Dave, you should be sorry
His most recent blog post is the worst one, but it isn't new. I read two-plus years of David Heinemeier Hansson's writing, and then watched someone I respect get excited to work with him anyway.
His most recent blog post is the worst one, but it isn't new. I read two-plus years of David Heinemeier Hansson's writing, and then watched someone I respect get excited to work with him anyway.
AI has made me measurably faster at my job. My salary hasn't moved. Here's where I think that value is actually going.
Meta's new AI manifesto makes an argument I've made myself, about concentrated power being the real danger. That's exactly why it's worth taking apart.
Webinar comments rage about AI ethics and never once mention the cobalt in the phone they typed it on. The selectivity is the actual story.
Elon Musk says AI and robots will make work optional and hand everyone a universal high income. He never explains how ownership gets from him to everyone else.
I use AI every day and I think AI art is bad for the world. Here's my attempt to think through why - honestly.
I used to have a clear answer to this question. Agentic AI is making it harder.
Anthropic's new identity verification policy is narrower than the takes suggest. The vendor they hired to do it is not.
I write a lot about using AI as a tool. That probably makes me look like a true believer. The answer is more complicated.
What happened at Newark's Delaney Hall ICE detention center over Memorial Day weekend, and the infrastructure that made it possible.
The repo is the surface. The instructions file is the manual. The human review keeps it safe.
The GM layoffs are a preview of a broader shift. AI isn't simply eliminating tech work - it's changing what tech work looks like. That's not necessarily good. But it's happening.
The viral image is sensationalized. The infrastructure behind it is not.
AI didn't change whether you need to understand what you ship. It changed how much of it you wrote yourself.
AI is powerful. That's not the interesting question. The interesting question is why we haven't solved these problems already.
The Technological Republic isn't political philosophy. It's the ideological scaffolding for technofascism, written by the people who profit from it.
AI can't accelerate your device management workflow if your config lives in a GUI. Here's why GitOps is the prerequisite for everything that's about to happen in this space.
AI is making code cheaper to write. That means the constraint in software development, and device management, has shifted to product design and QA.
The viral AI agent is everything we wanted. It's also a security researcher's nightmare. Here's what you need to know.
Scholars say the U.S. no longer qualifies as a democracy. The tech industry built the tools that made this possible.
A colleague recently found out that I use an LLM as part of my blog writing process, and they told me that this changes how they see my writing. But my blog posts haven't changed, only the knowledge of how I made them. Does that change the value of my writing? And if so, why? These questions are complicated, and I don't have the answers. Neither do you.
Think you can spot AI content? You're probably missing the good stuff. Like CGI in movies, you only notice bad AI - great AI-assisted work is invisible.