So I Got Back Into Reading
I started the year with a reading goal of 12 books because I wanted to become someone who reads again. One a month felt realistic. I didn’t have to know exactly how it would fit into real life.
Once it was obvious I would clear 12, I escalated the goal to 26. One every two weeks felt like a useful constraint without turning a hobby into a performance review.
After reading Real-World Kanban, I started tracking my reading on a kanban board in Obsidian. Nothing fancy, just a way to see what was next and what was stuck. When a book sat in the same column for three weeks, I either picked it back up or admitted I was done with it.
The Mixed Bookshelf Experiment
At some point I was reading three books at once: one fiction, one non-fiction, and one “technical.”
“Technical” is doing some work here. I mean anything related to the technology world I work in, from kernel internals to the social dynamics around who builds what and why.
Balanced reading was never the plan. Some books were just harder to get through, and I didn’t want to skip them or turn every reading session into a test of discipline.
The three-track setup solved that in a way I didn’t expect. Fiction gave me somewhere to go when I was tired. Non-fiction kept me honest about the world outside my own head. Technical books reminded me I still like my field, and occasionally that I needed to step away from it.
Chapter Switching
I rotated by chapter rather than by book.
One chapter of fiction. Next session, one chapter of non-fiction. Next, one chapter of technical. Then back around. Sometimes I read two fiction chapters in a row if I was tired or if the book had momentum. Sometimes I stalled on a technical chapter and tried to negotiate my way through it. But the rule was simple: never let one book monopolize the time.
This started as a coping mechanism. I hit a dense section, lost momentum, and picked up something else “just for now.” Eventually “just for now” was the only reason I kept moving at all, so I stopped treating it as accidental and made it the system.
With three books underway, I was always resuming rather than starting cold. One difficult chapter no longer made reading feel like an obligation.
What Worked, What Didn’t
The rotation kept me reading. I stopped waiting for the perfect mood to read a hard chapter. I read a chapter, any chapter, and trusted I could switch to something more forgiving next time.
Dense books became survivable. I didn’t have to face them in long, unbroken stretches, so I finished material I normally abandon with good intentions and vague guilt. The FreeBSD book is the clearest case. I first put it down about ten years ago, and a chapter at a time finally got me through it.
I enjoyed fiction more once it became the easy lane. I stopped asking it to be profound or instructive or secretly good for me.
Each category asked for a different kind of attention. Technical books demanded precision. Non-fiction kept me focused on incentives. Fiction reminded me that people do not behave like tidy models.
The downsides were real. Some books want immersion, and chapter switching works against that. When a book’s power is cumulative, breaking it up can flatten the emotional arc or dull the atmosphere. I also lost track of which book I was in and had to stop and ask whether I was reading about institutional trust, institutional collapse, or institutional collapse driven by incentives in technology. The answer was often all three.
Big emotional moments sometimes landed softer than they should have. It’s hard to sit with the aftermath of a powerful scene when the next chapter is about process design or the political economy of a complex system.
The trade was worth it because I stopped treating difficulty as a reason to quit. A hard chapter just meant the next one would be easier.
The Themes That Kept Showing Up Anyway
In hindsight, the categories blur. I kept circling the same questions from different angles, and the label on the book mattered less than I expected.
The technology books kept returning to power: who has it, how they explain it to themselves, and what happens when systems grow large enough that responsibility becomes abstract. Careless People, Empire of AI, and Extremely Hardcore are all nominally about technology companies. They’re really about governance, incentives, and what people do when they believe their decisions are inevitable rather than chosen.
The books also kept asking how trust is built, how it erodes, and how often it gets confused with compliance or silence. The Thin Book of Trust takes it on directly. On Tyranny and A Paradise Built in Hell come at it from opposite ends, one on how institutions lose it and one on what people do for each other when institutions fail. Once I started noticing it, I saw the same question in books about workplaces, governments, disasters, and technology companies.
Meaning under constraint was the third thread. Four Thousand Weeks makes the argument directly: you get a finite number of weeks, and no amount of optimization changes the number. Empire of AI and The Scaling Era approach the same tension from opposite directions: the costs of growth and the promise of more scaling.
It was strange to see the same three questions surface in a novel about octopuses and a memoir about Facebook in the same month.
The Books, in the Order I Finished Them
I’m listing these chronologically, with the category I used while reading. The notes are honest, but I’m not here to litigate novels.
A Philosophy of Software Design (John Ousterhout) #technical
Ousterhout’s case that complexity is the central problem in software design, and that most of it is self-inflicted.
I started reading this because several colleagues were reading it at the same time. I agreed with its focus on complexity, though parts of it changed how I thought about module boundaries.
On Tyranny (Timothy Snyder) #non-fiction
Twenty lessons on resisting authoritarianism, drawn from 20th-century history.
I read this as part of a book club that, like many book clubs, had exactly one meeting. What struck me most was the timing. The book focuses on the first Trump administration, but I was reading it during the early days of the second. It read like a replay.
Real-World Kanban (Mattias Skarin) #technical
Four case studies of teams adopting Kanban, with the practical details most process books skip.
I already knew Kanban, but the case studies got me to set up a board for my reading. I used it for the rest of the year.
Wool (Hugh Howey) #fiction
Post-apocalyptic sci-fi about humanity surviving in an underground silo, forbidden from knowing what’s outside.
This was a book I had started and abandoned more than once. I picked it back up after watching the Silo TV series on Apple TV. I’m glad I finally finished it. It was an easy world to settle into.
All Systems Red (Martha Wells) #fiction
A socially anxious security robot hacks its own governor module and mostly wants to be left alone to watch TV.
I picked this up because of the Murderbot adaptation and finished it quickly. I ended up reading several books in the series and still have more to come back to.
