Friday, September 18th, 2026
I didn't plan this week's theme. It assembled itself, which is usually how the good ones happen.
The book is about people who've spent their lives inside a structure and can't see out of it. The articles are about choosing your structure on purpose, refusing to choose one at all, and watching a rung quietly vanish from a structure nobody's looking at clearly. The podcast is a comedian arguing about conditions better than most business books manage. The show is about the container people spend forty years inside and mostly like.
Same question underneath all of it. What's the shape of the thing you're living in, who decided on it, and when did you last actually check?
Grab a drink. Let's get into it.
Book of the Week
I avoided this for years assuming it was dystopia by numbers. Grim world, plucky hero, evil council up top. I was wrong enough that I burned two weeks on it.
The setup is simple and the simplicity is the trick. What's left of humanity lives underground in a vast silo. There are rules, the rules keep people alive, and the strictest one is that you never speak about going outside. Break it and you're sent out, which is a death sentence dressed as an errand. When the sheriff commits exactly that sin, the job lands on Juliette, a mechanic from the down deep who's never met a machine she couldn't fix or a rule she wouldn't break.
That's the whole hook. A person whose entire training is taking things apart to see how they work, handed authority over a system specifically designed to not be taken apart.
Wool is the one people know, and it's the basis for the Apple TV show. But the reason to get the collection instead of the single novel is Shift, the second book, which is the one that reorganized how I think. Wool asks what it's like to live inside the silo. Shift asks who built it and what they believed they were doing. And the answer isn't a cackling villain. It's a series of reasonable people making defensible decisions under bad constraints, each one slightly worse than the last, until the accumulated result is monstrous and nobody involved can point to the moment it turned.
If you've ever inherited a business process and asked why it works this way, and traced it back through four people to a decision somebody made in 2019 for a reason that expired in 2021, you already know the emotional register of that book.
The line that follows me around is about how information is controlled. Not through lies, mostly. Through omission and framing. People in the silo aren't told falsehoods so much as they're given a view, and the view is the control. Everyone's operating on real data. The data is just curated.
For anybody running something, that's uncomfortably close to home. Your team has a view. Your clients have a view. You have a view, and yours is assembled from whatever surfaces reach you, which is not the same as what's true.
Fair warning. It's over seventeen hundred pages and the middle of Shift drags where it splits timelines. Push through, because the landing in Dust is one of the better ones in the genre. The collection also carries the author's essays and a chapbook of short fiction, and the essays are worth it on their own.
Read it for the story. Keep it for the question about who curates your view.
Get it here: The Silo Series Collection on Amazon
Three Articles Worth Your Time This Week
Article 1: Why the Smartest Solo Entrepreneurs of 2026 Are Staying Small
Source: Millennial Magazine
This one earns its spot by refusing to lie to you, which in this category is rarer than it should be.
The setup is the story you know. Somebody builds a small thing, it works, and the assumed next step is the office, the round, the headcount. The piece asks whether that arc is an obligation or just an option, and lands somewhere I've been circling for a year: you're allowed to build something small on purpose, and the permission was always yours to give.
Where it gets good is the data, and where it gets honest is what the data costs the argument.
The scale is real. The Census Bureau counts 30.4 million nonemployer businesses in the US, firms with receipts over a thousand dollars and no paid employees, together pulling roughly 1.8 trillion. That's not a niche. That's an economy.
Then it does the thing almost nobody does in this genre. It names the outliers as outliers. Yes, Pieter Levels reports around three million a year alone. Yes, Gumroad runs lean at a reported ten million in annual recurring revenue. And then: the average US nonemployer business earns closer to $47,800 a year.
Forty seven thousand eight hundred. Sitting in the same paragraph as the three million.
That's the whole piece in one juxtaposition, and it's why I'm sending it. Most solo business writing shows you the top of the distribution and lets you assume it's the middle. This one shows you both.
It also brings in the calm company idea, the 37signals lineage, which is the useful frame. The choice isn't between ambition and no ambition. It's between two definitions of it.
