Writing helps me think.

There’s a feeling you get when everything clicks into place. Not the eureka moment of solving a problem, but something more fundamental: the deep satisfaction of being exactly where you need to be, with exactly what you need, arranged exactly how it should be. It’s the feeling of home, but applied to the spaces where we create. This isn’t about aesthetics or luxury. It’s about something more primal: the human need to nest, to claim territory, to shape our environment until it shapes us back. And nowhere is this more crucial than in the spaces where we do our most demanding creative work. ...

As a hobby, I’ve been perfecting various burger recipes. It both feeds us and gives me something to incrementally improve. Add fake internet points to the mix, and now I’m motivated to post photographs of my creations. The feedback and reinforcement these fake points provide are enough to keep me consistently taking those photos, and making those burgers. A little push can be good. But here’s the thing about fake internet points: they work. Absurdly well. Here I am, a grown adult who builds software for a living, getting genuinely excited because 47 strangers clicked an arrow next to a photo of my latest smash burger creation. The dopamine hit is real, even when I know exactly what’s happening to my brain. ...

Your AI coding assistant hasn’t gotten worse; you’ve just hit the ceiling of what your current approach can handle. “My LLM coding tools have been serving me hot garbage lately” is a complaint I see weekly on Reddit. Same tool, same developer, suddenly worse results. These threads are full of developers convinced their AI assistant has degraded overnight. Same prompts that worked last month suddenly fall over. The tool must be broken, right? ...

After three months of AI-assisted development, I’m more productive than ever. I’m also mentally exhausted in ways I didn’t expect. The productivity gains are real—I can go from sketch to working prototype in hours instead of days. But there’s a hidden cost that we’re not talking about: the tools fundamentally change the cognitive load of programming. Instead of being tired from wrestling with implementation details, I’m exhausted from operating constantly at the architecture and design level. ...

Something weird happened the other day. I found myself sketching more than I had in years. The pages quickly filled with the doodles, wireframes, and half-formed ideas that used to live in my head until they died of neglect. The reason? I can finally get from sketch to fleshed-out concept fast enough that my ideas don’t evaporate. After a few months of serious AI-assisted development, I’m starting to understand what these tools actually do well, and more importantly, where they will cheerfully lead you off a cliff. They don’t replace thinking, they amplify it. But only if you understand the difference between augmenting your expertise and outsourcing your judgment. ...

After thirty-something years of working in terminals, I’ve developed a pretty strong bias toward doing things from the command line. It can be faster for me to navigate around using scripts and shortcuts than clicking through GUI hierarchies. The muscle memory is there, the automation possibilities are endless, and frankly, it just feels more direct. The problem is that GUI applications don’t always want to cooperate with this workflow. You end up breaking your flow to hunt through Spotlight, dig around in Finder, or, worst of all, try to remember the actual executable path buried somewhere deep in an .app bundle. It’s jarring when you’re in the zone. ...

There’s a familiar anxiety in the air among software architects and engineering leaders. AI coding tools are going to make us lazy programmers. They’ll erode our fundamental skills. We’ll lose the ability to think through problems. Sound familiar? It should. I’ve heard this exact song before, just with different verses. In the ’90s, visual development environments were going to create a generation of developers who couldn’t “really” code. Before that, higher-level languages were supposedly making us soft. Hell, I remember when debuggers were contentious—Linus famously ranted that real kernel developers should be able to think through their code without stepping through it line by line. ...

We’ve all seen the ergonomic warnings. Adjust your monitor height. Take breaks every twenty minutes. Use a proper keyboard. The IT department sends out reminders about repetitive strain injuries, and we dutifully nod while hunching over our laptops for another six-hour coding session. But here’s what nobody talks about: your brain is a muscle too, and we’re giving it the cognitive equivalent of carpal tunnel syndrome. The thinking muscle I’ve been writing software for over twenty years, and I’ve watched my industry evolve from “ship it when it’s done” to “ship it yesterday, fix it tomorrow.” Somewhere along the way, we forgot that thinking—real, deep problem-solving—is physical work. Your brain burns glucose. It gets fatigued. It needs recovery time. ...

I sat down a few weeks ago to ask Claude to help me analyze some user feedback data. I spent fifteen minutes crafting what I thought was the perfect prompt: > Please analyze this data, but don't make assumptions about causation, don't over-interpret small sample sizes, don't use corporate jargon, don't suggest solutions we can't implement, and definitely don't... The response was generic and cautious. Frustrated, I tried again with a simple request: ...

This is the 42nd version of my personal website and weblog, once again self-hosted at Dreamhost. I use the site as a forcing function; a way to learn various languages and tools, and a kick to get back into the habit of writing and publishing. This year I ported the content from Write.as to a self-hosted Hugo setup, which took about a day. Everything about Hugo has been better than expected, and the only hiccup was on the hosting side (SSL setup was broken for a few hours). ...