Read The Sequences As If They Were Written Today

post by Peter Berggren (peter-berggren) · 2025-01-02T02:51:36.537Z · LW · GW · 3 comments

Contents

  Key advantages of this approach
    1. It's easier to notice the contemporary applicability of the Sequences if you think of them as contemporary
    2. It lets you set aside misremembered ideas about what the Sequences actually say
    3. The Sequences were designed for today, not for when they were written
  A potential new series of posts
None
3 comments

If you've never read the LessWrong Sequences (which I read through the book-length compilation Rationality: From AI To Zombies), I suggest that you read the Sequences as if they were written today. Additionally, if you're thinking of rereading the Sequences, I suggest that your agenda for rereading, in addition to what it may already be, should be to read the Sequences as if they were written today.

To start, I'd like to take a moment to clarify what I mean. I don't mean "think about what you remember the Sequences talking about, and try to apply those concepts to current events." I don't even mean "read the Sequences and reflect on where the concepts are relevant to things that have happened since they were written." What I mean is that you should read the Sequences as if they were written today. You should imagine that, on January 1, 2025 (or whenever you happen to be reading this post), whatever post you're reading has just been released by some unknown 20-something-year-old alignment researcher, you saw it while scrolling down a bit on the front page of LessWrong, the title caught your eye, and you started reading it.

Key advantages of this approach

1. It's easier to notice the contemporary applicability of the Sequences if you think of them as contemporary

Many Sequences posts seem to be responding very directly to certain historical events within the rationalist community and Internet culture more broadly, at least if you approach them as a historical document. If you read them as if they were written today, however,  they will seem instead to be responding very directly to certain current events within the rationalist community and Internet culture more broadly.

Strictly speaking, I don't think either one of these understandings is accurate, especially given how heavily they were based on scholarly work from decades earlier (e.g. the work of Tversky and Kahneman) that was likely in the works for a long time before it was even published. However, it's natural to try and apply them to whenever they were written, given how much Eliezer Yudkowsky is idolized (however unintentionally) as an original thinker and how comparatively uncommon it is to fully account for the giants whose shoulders he stood on. 

While it may be intellectually interesting to see how Yudkowsky's work applies to the time in which it was written, there are a few key weaknesses to this approach.

1. It is easy to mistake the supposition that they were written for their time for truth.
2. The purpose that most people have in the modern day for reading the Sequences is to apply them to current events, not to apply them to events of the time.

In contrast, if one makes the supposition that they were written today, 

1. One knows that they weren't literally written today and so can more fully understand their applicability across a wider range of events than the specific ones in question, and
2. One can better apply them to current events rather than to events of the time.

To that end, setting aside the popular vision of Eliezer as a long-standing intellectual juggernaut is another purpose of reading the Sequences as if they were written today, and is one that Eliezer himself has endorsed on numerous occasions (including e.g. the "post-Sevarists" dialogue from Planecrash). This allows the ideas within them to stand on their own, apart from any positive or negative perception one may have of Eliezer, which leads into the second key advantage of reading the Sequences as if they were written today.

2. It lets you set aside misremembered ideas about what the Sequences actually say

Many concepts from the Sequences have entered into common usage within the rationalist community, and have been used to refer to things that the Sequences was not referring to. While this is not on its own a bad thing, it does mean that it's harder to read the Sequences without unnecessarily injecting this context. A few of the clearest examples are listed below, though I can point to countless others:

1. "Ethical Injunctions" is making a Kantian argument about certain patterns of behavior being inherently self-contradictory and thus impossible to consistently follow, not a rule-utilitarian argument about certain patterns of behavior causing bad outcomes if everyone were to do them.
2. "Fake Justification" can be seen as a criticism of the modern use of the concept of "ethical injunctions" to justify non-consequentialist intuitions using consequentialist-sounding language.
3. "Reversed Stupidity Is Not Intelligence" is not making a generic argument about the fallacy fallacy or guilt by association; it is making a very specific argument about the dangers that arise when debates become politicized.
4. "Cultish Countercultishness" does not argue that the drive to avoid cults is irrational. It argues that it is very rational on its own, but is often misdirected to target aesthetic signifiers of cult behavior rather than the core problems (e.g. outgroup hostility). And the title is far more than just a rhetorical flourish; it very literally argues that the drive to avoid cults is used by actual cults as a recruiting tactic.
5. "Why Our Kind Can't Cooperate" very directly calls for deliberate practice within the rationalist community on how to effectively disagree with other people, as well as how to effectively agree with them. 

When reading the Sequences as a historical document, it's hard to avoid injecting these ideas with their modern meanings and assuming that Yudkowsky defined them as they are used today. If one instead were to read the Sequences as if they were written today, one can instead simply suppose that Yudkowsky is re-defining these concepts for what he thinks are more instructive meanings, at least in the particular context in which he presents them.

3. The Sequences were designed for today, not for when they were written

I understand this is a pretty controversial assertion, but hear me out. Much of the Sequences, particularly the Sequence titled "The Craft and the Community," discuss the sort of problems that could befall a large organized rationalist community and how to prevent them. Additionally, many posts are written in a way that expects them to reach a very large number of people, far larger than the original audience. This means that when approaching the Sequences as a historical document, one is stuck between either seeing arrogance on Yudkowsky's part in assuming he'd become very popular (and by pure coincidence being right) or seeing an extreme degree of foresight instead. This further props up whatever pre-existing vision of Yudkowsky you have and prevents you, as discussed earlier, from evaluating these ideas on their own. If one were to instead read the Sequences as if they were written today, one doesn't have to think that way; in this framework, one can imagine that they were written to a large number of people because LessWrong is a large community and well-written posts, even by unknown authors, can go viral quite often.

A potential new series of posts

If this idea seems interesting, I'll probably be writing my own series of posts in the format of "Reading [Post from the Sequences] As If It Were Written Today." I already have at least a dozen post ideas lined up for that, and I'm not sure how frequent they'll be, but I expect this would be interesting.

I'm planning to take a very literal "as if it were written today" framing in these posts (e.g. talking about MIRI as if I don't have the benefit of hindsight but maybe nudging and winking a bit because I do, speculating on whether the early MIRI will be able to get Open Philanthropy funding or not even though that's anachronistic to what actually happened here, acting like posts are vagueposting about things that happened over a decade later) both to better get myself into the "as if it were written today" framing and to add small amounts of humor throughout the post. However, I'm also willing to abandon that framing if it becomes too grating or cumbersome.

3 comments

Comments sorted by top scores.