Gears Within Gears
No man but a blockhead ever wrote, except for money. — Samuel Johnson I like the act of writing , but often find actually publishing paralyzing. I’ll sit with a piece for weeks, hem and haw over it and chew on it and maybe pass a draft around, but by the time I’m done I’ll have lost whatever conviction I started with. A fair number of pieces never finish, and the drafts folder…
One of the more useful definitions of software architecture comes from Building Evolutionary Architectures : architecture is definitionally the stuff that’s hard to change. 1 I’ve always found this definition to be the most honest framing available, to say nothing of the simplest. It doesn’t pretend architecture is about beauty or correctness or your resident architect’s…
When a company hits a ceiling on its core business, it often tries to grow demand. The Michelin Guide wasn’t about restaurants; it was about miles driven. Facebook’s internet.org wasn’t about altruism; it unlocked expansion by growing the size of the entire internet. The ceiling on the business was the medium, so they expanded it. AI companies face the same structural problem:…
We have entered a period in which AI is able to author more code than is reviewable by humans. Many teams are already past the point where each line can realistically be scrutinized; the volume of code being generated is increasing faster than our capacity to inspect it. Every AI-assisted coding session produces choices: naming decisions, structural decisions, abstraction decisions, dependency…
This is a transcription of a talk given by Mary Poppendieck , The Tyranny of the Plan , lightly edited by me for readability and brevity. A similar transcription was previously hosted here , with Mary’s permission, but has since been taken down. I share it here because I think it’s an important talk and worth preserving, and I hope it will as be useful to others as it has been to me.…
By the end of 2025, a question that’s hung over the industry for the past year—is agentic software development a novelty, or is it here to stay?—now has its answer: it is. In 2026, it becomes the new normal. Credible , senior engineers started making real cases for productivity improvements: demonstrated improvements in throughput, fewer dead-end branches, less yak shaving, more time spent…
I’m Brian Guthrie, a software engineer, leader and entrepreneur based in New York City. I write about the intersection of software, talent, complexity, and systems thinking. Presently I lead a team at Justworks that helps ambitious firms grow globally. Prior to this, I founded Orgspace , led engineering at Meetup , and hacked all over the world with Thoughtworks , including, for a time, with…
Hi there! I’m excited to be working with you. This is meant to give you a rough introduction to me—how I think, what I value, what works for me and what doesn’t. What success looks like I want us to ship valuable software, quickly, sustainably, and at high quality. I want the teams I work with to be known for getting things done by consistently delivering work that matters. I want the…
I’ve been spending a lot of time recently thinking about good generative AI practices, most particularly for code. This is obviously a rapidly evolving topic in a nascent space, but I have a metaphor here that I’ve used repeatedly, and I stand by it: treat generative AI like a bicycle, not a taxi. Bicycles require skill and finesse to operate, but provide mechanical advantage to the operator. They…
The myth of objectively good code For as long as people have been writing software, people who write software have been arguing about how to write it. Because code has to balance the need to do something with the need to communicate to humans what it does , we say that it must express intent : good code makes it easy to understand. Consequently, programmers often make aesthetic choices which help…
The quest to kill meetings A few years ago I was leading the consumer product engineering division at a startup, and one of the teams I was leading was repeatedly flagging the same issue: too many meetings. Clawing back time to focus on day-to-day work is a recurring pain point for engineers, because for many of us it’s difficult to regain our focus and concentration after an interruption . I…
Happy Gilmore is a 1996 Adam Sandler movie about a guy who’s weirdly good at driving golf balls, and so becomes a golf celebrity even though the rest of his technique is terrible. I’m not into golf, but there’s a scene in the movie in which Happy’s coach advises him to treat the ball like it wants to go home. (Happy fails to sink the putt, loses it, and yells at the ball instead.) I think of…
I used to run an internal software platform department (infrastructure, tooling, devops), and every so often, it was our immense honor—and, indeed, joy—to delete old code. But the trouble with old code is that it sometimes still does things, and even the undesirable things often matter to someone somewhere. The practical mechanics for doing so without irritating the bejesus out of everyone were…
I built an entire startup on top of an event-sourced data architecture. In the end we didn’t win in the market, but I walked away from the experience thoroughly convinced by the power of the model. In this post I want to sketch out the essential elements of the pattern, and what makes it so compelling; in particular, I want to make an argument that more teams should take it seriously, even…
From time to time I come across a problem that I don’t know how to solve but that I’m certain someone else must have already solved. One of the benefits that experience brings is that it’s relatively safe to make the inference that if a problem seems interesting or important, someone else has already taken a whack at it publicly: it’s a big internet out there, and in…
I’ve worked in the software industry for over twenty years. I have been a consultant, a line engineer, a director, a founder, and a CTO. I have worked at firms as small as two engineers and as large as fifty thousand. I frequently get asked how to move faster in software. This is what I say. It is easier to describe than it is to do. Rules for moving faster I am not some kind of process…
I originally wrote this on Twitter in 2017 . I’m reposting it here, on my own site, to preserve it. Rant time: today I’d like to talk about feature branching and GitHub—how we got here, what they are (and aren’t) good for, why that matters, and what you can do instead. Join me, friends! First, a definition: “Continuous Integration is the practice of merging all developer…
I originally wrote this on Twitter in 2017 . I’m reposting it here, on my own site, to preserve it. Today I’m going on a rant about the term DevOps. Join me, friends! Here’s the rant: the term DevOps originally had a completely different meaning from how it’s now commonly understood, and I’d like to talk a little bit about why that matters, and what you can do to…
The end of an era It’s been rough couple of years for Thoughtworks. After a long-awaited IPO in 2021, their stock price declined steadily, prompting the board to replace longtime CEO Xiao Guo with Mike Sutcliff, who has now taken the firm private again. Xiao was a protégé of Thoughtworks’ founder Roy Singham and a 25-year veteran of the firm, and as such was deeply immersed in its…