The value of content density

post by Joey · 2022-07-20T15:34:55.442Z · EA · GW · 10 comments


  Why we should aim for higher content density
  Possible heuristics
  More information

Different communities have different norms when it comes to writing and content production. One stylistic difference the EA community has, is that forum posts are often comparatively long, deep and advanced. I really like aspects of this and I think that having a relatively high level of depth can be very nice to offset the other mediums of communication EA uses (e.g., facebook). However, I wish EA forum authors would give content density a little bit more consideration.

Different books have different levels of content density. Some have a strong concept on almost every page (a good example of this would be Ray Dalio's Principles), while other books could be summarized fairly easily into a single, short blog post (I think Carol Dweck’s Mindset falls into this category). There is often a strong social pressure to make something ~book length, so even if it's a bit concept-light it will get stretched to meet a word count. In the world of blog posts, no such limit exists but there are social norms that tend to affect the average duration. I think for entertainment blog posts there can be a lot of value in longer-form writing (e.g., Wait But Why’s writing), but for blog posts that are aiming to get a concept across (like this one) a leaner form of writing would be better. I do agree we should try to make writing entertaining [EA · GW], just not with huge length increases as a trade off. (A good example of both high content density + low wordcount). 

Why we should aim for higher content density

Time limited: Many EA forum readers are time limited and thus it is often a worthwhile trade to write a post that covers 90% of the value and is half as long. I often see blog posts that have a good concept or two, but due to how long they are they get far less engagement and readership than they would otherwise. 

Counterfactuals: There is quite a volume of posts on the EA forum and this directly ties into most readers' opportunity cost. Oftentimes the trade-off might literally be reading two EA forum posts vs one that is double the length.

Approachability: EAs are a fairly technical bunch and although the use of jargon has been written about, having a high volume of content with relatively sparse concepts also makes it harder for new people to get up to speed. It can make the forum as a whole seem more intimidating than it has to be. 

Possible heuristics

Length: The rule of thumb I use for blog posts is ~0.5-1 page per major concept and 1-3 pages per blog post. This kind of cap creates a strong force toward brevity and conveying concepts concisely. It does sometimes require major editing (the quote “I would have written you a shorter letter but did not have the time” comes to mind). 

TL;DR and summaries: I think the EA forum makes great use of these and they are worth including in almost every post that is over one page. 

Top concept leveraging: I think blog posts often try to cover a large number of topics that  could easily be broken into multiple posts. For example, this data on value drift [EA · GW] post being separate from this strategies for preventing value drift [EA · GW] post makes each piece of writing much more individually referable and digestible. 

More information

Some examples of content-dense writing include:

Further reading: Steven Pinker has some really well-articulated and pointed suggestions for community writing norms, many of which I think would apply directly to the EA forum.


Comments sorted by top scores.

comment by Chris Leong (casebash) · 2022-07-20T16:41:53.032Z · EA(p) · GW(p)

Footnotes are your friend here - they allow you to add detail for those who need or want it - whilst not wasting everyone else's times.

Replies from: vaidehi_agarwalla
comment by Vaidehi Agarwalla (vaidehi_agarwalla) · 2022-07-20T20:42:27.323Z · EA(p) · GW(p)

And key-takeaways at the top to let users skip to relevant sections.

comment by AndreFerretti (andreferretti) · 2022-07-21T13:30:20.679Z · EA(p) · GW(p)

TL;DR Write shorter posts.

comment by Ryan Beck · 2022-07-20T17:55:28.480Z · EA(p) · GW(p)

Thanks for writing this, as a new-ish user of the forum it's been frustrating trying to find previous posts that address questions I have or things I want to learn more about, only to find sprawling or multi-part posts with half hour or longer read-times that may or may not address the specific thing I'm interested in.

Also you mentioned jargon and I think there's room for a lot of improvement there, it seems to me like there's more jargon than is justified and it made the forum daunting for me. This previous post [EA · GW] has some good recommendations and in my opinion it would be valuable for more people to try to simplify their language where possible.

comment by Linch · 2022-07-21T02:01:26.061Z · EA(p) · GW(p)

Doesn't everybody skim these days? It seems maybe better for writers to do posts that facilitate skimming, rather than posts that are short/incomplete. 

