A Suggested Reading Order for Less Wrong [2011]

jimrandomh8 Jul 2011 1:40 UTC

38 points

Less Wrong contains over four thousand posts. This is awesome. For newcomers, however, it’s quite intimidating. Rather than leave newcomers to figure out which posts are worth reading themselves, we provide some guidance, in the form of suggested reading orders. Previously, this has mainly meant the sequences, which are a list of posts that fit neatly into topics, sorted by topic. Unfortunately, reading in topic-sorted order is less than ideal, and many of Less Wrong’s best posts weren’t part of any sequence. Therefore, I have put together a suggested reading order: the hundred best posts on Less Wrong, in my purely subjective and unofficial judgment, arranged in a sensible order. Thanks to Student_UK for pointing out the need to reconsider how Less Wrong’s archived content is presented, to XiXiDu for creating a reading list and to Academian for another reading list; my reading list draws on (but is not a strict superset of) both these lists.

This is only a draft. Since the set of posts newcomers read has a significant impact on the community, I am soliciting feedback. After feedback has been collected, I will move (or clone) this list into the wiki, with feedback incorporated. Note that after my first pass, I had more than four hundred candidates for the top hundred; there were many excellent posts that I had to cut, and doubtless many more that I simply overlooked. I used karma as one consideration when deciding what to include, but this list is not karma based. Additional notes about what I did and didn’t include at the bottom.

And with all that out of the way, the top five Less Wrong posts that everyone should read:

Top 5
What Do We Mean By “Rationality” by Eliezer_Yudkowsky
Scientific Self-Help: The State of Our Knowledge by lukeprog
Cached Selves by AnnaSalamon and Steve_Rayhawk
Efficient Charity: Do Unto Others… by Yvain
Making Beliefs Pay Rent (in Anticipated Experiences) by Eliezer_Yudkowsky

A good pace is one or two posts per day, so that you have time to digest and internalize them. If you enjoyed the top five, take a moment to bookmark this page before continuing to the rest of the top 25. If you have questions or thoughts on a post, leave a comment; readers who follow the recent comments page will see it and may reply.

And that’s it for the top 100! If you’ve read up to this point, now is a good time to pause, reflect, and start incorporating the better parts into your life and your thinking.

And now, some notes about this list. I’ve tried to include a balanced mix of epistemic rationality, instrumental rationality, and motivations to learn about rationality, spread out through the list, except that the motivations are front-loaded. I have omitted some topics entirely; now that I’m done with the top 100, I will be starting on an “overflow list”, containing all the posts that deserved to be here but which didn’t fit in the top 100. This will include the Quantum Physics, Luminosity, Zombies, and Metaethics sequences, plus skipped portions of the main sequences and probably at least a hundred other miscellaneous posts.

I am looking for three types of feedback. First: point out if any posts are using concepts without sufficient introduction, so I can change the order to put the introduction in the right place. Second: if there are any posts in this list aren’t pulling their weight, that might be profitably replaced with the best of the overflow list, point them out. (Note that a post can meet this criterion and still be excellent.) And third: point out any posts that should be here but aren’t, especially posts that introduce concepts or jargon that occur frequently in comments.

What links here?

jimrandomh8 Jul 2011 1:40 UTC

38 points

10 comments4 min readLW link Archive

List of Links

Zed 8 Jul 2011 11:56 UTC
15 points
You list a number of great posts but the trouble with the sequences and most of LW is Death By Dependency. Nothing makes sense until everything else makes sense after which everything makes sense.

For that reason alone I think every post that links to related material every second sentence should be disqualified from the suggested reading list. People who tend to read in a depth-first manner (I think most people do?) will open the #1 post “What do we mean with Rationality” only to immediately start reading about “Your map and the territory” and will keep circling until they’ve read over 2 million words.

The core of the problem is that every post is entangled with every other post and this cannot be solved by creating a new post that itself refers to a 100 other posts. You cannot first explain that Curiosity is a virtue to people who are already curious by nature and then casually link to Newcombike problems (as is the case in reading list post #1) and not expect them to get stuck on Newcomb’s problem and TDT for the next week of their life.

Wikipedia has pages in English and pages in Simple English. Example: average and average simple. I think this is the approach we need to take. A separate Introduction to LessWrong sequence that can be read linearly and covers the subjects in a direct and accessible manner.