Punishment-Free Parenting (Jon Fogel) #non-fiction
Parenting built on connection and understanding rather than discipline and consequences.
I found this through TikTok, which was unexpected. My son has ADHD and is hyperactive, and his challenging behaviors feel familiar for obvious reasons. The book gave me language for the gentler approach I was already trying to follow.
Artificial Condition (Martha Wells) #fiction
The second Murderbot novella, featuring an unlikely partnership with a research transport ship.
I kept going with Murderbot. It was exactly the fiction I wanted between heavier books.
Design and Implementation of the FreeBSD Operating System (Marshall Kirk McKusick, George V. Neville-Neil) #technical
Everything about the FreeBSD kernel, from process management to the network stack, in considerable detail.
I first picked this up about ten years ago, put it down, and then restarted it recently. It is dense and demanding. I’m glad I finished it, though I respected it more than I enjoyed it.
Rogue Protocol (Martha Wells) #fiction
The third Murderbot novella. A decommissioned terraforming facility and the corporate espionage around it.
Still effective as a pressure-release valve.
Careless People (Sarah Wynn-Williams) #non-fiction
A memoir of working at Meta through a run of global crises, from someone who was in the room.
In my work I heavily overlap with trust and safety, which made the details about Meta’s influence on behavior uncomfortably concrete.
Project Hail Mary (Andy Weir) #fiction
An astronaut wakes up alone on a spaceship with no memory and a planet to save.
I owned this for years before finally reading it. The movie trailer spoiled a major reveal, but I still tore through it.
Ministry of Time (Kaliane Bradley) #fiction
A government time-travel program pairs civil servants with “expats” pulled from the past. Mostly a slow-burn romance.
A friend recommended it. It took a little time to settle into, but I enjoyed having it alongside heavier non-fiction.
Orbital (Samantha Harvey) #fiction
One day aboard the International Space Station, told as literary fiction.
This was the closest I came to not finishing a book this year. It didn’t click for me. It felt more descriptive than cumulative.
In This Economy? (Kyla Scanlon) #non-fiction
An accessible explainer on inflation, economic anxiety, and how the economy actually works.
Another TikTok find that paid off. It helped contextualize a lot of ambient economic anxiety without talking down to the reader.
Cybersecurity Myths and Misconceptions (Eugene H. Spafford, Leigh Metcalf, Josiah Dykstra) #technical
Debunks the security fallacies that persist anyway, and offers better mental models for thinking about risk.
Much of this overlaps with my day-to-day work, but I still came away with examples I’ll reuse.
Working in Public (Nadia Eghbal) #technical
How open source actually gets maintained, and the economics of maintainer labor.
I hoped for a broader examination of open source, but it focused heavily on GitHub-centric models.
The Scaling Era (Dwarkesh Patel) #technical
Interviews on AI scaling, where the progress is coming from, and what it implies.
I preordered this because I like the Dwarkesh podcast. The interview format worked well with the rotation because I could read one conversation at a time.
Empire of AI (Karen Hao) #non-fiction
An investigation into AI’s global footprint: labor exploitation, geopolitics, and the human costs behind the technology.
I sought this out to balance the more optimistic AI books on my list.
Abundance (Ezra Klein, Derek Thompson) #non-fiction
A policy argument for removing the barriers that block housing, energy infrastructure, and other things we need built.
I don’t agree with everything, but it helped me understand a policy argument that will keep showing up in left politics.
The Thin Book of Trust (Charles Feltman) #non-fiction
Four behaviors that build or break trust at work, in about eighty pages.
My company’s leadership team read this together. It gave us a shared vocabulary for discussing trust.
The Mountain in the Sea (Ray Nayler) #fiction
An octopus species develops intelligence, and the scientists and corporations race to understand or exploit it.
One of my favorite fiction reads of the year. It put the AI and consciousness questions from my non-fiction into a great story about octopuses.
Four Thousand Weeks (Oliver Burkeman) #non-fiction
You get about four thousand weeks. Burkeman argues for accepting that instead of optimizing against it.
The argument about finitude stuck with me, but much of the rest did not.
Mickey7 (Edward Ashton) #fiction
A disposable clone on a colonization mission, printed fresh each time he dies doing the dangerous jobs, until the system glitches.
I added this to my list after seeing the movie, and I preferred the book’s arc. I enjoyed it enough that I immediately picked up the sequel, Antimatter Blues.
Extremely Hardcore (Zoë Schiffer) #non-fiction
An inside account of Elon Musk’s acquisition of Twitter and the chaos that followed.
I knew the broad story, but not what it looked like inside Twitter while it happened.
A Paradise Built in Hell (Rebecca Solnit) #non-fiction
Solnit’s argument that disasters produce cooperation and mutual aid, not the chaos and looting we expect.
Given my work at GoFundMe, I have a strong interest in mutual aid. The core argument stayed with me even where the framing felt heavy.
Going Infinite (Michael Lewis) #non-fiction
Michael Lewis’s account of Sam Bankman-Fried’s rise and the collapse of FTX.
Familiar ground, but compellingly told. I came away partly convinced that Sam Bankman-Fried started with overconfidence and a belief that discipline was optional, not an explicit plan to commit crimes.
Personal Rules I’m Keeping Next Year
I’m keeping the three-track rotation: one fiction, one non-fiction, one technical. It prevents stalls from turning into full stops.
I’m keeping chapter switching, along with permission to break it whenever a book has momentum.
Fiction stays the relief valve. I’m allowed to enjoy it without trying to extract a moral.
Technical stays broad. If it helps me think clearly about technology, incentives, systems, or how work actually happens, it counts.
The nicest surprise this year was that none of it felt like discipline. It’s just a rotation I come back to, one chapter at a time.