100
What I’ve learned in my first two+ years on Substack
This is my hundredth Substack post.
In my defense, your honor, twenty-two of the hundred have been quotes posts (i.e., I’ve only written seventy-eight original posts).1 Still, I thought this would be a good opportunity for a Slate Star Codex–inspired round-up of what I’ve learned so far. At the end, I’ll also have some more general reflections on writing and the state of the blog.
Looking back, it seems to me that the central intellectual project of this newsletter thus far has been navigating my ambivalence about the limits of human rationality. I might not have expressed it in quite these terms in January 2022, when I first began writing, but the pattern is easy to trace in retrospect.
For example, early on, I planted my flag on the “believing game”—a direct challenge to rational skepticism—and I continue to find great epistemic and instrumental value in this vision of mature credulity. Over the past couple years, I’ve returned to this theme repeatedly, most notably in an updated, abridged version of that original post, and most recently in last month’s mini book review of Gilead. I’m increasingly frustrated by those who pooh-pooh the believing game—who, for example, resist basic perspective-taking, or fail to grasp that their argument could only be compelling to someone who already agrees with them. I believe that we are unreliable narrators of our own behavior, that we do ourselves a disservice by looking only at what we can measure, and that the simple explanation is often the correct one (or at least that, to a first approximation, the “smart” explanation is not reliably better).
In one way or another, these are all ways of saying that truth is subjective and ultimate rationality is not attainable.
At the same time, I don’t think we are totally in the dark about the nature of the universe; clearly humanity’s collective knowledge and ability to reason has increased significantly since the Stone Age. And on a personal level, I can’t help myself from trying to systemize fundamentally unknowable processes. To that end, in this newsletter I have…
argued in favor of economizing psychological resources for a happier life;
theorized about the idealized shape of the creative progress and optimal strategies for creative success;
made recommendations, drawn from the science of judgment and decision-making, for more effective revision of writing;
analyzed punctuation marks and paragraph structure in graphical, quasi-measurable terms;
and defended more rational approaches to everyday kindness and philanthropy writ large.
If pressed, I might concede that some of these posts are thought experiments or speculation; but the truth is, they are views I sincerely hold—and ones at least superficially grounded in empirical findings or a chain of reasoning.
Of course, having a rational model of how the world works does not mean that one will actually behave rationally in the heat of the moment. In this vein, I’ve argued at times for a compromise position between skepticism and credulity, rationality and intuition: The quirks of human nature mean that sometimes rational ends are best achieved through irrational (or, more often, a-rational) means.
I laid out this vision most thoroughly in “If It’s Worth Doing, It’s Worth Doing with Made-up Graphs”—in which I highlighted graphs as one of a small handful of tools by which the rational mind can speak directly to the intuitive mind. At other points, I’ve argued that it’s easier to act your way into a new way of thinking than think your way into a new way of acting; and that sometimes the best way to work smarter is through various dumb tricks and constraints. The unifying theme here is that while true rationality may not be possible, we can often construct situations or environments that nudge us closer to that ideal.2
Finally, while I can’t say I’ve found an answer to where the limits of human rationality lie (surprise…), I can say that my ambivalence has led to some of the creative pieces I’m proudest of: a personal essay grappling with the logic, or lack thereof, of moral development; a poem exploring the relationship between technological progress and psychospiritual wellbeing; a short story about the complex interplay between agency, causality, and the inexorable march of history killing baby Hitler. As always, mixed feelings yield greater artistic fruit than certainty.
If the driving beat of the blog has been ambivalence about rationality, then the refrain—the message I do feel confident in—is what one researcher terms “the mundanity of excellence.”3 (A more colloquial framing of the same idea: “To be great, be good repeatedly.”) Put simply, I believe that in almost any domain we care about, there is no trick per se; expertise is simply the confluence of many many sub-skills, acquired through lots of practice and iteration.
I’ve applied a version of this thesis to the three domains in which I have the most first-hand experience: writing; teaching; and operations. I’ve similarly argued toward treating happiness as an emergent property of a bunch of little things. More broadly, I’ve tried to spell out the practical implications of this thesis for those seeking to learn or instill new skills: Specifically, I think metaphors and hands-on experiences are crucial for building implicit knowledge; and that curricula ought to reveal vertical structure through horizontal exploration.
For my own part, I know that one of the greatest rewards of the past two+ years has been iterating enough to feel my own writing improve. Most of these improvements are vanishingly small on their own, but their cumulative effect is a writing process and product that are categorically better. To single out a few:
Concision — This manifests on a few levels.
In 2022, the mean length of my posts (excluding quotes posts) was about 2,300 words; in 2023, it was about 1,300. In part, this points to more economical prose, but it’s also a reflection of process—namely my desire to iterate faster: Shorter posts generally take less time to write, so I’m able to do more of them. (The data support this: I published forty-one posts in 2022 and forty-nine in 2023.)
On the line level, I’m more willing to break up sentences. For example an early post might have included syntax like Although snorks suffer from globification—which can lead to anxiety and paralysis—they are advantaged in the long-run by a societal bias toward pludge, now, on the margin, I’d be more likely to write: Snorks suffer from globification. (This can lead to anxiety and paralysis.) But snorks win in the long-run because society has a bias toward pludge.4
My titles have gotten punchier. For example, two early titles were “Gossip Is Good for Us—Not Just a Guilty Pleasure” and “Why (and How) You Should Create a Skimbox.” If I were rewriting these posts today, I would title them, respectively, “Gossip Is Good,” and “You Should Make a Skimbox.”
Patience — While the total time spent on any individual piece has tended to decrease, the total time from conception to publication has tended to increase. In other words, I let pieces incubate longer. The most extreme example of this is a portrait of my grandparents I’d originally written three years before publishing—and indeed well before I’d started writing on Substack. To date, this has been my one of my best received posts. However, even for more prosaic topics, I’m now much more likely to let drafts sit and wool-gather for many months—adding stray lines and examples as the mood strikes me until I’m finally ready to bang out a real draft. I wrote the original seed of this paragraph over a year ago, in January 2023.
Aesthetics — My posts have become more visual, I now have a background and accent color on my landing page, and I created a punctuation-suggestive logo for my quotes posts:
New Interests — I’ve become curious about a lot of new topics in the past couple years, to the point where I added a couple new tags/sections to the blog: one on technology/artificial intelligence, and one on operations/productivity. These aren’t topics that I’d planned to write a lot about, but I found I kept gravitating toward them and had a lot to say. Here’s the totals for all categories:
Fiction — 2 posts
Book Reviews — 3 posts
Miscellanea — 3 posts
State of the Blog — 4 posts
Personal Essays — 4 posts
Poetry — 6 posts
Technology/AI — 8 posts
Operations/Productivity — 10 posts
Education — 12 posts
Writing/creativity — 17 posts
Politics/culture/news — 21 posts
Quotes I Like — 22 posts
Psychology — 33 posts
The joy of new discoveries and small improvements is both one of the best reasons to write a blog and the most difficult to anticipate from the outset. My advice for anyone just starting out: Commit to a trial period of at least six months, and stick to it!5
Finally, a HUGE thank you to the friends, family, and internet strangers who have been so supportive through these first hundred posts. I’m grateful for your readership and hope you will get value from the next hundred as well <3.
-Gilad
My hundredth original post is projected for late summer or early fall 2024.
In Jonathan Haidt’s terminology: The rider is more likely to win by shaping the path than by trying to steer the elephant directly.
H/T to Brendan for sharing this paper with me.
I don’t think simple syntax is universally better; I just think that all else equal, it’s more online-friendly
Also, read Steve Yegge and Alexey Guzey’s posts on this topic!