Writing On The Pareto Frontier
I have a personal rule: don’t write something which someone else has already written better.
This is easier than it sounds. For instance, suppose I’m writing an intro to systems biology. I don’t need it to be the most comprehensive intro ever written, or the most accessible intro. I just need it to be good enough on each of these axes that no other work is better on both at once.
In other words, I try to always write things which are the Pareto Best In The World.
Of course this generalizes to more dimensions as well: I might also care about writing something rigorous, or communicating intuitions, or making the piece enjoyable to read, or including good visualizations, or …. I don’t need to write the best piece in the world along any particular dimension. I just need to write a piece good enough on enough dimensions that nothing else beats it on every dimension which I care about.
Ways To Be On The Pareto Frontier
Novel Ideas
One natural way to be on the Pareto frontier is to write about a new idea, or at least an idea unusual enough that few people have written about it.
As with writing, new ideas are not necessarily that difficult to find. The trick is to combine things: finding a novel and interesting idea in complexity theory is a lot harder than finding a novel and interesting application of complexity theory to caricature art. On the LessWrong frontpage right now, there’s an interesting post about applying the idea of film study (i.e. athletes watching films of their games) to become better at research. I’ve been writing a whole series of posts intended to practice applying frames from math in unusual ways, and some interesting novel ideas have already come out of them—e.g. optimizing multiple imperfect filters.
Distillation
Research goes through a pipeline. First, researchers write papers, packed with jargon and assuming lots of background context. As results come together, they get boiled down into overview papers. Then come textbooks and programming libraries. Eventually, there might even be courses.
At each of those steps, work is done to distill the results—to explain them, draw analogies, add visuals, suggest exercises and applications, etc. In general, there’s a Pareto frontier with new cutting-edge results along one axis, and well-distilled results along the other. Pushing that Pareto frontier outward means finding some result which hasn’t been explained very well yet, understanding it oneself, and writing that explanation.
Note that Pareto optimality is again relevant to choosing examples/explanations: different examples will make sense to different people. Just offering very different examples from what others have written before is a good way to reach the Pareto frontier.
<Topic> For <Audience>
Probably the most successful statistics book in the first half of the twentieth century was Fisher’s Statistical Methods for Research Workers, which was essentially a repackaging of statistics for biologists. Glancing at my bookshelf, I see Basic Category Theory For Computer Scientists. COVID posts aimed at the rationality community have been a hot topic over the past year-and-a-half.
In general, a natural way to find a Pareto frontier is to pick a topic, pick an audience which doesn’t usually specialize in that topic, and write an intro to the topic for the audience.
Of course, this requires some background knowledge on both the topic and the audience—e.g. writing “Statistics for Biologists” requires background in both statistics and biology. So, it overlaps nicely with Being The Pareto Best In The World. The writing aspect also adds another dimension: it’s not just statistics and biology skills which are relevant, but writing skills as well. That means there’s three axes along which our skill could be Pareto optimal—and the more axes, the more “elbow room” we have on the Pareto frontier.
Context & Background
One question I get all the time about my work is “Why is <particular result> interesting?”; I sometimes write posts which give more context, but often I just write up results.
Some writing optimizes for communicating an idea clearly. Other writing optimizes for explaining why the idea is interesting/useful, or where it came from, rather than explaining the idea itself. These are both useful, so they’re both axes relevant to Pareto optimality of writing.
Another example: if you want to know why high-school calculus is interesting, then a physics class or a history book (like History Of √-1) will give more context than studying calculus itself. (In fact, this was how I first picked up calculus in high school—the actual calculus course came a few months later.)
Fun Writing
Finally, obviously, writing can be optimized to be fun to read. Embedding interesting ideas in fiction, for instance, is one cool way to reach a Pareto frontier. Just writing fun-to-read nonfiction is also great—whether it’s James Mickens’ rants on computer security, or visual proofs from Euclid’s elements, or drawings of the internals of everyday objects with cartoon mammoths.
Takeaway
All of these different dimensions—novel ideas, distillation, targeting, context/background, fun writing, and plenty of others which I didn’t get to here—are different things one might want from a piece of writing. They are all different ways to be on the Pareto frontier, and any combination of them is also a way to be on the Pareto frontier.
To achieve Pareto optimality, a piece of writing does not need to have completely novel ideas or amazing distillation or be super fun to read or …. It just needs to be good enough on enough of those axes that no other piece of writing is better on all of them.
I think that this is an important point. Personally, I didn’t realize it until I read Non-Expert Explanation.
The way I think about it, the clarity of an explanation is a 2-place word. It doesn’t make sense to say that an explanation is clear. You have to say that it is clear to Alice. Or clear to Bob. What is clear to one person might not be clear to another person.
In the language of pareto frontiers, I suppose you could say that one axis is “clearness to Alice” and another “clearness to Bob”, etc. And even if you do poorly on the other axes of “clearness to Carol”, “clearness to Dave”, etc., it could still be a pareto frontier if you can’t do better along eg. “clearness to Carol” without trading off how well you’re doing on eg. “clearness to Alice”. There’s no opportunity to do better along one axis without doing worse along another. You wrote the best article out there that targets Alice and Bob.
