What (standalone) LessWrong posts would you recommend to most EA community members?

I'm a big fan of some of the early LessWrong content, e.g.

• Confidence levels inside and outside an argument
• Humans are not automatically strategic
• Parapsychology: the control group for science
• Politics as charity
• Why the tails come apart

More generally, I'd recommend much of the content by Scott Alexander ("Yvain"), Paul Christiano, Wei Dai, Gwern, Greg Lewis ("Thrasymachus"), Anna Salamon and Carl Shulman (I'm probably forgetting other names).

Maybe the thing to do would be to start a low-commitment LW book club? There's so many old posts that it doesn't feel fresh to comment on them, but having a way to put some group attention on a couple posts at a time might help.

I think a lot of old Scott Alexander posts are quite accessible. Some top ones are: 

https://www.lesswrong.com/posts/gFMH3Cqw4XxwL69iy/eight-short-studies-on-excuses

https://www.lesswrong.com/posts/895quRDaK6gR2rM82/diseased-thinking-dissolving-questions-about-disease

Some other good classical posts: 

https://www.lesswrong.com/posts/aHaqgTNnFzD7NGLMx/reason-as-memetic-immune-disorder

https://www.lesswrong.com/posts/PBRWb2Em5SNeWYwwB/humans-are-not-automatically-strategic

https://www.lesswrong.com/posts/yA4gF5KrboK2m2Xu7/how-an-algorithm-feels-from-inside

Privileging the Question changed my life in college. I don't know how useful it would be for the average person already involved in EA, but it played a huge role in my not getting distracted by random issues and controversies, and instead focusing on big-picture problems that weren't as inherently interesting. I'd at least recommend it to new members of university EA groups, if not "most community members".

This got me to leave my girlfriend and has remained a permanent way that I think:

https://www.lesswrong.com/posts/627DZcvme7nLDrbZu/update-yourself-incrementally

I read it as part of all the sequences, no idea how helpful it will be to others or as a standalone post

My take is that LessWrong is best understood as a mix of individual voices, each with their own style and concerns. The approach I'd recommend is to select one writer whose voice you find compelling, and spend some time digesting their ideas. A common refrain is "read the sequences," but that's not where I started. I like John Wentworth's writing.

Alternatively, you might find yourself interested in a particular topic. LessWrong's tags can help you both find an interesting topic and locate relevant posts, though it's not super fine grained or comprehensive.

One of the key sources of value on LessWrong is that it provides a common language for some complex ideas, presented in a relatively fun and accessible format. The combination of all those ideas can elevate thinking, although it's no panacaea. My intuition is that it's best to slowly follow your curiosity over a period of a few years, rather than trying to digest the whole thing all at once, or pick a couple highlights.

Here are a few from Eliezer:

• Making Beliefs Pay Rent (in Anticipated Experiences)
• Politics is the Mind-Killer
• Money: The Unit of Caring
• Taboo Your Words
• Purchase Fuzzies and Utilons Separately
• Leave a Line of Retreat
• Dissolving the Question
• The Bottom Line
• Is That Your True Rejection
• Avoiding Your Belief's Real Weak Points

And a few from others:

• Reason as a Memetic Immune Disorder
• Is the potential astronomical waste in our universe too small to care about?
• Beyond astronomical waste

I might add more later.

I really like Ends Don't Justify Means (Among Humans) and think it's a bit underrated. (In that I don't hear people reference it much.)

I think I find the lesson generally useful: that in some cases it can be bad for me to "follow consequentialism," (because in some cases I'm an idiot) without consequentialism being itself bad.

The noncentral fallacy nicely categorizes a very common source of ethical disagreement in my experience.

[Edit:] Somewhat more niche, but considering how important AI risk is to many EAs, I'd also recommend Against GDP as a metric for timelines and takeoff speeds, for rebutting what is in my estimation a bizarrely common error in forecasting AI takeoff.

Saved these all to pocket, thanks for the recommendations!