Notes on SubStack These Days
The Substack Paradox: community, control, the myth of lock-in, golden handcuffs, platform power, writer agency, exit, & voice…
The Substack Paradox: community, control, the myth of lock-in, golden handcuffs, platform power, writer agency, exit, & voice…
I went to the SubStack Party at Shotwell’s Saloon <https://www.shotwellsbar.com/> in San Francisco’s Mission neighborhood <https://missionstreetsf.com/> last week. SubStack CEO Chris Best <https://substack.com/@cb > gave a little talk. I took notes. He said the right things.
I then took my notes and fed them into the skill-prompt I call “thucydides-speech” I have on my Dia Browser <http://diabrowser.com> installation. This is what came out:
SubTuringChrisBestBot: What is the real engine of our shared endeavor? It is tempting, in an age of digital marvels, to attribute success to the cleverness of algorithms, the robustness of code, or the magnitude of capital investment. But such attributions, while comforting to the technocrat and the financier, miss the mark.
The true foundation of Substack’s vitality is not the machinery humming in the background, but the active engagement, criticism, and care of its users.
Every bug report—no matter how minor—every complaint about statistics that refuse to align, every suggestion for improvement, these are not mere irritants, nor are they distractions from the “real” work of platform-building. On the contrary, they are the lifeblood of the collective project. Each is a signal, a pulse, a demonstration that the community is alive and alert, unwilling to accept mediocrity, and eager to shape the environment in which it dwells.
I say this is not mere rhetoric. The platform’s vitality depends, in the most literal sense, on the willingness of its writers and readers to invest not just their time, but their energy, and—dare I say it—their emotion, in making it better. If the users were to withdraw their attention, to resign themselves to passivity, the site would become a mausoleum of abandoned newsletters and silent comment sections. It is the ongoing, sometimes unruly, sometimes passionate, sometimes exasperated participation that keeps the platform alive.
It is easy, in the age of scale and automation, to forget this. The myth persists that progress is a function of size, that the larger the network, the more inevitable the improvement. But the history of human endeavors—whether in Athens or in San Francisco—teaches us otherwise. The real engine of progress is the community itself: the people who show up, who argue, who demand better, who offer their labor and their criticism in equal measure.
Substack’s achievements—such as they are—derive not from a passive audience, but from a citizenry of active participants. You are not mere consumers, here to be fed content by invisible hands. You are the authors of posts, the sources of critique, the shapers of the ecosystem with every comment, every suggestion, every act of engagement. The platform’s team is acutely aware of this; their appreciation is not a marketing ploy, but a recognition of dependence.
Let us not deceive ourselves: the economics of digital platforms may be driven by network effects and economies of scale. These are the iron laws of the marketplace. But the soul of any platform—its animating principle—resides in the feedback loops of care, critique, and collaboration. Without these, there is only stagnation.
The work is ongoing. Let us, then, continue as we have begun: active, critical, invested—not because it is easy, but because it is necessary for the health of the community, and for the promise of the project we have chosen to build together…
And this is the “thucydides-speech” skill prompt:
please take the highlighted bullet point notes and expand them into a talk, following the procedure of Thucydides: Since it is difficult to remember the exact words that were spoken, either those which one heard myself or those which were reported at second hand, the idea is to make the speakers say what, in your opinion, was called for by each situation, while keeping as closely as possible to the general sense of what was actually said.
What do I think of this?
First, register this: These are not Chris Best’s words. These are not Chris Best’s thoughts. These are not, really, thoughts at all. These are a remix of sentence and paragraph fragments constructed by a new cultural summarization technology that presents you with a blurry jpeg of a fragment of the text of the internet. To the extent that there are ideas and thoughts behind these, they are the thoughts of the people who wrote the words on which the models have been trained, perhaps in a way similar to the way in which the motions of the character Gollum in the Lord of the Rings movies are distorted reflections of the underlying motions of the actor Andy Serkis.
Ascribe these to an entity of some sort I will call SubTuringChrisBestBot.
And, well, it really does not sound much like Chris Best. It sounds more like Perikles of the Athenai, if he worked for a SOMA San Francisco startup today. (That is not necessarily bad.)
And it does, I think, accurately convey the vibe I got from the event: that the writers on SubStack who showed up and the staff of SubStack who showed up are seriously invested in the project, as a way of pushing back against clickbait social-media shouting-class rage-engagement slop, and pushing for a rational, thoughtful infosphere of accurate information, rational discourse, and public reason. All are damned certain that they are not just hoping to do well but actively doing good.
So I left more encouraged and hopeful about SubStack and its future than I had been before. Assibayah it has.