The part I'd push on: staying small is only a strategy if you've decided it. Most people aren't choosing small, they're stalled at small and telling a story about it afterward. The article is generous about that distinction. I'd be less generous. The difference between deliberate and stuck is whether you can name what you turned down.
Article 2: Keeping Your Options Open Is the Most Expensive Thing You Can Do
Source: Scott D. Clary
Direct counterweight to the last one, and the pairing is the point.
Clary opens with two years of hedging. He'd left the corporate career, the podcast was growing, the newsletter was working, the business looked like a real business. And he hadn't committed. Still taking consulting calls. Still saying yes to freelance that had nothing to do with what he was building. One foot in the old world because the new one wasn't guaranteed.
He told himself this was prudence. Diversified risk. Responsible. What a reasonable person does.
Then he does the arithmetic, and the arithmetic is the piece. The podcast got sixty percent of his attention. Consulting twenty. Freelance ten. The last ten went to anxiety about whether he was making the right call. Which means the thing he was supposedly building his future on was receiving barely more than half of him.
And, he writes, he wondered why it was growing slowly.
I sat with that longer than I wanted to. Because I've run that exact spreadsheet on myself and the numbers weren't better.
The turn wasn't a growth hack or a break. He killed the options. Stopped consulting, stopped freelancing, stopped hedging, decided this is the thing and if it fails it fails but it gets all of me. Six months later the podcast had doubled. Not because he got smarter. Because he stopped dividing.
Honest note: it goes behind a paywall shortly after he names the optionality trap. You get the story, the arithmetic and the turn for free, which is the part that does the work.
The reason this sits next to the staying small article is that they look like opposites and aren't. Both are arguing for choosing a container. One says the container can be smaller than you were told. The other says you have to actually get in it. Optionality feels like freedom and functions like a tax, collected in ten percent increments you never see itemized.
Article 3: AI and Entry-Level Jobs, What Hiring Data Really Shows
Source: IntuitionLabs
This is the most useful thing I've read on the subject and it's useful for an unfashionable reason. It refuses to give you a headline.
It's current as of this month, and instead of asking whether AI is taking entry level jobs, it asks a narrower one: what does the hiring data actually show, how is AI exposure defined and measured in the research, and which part of the decline can be causally separated from AI adoption with the evidence published so far.
That third clause is the whole thing. Most coverage skips it entirely.
The finding that reorganized my reading of the last year isn't about AI at all. It's about methodology. Payroll and job posting data, the Stanford and Indeed style sources, measure hiring that actually happened, and those consistently show contraction concentrated in junior, AI exposed roles. Forward looking executive surveys measure stated plans, and those skew more optimistic. Neither is wrong. They're measuring different things, and almost all the apparent disagreement between studies collapses once you know which type you're reading.
That's why you can find a credible report saying entry level hiring is collapsing and a credible Forbes piece from July saying new data says no, and both are defensible. One is counting what happened. The other is counting what people say they intend.
It also catalogs the counterexamples rather than burying them, which is how you know somebody's being straight with you. McKinsey publicly committing to grow non-partner staff and lift 2026 hiring, on the argument that AI doesn't substitute for the analytical work juniors actually do.
Why it belongs in this roundup: it's a case study in the silo problem from the book. Everybody's working from real data. The data is curated by the method, the method is invisible in the headline, and the view you end up with was decided by somebody upstream who never explained the decision.
If you employ anybody, or you're advising somebody young, read the methodology section and skip the rest.
Read it here: AI and Entry-Level Jobs, What Hiring Data Really Shows
Podcast of the Week
We're At The Beginning Of A Revolution, Jimmy Carr, Modern Wisdom #1135
Second Williamson episode I've sent in a row, which I noticed and am flagging so you know I noticed. I'll rotate next week. This one's worth the repeat.
Two hours with Jimmy Carr, recorded in August. If you only know him as the deadpan one liner guy from British panel shows, this will reset that. He's spent thirty years in a craft where you find out within a second and a half whether the thing worked, and people who live in that loop think unusually clearly about process.
