definitely not cause-neutral. EAF, even with all its flaws, remains the only home for discussion of cause-neutral altruism. Given that EAF is the only game in town in that respect, I don't want to give up on it; I want to make it better.

The link does not provide any argument or evidence that EAF is a trash fire? It's just a comment by Buck saying that EAF is a trash fire. This is also a very non-specific and unhelpful sort of criticism.

My attempt at interpreting "trash fire" is something like

• EAF regulars are insufficiently ASI-pilled, do not take misalignment risks seriously enough, and don't pay enough attention to discussions of AI risk
• EAF regulars are too deferent to EA leaders, who are importantly wrong about a ton of stuff (mostly AI-related stuff)
• EAF is not sufficiently oriented around truth-seeking
• People on EAF care too much about PR and power-seeking

Mostly these are criticisms I see people make of "EA", not the EA Forum specifically, so I'm trying to extrapolate.

I think that running things like the EA Forum prizes, contests, and other events usually significantly improve post quality. The Forum could go back to doing way more of this!

I have really enjoyed the debate weeks, and I think the forum would benefit from having them >3x more frequently.

Hmm, I think a factor here is that AI safety orgs have more of a culture of writing a lot, and a lot of them have grievances against the EA community which is why they don't post on EAF. I think their grievances are largely justified in the sense that they can correctly point to prominent EAs who have acted badly, but I disagree that "stop engaging with the EA community" is the correct response. EA is the only community that's dedicated to cause-neutrally doing good; if EA dies, there will be zero such communities. If you care about cause-neutrally doing good, then I think the move is to engage with the community and try to be a model EA.

This is a hard problem to fix because there's a network effect. If AI safety people mainly post on LessWrong and not EAF, then other AI safety people don't have much reason to post on EAF—although I'm not sure that's really true because cross-posting is easy.

There's also a question of why AI safety people write a lot. I don't know the answer to that one.

Agreed that it's worth trying to improve the EA Forum. Most of your suggestions are about how to screen out low-quality posts. What I really care about is how to get more high-quality posts and comments.

Why aren't there more high-quality posts?

• Writing high-quality posts is hard in general.
• Maybe people who used to write good posts are now doing other important work, and don't have time to write.
• Maybe some high-quality posts have moved off the EA Forum and onto other places.

Why aren't there more high-quality comments?

• Engaging with other people's work is hard, and usually you don't get any benefit from it.
• ? (LessWrong gets a lot of high-quality comments, even though writing comments is hard. How can EAF comments be more like LW?)

"Get people to write more good content" is sort of a hard problem in general—there is no "good content dial" that you can turn. One answer that comes to mind is that grantmakers should pay more people to write, and especially to write comments / engage with other people's work. I don't know if anyone in history has ever gotten a grant solely for the purpose of writing forum comments, but it seems like the sort of thing that's totally worth funding under the right circumstances.

Or there could be weekly comment prizes for the top 3 comments of the week, or something like that.

Inkhaven is an attempt at getting people to write more. Few Inkhaven residents post on EAF. Why is that?

I would accept this bet in theory, but there are two reasons why I don't want to:

1. When I've made bets before, it turned out that the headache of managing the bet and the worry of losing were not justified by the expected gains. The former is especially true when the bet is for only $20.
2. If the bet resolves YES and I have to pay out, then I expect my investments will have lost a lot of value. Given the basic theory of diversification (this bet is not perfectly correlated with my investments), there's still some level at which the bet makes sense to make, but it's a lot of work to figure out that point, and the expected payoff is probably not enough to justify the work.

In recent years I have become more bearish on bets in general, mainly for reason #1. I used to be very pro-betting but now I believe it's rarely worth the effort.

hmm it looks like the HTML is being treated as part of the text