Replies from: Stefan_Schubert
comment by Stefan_Schubert · 2022-07-21T07:52:51.446Z · EA(p) · GW(p)

My sense is that when you try to skim, you often miss points.

To the extent that one wants something like that, I'd rather employ features like abstracts, appendices, etc, which allow those who want to spend less time on a post to skip parts of it entirely.

Replies from: __nobody
comment by __nobody · 2022-08-05T14:41:35.633Z · EA(p) · GW(p)

An abstract or summary is a different text. If you want information that's plausibly in the long version but not in the abstract, then you're back where you started. If the text is structured to facilitate skimming, you can probably quickly find the stuff you want (and it likely wouldn't have needed the summary), but if it's more or less an opaque blob, then "decorating" it with a summary doesn't make it more usable. Adding large-scale structure doesn't reduce the importance of small-scale structure.

If there's several levels of headings, and maybe even highlighted keywords within the text to provide hierarchical structure (or at the very least some anchor points), that makes both skimming and general navigation much easier. This can go all the way to a fairly strict paragraph structure, where the first sentence makes some claim or observation, the following sentences elaborate on that, and the closing sentence of the paragraph wraps up what was said. (In some fields of academia, that seems to be the common style, in others they never even heard of it.) In such a text, you can just read/skim the first sentences of each paragraph and you'll end up with a pretty clear picture of the whole text. More generally, well-structured texts tend to make it easy to see what parts are less important and which parts you should read fully, even if you're in a hurry.

While not every text is compatible with strict structure, even just going through at the end and bolding key points can make it easier to navigate – no heavy editing required. (Sure, it'll look less pretty if there's 2-3 black blobs scattered about each page instead of everything being a perfectly uniform gray level, but it's useful and saves time.) Even the already pretty short post here could benefit: Highlighting e.g.

  • "There is often a strong social pressure to make something ~book length,"
  • "write a post that covers 90% of the value and is half as long"
  • "~0.5-1 page per major concept and 1-3 pages per blog post"
  • "easily be broken into multiple posts"

would mean someone skimming and reading just these short snippets would already get a rough idea of the argument, plus some mitigation strategies. (It may not be 90% of the value in 10% of the time, but it's probably around 60% of it.) If you need more details, you can just read on past the bold bits, or maybe jump up a few lines and read from there. (That's a harsh difference to a summary, where you first have to find the spot(s) in the main text that expand on the thing you want to know more about.) Intentional highlighting also gives the author control about what parts will most likely be picked up by someone skimming the text, at least partially mitigating the risk of missing important points.

comment by Grayden · 2022-07-29T09:05:23.840Z · EA(p) · GW(p)

Great post!

One nuance I would add is not to achieve density through jargon. As a long-term EA focusing most of my time on E2G, I’ve often found the Forum (and EA debate in general) to have a high barrier to entry. EA has its own language, which sometimes gives helpful precision, but can just make in inaccessible to those not spending 100% of their time in the EA universe.

I also think the concept of an elevator pitch is a useful one. You should be able to summarise your point in 30 seconds or less.

comment by Devon Fritz (DevonFritz) · 2022-07-21T07:46:20.040Z · EA(p) · GW(p)

Glad to see this and strongly agree. 

I think Paul Graham does this well in his essays, both in terms of 1) not using any extra words to communicate his point and 2) doing so in a jargon-free way. 

comment by fredzannarbor · 2022-07-21T05:47:41.174Z · EA(p) · GW(p)

Content density has been measured by various NLP methods. While I would not claim those methods are definitive or exhaustive, they are a starting point for independent benchmarking and self-regulation.

Brown, Cati, Tony Snodgrass, Susan J. Kemper, Ruth Herman, and Michael A. Covington. "Automatic measurement of propositional idea density from part-of-speech tagging." Behavior research methods 40, no. 2 (2008): 540-545.

Snowdon, David A., Susan J. Kemper, James A. Mortimer, Lydia H. Greiner, David R. Wekstein, and William R. Markesbery. "Linguistic ability in early life and cognitive function and Alzheimer's disease in late life: Findings from the Nun Study." Jama 275, no. 7 (1996): 528-532.