Three things in it that I've been chewing on since.
The first is his reframe on difficulty. His argument isn't that your life should be easier. It's that it should be simpler. Those aren't the same request and I'd been making the wrong one for about a decade. Easier means less weight. Simpler means fewer things to carry at once. You can have a genuinely heavy life that's simple, and it'll feel better than a light one that's fragmented across nine obligations.
The second is a question he puts bluntly: are you actually depressed, or is your life depressing? He's careful about it, and not glib about clinical depression. The point is that people go looking for an internal explanation for a feeling that has an external and fixable cause, and spend years working on themselves when the situation was the thing that needed changing.
The third is the one I'd tattoo somewhere if I were the type. Favourable conditions never come. Not "conditions are usually bad," not "push through adversity." Never come. If your plan has a step in it where things settle down first, you don't have a plan, you have a wish with a timeline attached.
The section on optionality is what made me put the Clary article in the same week. Carr comes at it from the other end, arguing that having endless choices is itself a source of misery rather than the cure for it.
He's also good on using your unfair advantages, which people either overclaim or won't name at all, and on what to do when you catch yourself overthinking.
Listen here: We're At The Beginning Of A Revolution on Spotify
Track on Repeat
Notorious Thugs, The Notorious B.I.G. featuring Bone Thugs-n-Harmony, 2007 remaster
This has been on since roughly Tuesday and I want to be specific about why, because it isn't nostalgia.
The track is off Life After Death from 1997. The premise is that Biggie, whose whole identity as a rapper was unhurried, conversational, heavy on the beat, decided to get on a record with Bone Thugs-n-Harmony, whose entire style is the opposite. Rapid, melodic, triplet heavy, a flow so specific that for years almost nobody else could execute it convincingly.
And he does it. Not approximately. He gets on their beat, in their cadence, in their pocket, and he holds it.
What I keep thinking about is that he was already the best in the world at his own thing. No professional reason to do this. Nothing to prove and a lot to lose, because the failure mode is embarrassing and permanent. He did it anyway, and you can tell he did the work, because he doesn't sound like a guest imitating the hosts. He sounds like he learned the instrument.
The 2007 remaster is the version I'd point you at. The low end sits better and the vocal separation on the Bone verses is noticeably cleaner, which matters on a track where three people are moving that fast.
I put it on when I'm about to attempt something I'm not good at yet, which at the moment is most of what's on my calendar. It's a two and a half minute argument that being excellent at one thing is not a reason to refuse to be a beginner at another.
Listen here: Notorious Thugs on Spotify
Show of the Week
The Office, on Peacock
I know. Bear with me. I went back deliberately this month and found something I'd missed the first four times through.
I'd always filed it as a comfort watch. Background television. And it is that, which is most of why it's been streamed into the ground.
But watched with any attention, it's a show about a container. These people are inside a mid-size paper company in Scranton, an industry that is visibly dying around them across nine seasons, doing work that matters to almost nobody, in a building none of them chose for reasons most of them can't articulate. Nobody in that office is where they thought they'd be. Several of them know it.
And the show's actual argument, which it never states, is that the container turned out not to be the point. The people in it were. The container was arbitrary and the life inside it wasn't, and nine seasons is long enough that you feel that rather than being told it.
The other thing on rewatch: Michael Scott is the most accurate portrait of a certain kind of leader I've seen on television. Desperate to be liked, occasionally excellent at the actual job, catastrophic at anything requiring self-awareness. I've been all three on different Tuesdays.
If you're going back in, start at season two. Season one is four episodes of finding the tone and it puts people off.
Watch it here: The Office on Peacock
Resources Section
Track: Notorious Thugs, The Notorious B.I.G. featuring Bone Thugs-n-Harmony, 2007 remaster (Spotify)
Show: The Office (Peacock)
One step, one day. Grace over guilt. — Dan Kaufman
Tools I use and trust. If you sign up through these links I may earn a small commission at no extra cost to you. I only recommend things I actually use.