All of this is of course related to what was said in the <Topic> For <Audience> section as well.
It may also be worth noting that being on the pareto frontier doesn’t necessarily make it a good post. Eg if you write a post that is incredibly good at explaining calculus to John Doe, but terrible at explaining it to everyone else in the world, and John Doe has no interest in calculus, that post would be at the pareto frontier, but would also be silly to write.
Also, if John Doe is interested in calculus, but never finds your post, then it will also be a silly post to write. In general, the ability of writing to produce value is bottlenecked by our ability to get the right piece of writing to the right person at the right time.
It’s also relevant to worry about externalities and information asymmetries.
Persistent frustrations with social media originate from posts that are at the Pareto frontier, having traded a lot of nuance and accuracy off in exchange for fun and signaling. Such posts do this because the writer gets more clicks and shares by writing posts like this. This is “good” for the individual readers and sharers in the moment, if we believe their behavior reflects their preferences, but it may be bad for society as a whole if we’d prefer our friends to focus more on accuracy and nuance.
Readers may use signals of credibility when they pursue nuance and accuracy in order to judge the accuracy of a text. They optimize, therefore, for credibility, because they can’t directly optimize for accuracy. Perhaps they also want accessibility. If you then write a post optimized for credibility and accessibility, but the post isn’t accurate, then you can be at the Pareto frontier while also doing the reader a disservice.
That being said, the basic concept here seems right to me. Being at the Pareto frontier is correlated with creating value for the reader, and a search for such correlates of value is helpful.
In Being The Pareto Best In The World you mention the problem of elbow room:
I notice the converse of a multi-dimensional skillset is multi-dimensional assessment. In the same way it is hard to hire good programmers without knowing anything about programming, it will be hard for anyone else to assess a pareto-optimal product or skillset along multiple dimensions simultaneously.
It seems to me this challenge is pareto legibility. The more dimensions on the frontier, the noisier the assessment will necessarily be. This introduces a meta-problem where one of the skills on which you want to get good-enough is making your pareto frontier position legible enough for others to benefit from it.
As a practical matter this doesn’t seem like that big a deal for consumer goods like books, where even laypeople can take reviews of “X about this book was so good” and “I liked Y about this book” and round this off into a feeling of “muchly good.” By contrast, legibility seems exceptionally important for something like the econometric modeling applied to proteomics example.
Alternative title: Being The Pareto Best In The World for Writers.
The same thing is called Talent Stacking by Scott Adams. Your explanation is higher on the theoretical dimension vs. the usual Talent Stack explanations you see e.g. on YouTube are higher on the relatable/engagement dimension.
I notice that you didn’t give any arguments, in this post, for why people should try to write on the pareto frontier. Intuitively it seems like something that you should sometimes aim for, and sometimes not. But it seems like you find it a more compelling goal than I do, if you’ve made it a rule for yourself. Mind explaining briefly what the main reasons you think this is a good goal are?
Also, do you intend this to apply to fiction as well as nonfiction?
Good question! Rough argument: if someone else has already written it better, then do your readers a favor and promote that to them instead.
Obviously this is an imperfect argument—for instance, writing is a costly signal that you consider a topic important, and it’s also a way to clarify your own thoughts or promote your own brand. So Pareto optimality isn’t necessarily relevant to things I’m writing for my own benefit (as opposed to readers’), and it’s not relevant when the writing is mostly a costly signal of importance aimed at my social circle. Also, even if we accept the argument, then Pareto optimality is only a necessary condition for net value, not a sufficient condition; plenty of things are on some Pareto frontier but still not worth reading for anyone.
Not exactly a counterpoint to anything, but I am reminded: there’s a bunch of times Scott Alexander has said “please, read the original thing! it’s very good!” and then goes on to summarize it and discuss it and each time I’m like “ugh I’d rather just read Scott’s take on this thing than This Thing”.
I don’t super endorse this.
I mean, the argument does kinda rely on someone else having written it better, which does not often happen when “better” is comparing to Scott.
Robert Heaton calls this (or a similar enough idea) the Made-Up-Award Principle.
Wholeheartedly agree. My own frustrations with writing come when I forget that point, and try to be the best in each and every dimension. Thinking of it as a Pareto Frontier is a good mental tool to debug this kind of mindset when it arises.
Tim Chow calls such results “open exposition problems”, which I quite like.
I have two thoughts related to this:
First, there’s a dual problem: Given a piece of writing that’s along the Pareto frontier, how do you make it easy for readers who might have a utility function aligned with the piece to find it.
Related to this, for many people and many pieces of writing, a large part of the utility they get is from comments. I think this leads to dynamics where a piece where the writing that’s less optimal can get popular and then get to a point on the frontier that’s hard to beat.