Some More Stray Thoughts Gleaned from Conversations at the Event:
Which may or may not be based in reality:
Normalizing the “tip jar” is the key difference between SubStack and earlier weblogging efforts: that is the reason that one might hope that the SubStack information ecology will be stable and expanding long-term, while the weblogging information ecology was not.
Front-line workers for SubStack have no idea what SubStack is going to do with its new money other then bank it and so extend the runway.
But they do say that the newsletter-distribution-and-weblog-website-with-an-easily-tuneable-tip-jar core of the system makes figuring out what other surfaces they should work on a complex problem, and when they have to solve because they cannot lose their core identity without losing their essential mindshare edge as the central place to run a weblog in the 2020s.
The problem of finding ways so that potential audience members can discover what SubStacks would enrich their lives is the central, largely unsolved one.
In particular, adding any discovery layer within Substack that is more than rely on individual writer shout outs runs, very quickly, that any attempt to cast a broad net winds up, suggesting that people read either (a) writers who are very eager, for what seem like minor failures to toe some line, to drive others out in the wilderness to be food for Azazel; or (b) writers who are very eager to classify others as a blight on humanity who need to be removed from the gene pool, humanely (or not) not; or (c) both.
The “Nazi Bar” problem is worse than the “cancellation two-minute hate problem”, but both emerge remarkably quickly whenever you attempt a wide-net discovery layer that is anything other then a defensive crouch.
The problem of figuring out how to balance capturing enough of the value-added for the true fans that writers can eat without destroying the usefulness of the SubStack for the non-true fans is the second, still largely unsolved one.
And without free-subscriber non-true fans to serve as a SubStack’s marketing department, the paid subscriber numbers will be astronomically unlikely to look good in the long run.
The large free audience is not just an amplification play, but essential as the ground in which a healthy crop of paying subscribers can grow..
The solutions? The line is that it lies in a combination of improved platform design, smarter algorithms, and the active cultivation of networks to connect the right readers with the right writers. But the confessed reality is that the strategy is to hire 100 people (so far), cook huge amounts of spaghetti, and throw it against the wall in clumps to see what sticks.
Workers for SubStack are very curious about academic publishing—a system with a very broken business model but one in which there are a large group of people who are institutionally driven to write frantically, and then find some kind of audience.
Is SubStack a way to shift from traditional academic publishing to the open web, fulfilling the university’s mission?
And nearly everyone at the event is firmly committed to the creation of a rational, accurate information- and analysis-full public sphere.
The perennial questions about whether a 10% revenue cut can ultimately support a profitable business model.
The belief that a company with only $45 million in current recurring annual revenue—and spending $20 million a year or more on its people—has any business raising $100 million, no matter how much of that is actually not just a 9% upside equity share but super-senior debt as well.
A general belief that BOND, The Chernin Group, Andreessen Horowitz, Rich Paul, Jens Grede, Mood Rowghani, and the others are to some degree using this not as an investment so much as a public-relations reputation-washing effort, given some of the other things they have been doing.
Lots of curiosity about Mood Rowghani.
Substack’s current structure is fundamentally a one-to-many broadcast model, centered on essays and newsletters. There is growing demand for more interactive, community-driven formats—such as live video and richer discussions—that existing technology does not yet fully support. And that raises all of the issues of para-sociality as applied to (very minor degrees of) celebrity.
Navigating and surfacing meaningful content amid near-infinite digital “noise” is a persistent and growing challenge in the online information ecosystem, only growing larger with the tsunami of AI-slop now bearing down on us.
I have nor felt much response to the suggestion (not mine alone) that authors be permitted to sell single posts or small packages of posts. The problem with the subscription only method is that comments can come only from those that pay which means they will be biased toward some measure of agreement with the author's POV.
Best of luck to the Substack team --a solution would have positive externalities as a useful social technology-- but I suspect from experience and first principles that there's no healthy algorithmic discovery mechanism without strong, ideologically opinionated manual intervention, essentially because:
(1) Prosocial truth and poisonous bullshit can't be told apart through purely algorithmic means, so there's no algorithmic filtering out of the latter.
(2) Poisonous bullshit has more degrees of freedom than prosocial truth (ethical and epistemological constraints don't apply) so they have an inherent advantage in seeking and gaining algorithmic favor.
So far every large scale centralized platform that has attempted to figure this out has first failed and then given up. From what you say, there's no specific reason to think Substack will have better luck [happy to be proven wrong here] but also it doesn't look like they will limit themselves to the passive role they have figured out but it's not ambitious enough for who they are.
I do agree Substack has a positive replacement value over personal blogs when it comes to monetization (which can in turn help with long-term sustainability) but my cynical bet is that it'll fall into the same failure modes as every other large-enough centralized platform who has tried to implement discovery mechanisms, feeds, and so on.