Feedback on the reading list itself:

“Absence of Evidence Is Evidence of Absence” is probably one of the strongest posts in the MATMQ sequence and I think it should replace or at least precede Conservation of Expected Evidence (that post simply doesn’t make sense even to people who are familiar with basic statistics).

A link to Luke’s Reading Yudkowsky series would probably also be a good addition.

Edit: I do think the list as it is now is tremendously valuable because it’s very easy to miss a few of the great posts even if you think you’ve read about everything.
- tetsuo55 9 Jul 2011 17:00 UTC
  1 point
  Parent
  I could not agree more.
  
  That has to be the biggest problem: The lack of teaching stuff on a concept by concept basis, and lack of building a pyramid of concepts from the ground up.
  
  khanacademy is on the right track in this regards, they try to teach everything in the smallest concepts possible.
  - Metus 12 Jul 2011 18:12 UTC
    0 points
    Parent
    Khanacademy is excellent since it has a tree diagramm of subjects on which you can see how far you have come and what material you need to learn in order to understand the topics. The big problem with LessWrong’s sequences is that they are highly entangled in contrast to the highly hierarchic way people usually learn. The other problem is that the useful or better articles came later in the sequences so that one has to read older, less interesting ones in order to understand the vocabulary of the newer articles. Maybe the sequences should be rewritten with this in mind?
Dr_Manhattan 8 Jul 2011 12:21 UTC
7 points
This assumes not reading things in sequence order, which I disadvise (from experience of doing that myself at the start). Sure, top 10 posts would be a useful enticer for people to check out more of the site, but if someone is reading 100 posts they would be better off slogging through the Sequences. The problem of going through the sequences is better served via Luke’s summarization project IMO.
lukeprog 8 Jul 2011 5:45 UTC
3 points
This is a great project. A few notes:

The new ‘About Less Wrong’ you link to was written by me, not Eliezer.

I think ‘How an Algorithm Feels from the Inside’, ‘Reductionism’ and ‘Taboo Your Words’ need to be higher.
- jimrandomh 12 Jul 2011 22:06 UTC
  0 points
  Parent
  Thanks for the feedback!
  
  The new ‘About Less Wrong’ you link to was written by me, not Eliezer.
  
  Fixed, sorry about that. (I scraped the authorship information from the Recent Posts feed, which also has it wrong, and didn’t notice.)
  
  I think ‘How an Algorithm Feels from the Inside’, ‘Reductionism’ and ‘Taboo Your Words’ need to be higher.
  
  I’ve moved these up to the top 25, pushing down The Fallacy of Gray, Practical Advice Backed by Deep Theories, and Mind Projection Fallacy.
skepsci 24 Jun 2013 14:40 UTC
2 points
After the top 5 or 10 or so, rather than just presenting a list of articles, it may make more sense to split things up by topic. Being presented with a list of 100 articles is kind of intimidating. Being presented with five lists of twenty articles each on five different topics is less so, as it’s easier to divide and conquer. Readers may be interested in some topics but not others (at least at first), or may decide to read a few articles on each topic.

Some natural subdivisions might be:
- Human psychology and biases.
- Bayesian inference.
- The scientific method.
- Quantum mechanics.
Also, others have mentioned sequences. Rather than just ignoring that sequences exist, it would be useful to new readers both to point out the sequences (which would happen more naturally if you broke things up by subject), and to provide the reader with some idea of how important it is to read a given sequence in order.
Manfred 8 Jul 2011 4:46 UTC
1 point
I would advise against simply pointing people to the Quantum Physics sequence. The first, most quantum-ey part is largely good. The other stuff is much less so.

AAanyhow, I like the list. It’s a good thing to point people to if they just want to explore.

EDIT: I shall start a Notable Standalone Posts (or something like that) sequence (Part 1) on the wiki soonish—seeing all these has made me realize that it’s tough to not lose good solo posts.
fburnaby 8 Jul 2011 1:48 UTC
1 point
Thanks so much for doing this, jimrandomh.

I’ve read some unknown portion of the sequences (~40 of them?), and many random articles on here (~100?) but I’ve never even tried to do it in some meaningful order. It looks like I’ve been missing out because of this; even one of your top 5 (Cached Selves) is new to me.
Armok_GoB 8 Jul 2011 18:11 UTC
0 points
Excellent resource, I’ll be linking this a lot!