
Erik Zettersten · Feb 23, 2026
Why I Moved Back to CLI Development
After years of IDE-heavy workflows, I moved most of my coding back to the terminal. Here’s what changed, what got better, and where full IDEs still win.
Read More →Journal
Strong opinions, practical details, and real implementation context.

My hot take: despite the public standoff, Washington and Anthropic will reconcile sooner than later. My spicier take: Dario’s pride is making this phase louder and longer than it needs to be.
Read More →
Erik Zettersten · Feb 23, 2026
After years of IDE-heavy workflows, I moved most of my coding back to the terminal. Here’s what changed, what got better, and where full IDEs still win.
Read More →
Erik Zettersten · Feb 23, 2026
If foundation-model labs trained on the open internet, they should expect competitors to optimize extraction from public model interfaces. You can dislike it and still admit the symmetry.
Read More →
Erik Zettersten · Feb 22, 2026
Taalas is making a sharp bet: stop simulating models on general-purpose computers and turn them into hardware. The model quality may not be frontier-tier yet, but the latency profile changes what AI can do in production.
Read More →
Erik Zettersten · Feb 22, 2026
Twelve years in, I can say this with zero hesitation: marriage is one of the best decisions I have ever made, and Wendy is the reason our family stands strong through every season.
Read More →
Erik Zettersten · Feb 21, 2026
The market reaction to Anthropic’s Claude Code Security launch says less about one feature and more about how fragile security narratives are right now.
Read More →
Erik Zettersten · Feb 21, 2026
My career only looks chaotic from the outside. From the inside, it’s one long pattern: find unfinished systems right before they matter, then build under pressure.
Read More →
Erik Zettersten · Feb 21, 2026
An Oscar-winning actor geeking out over niche Italian watch design is not cringe. It's the best kind of enthusiast energy.
Read More →
Erik Zettersten · Feb 20, 2026
Anthropic’s new security preview is interesting and useful, but the 'AI replaces security teams' takes are still unserious.
Read More →
Erik Zettersten · Feb 20, 2026
The viral post isn’t just about a feature launch. It’s about packaging, timing, and distribution discipline.
Read More →
Erik Zettersten · Feb 19, 2026
Phil Spencer leaves behind both unfinished promises and undeniable wins. Xbox’s next era has upside, but also a dangerous experience gap at the top.
Read More →
Erik Zettersten · Feb 19, 2026
Open, practical, geospatially grounded AI tooling beats another generic chatbot demo every time.
Read More →
Erik Zettersten · Feb 16, 2026
A practical operator’s view on AI regulation risk before it becomes a board-level emergency.
Read More →
Erik Zettersten · Feb 13, 2026
Safety is too often treated as a PR category instead of a measurable engineering discipline.
Read More →
Erik Zettersten · Feb 10, 2026
If you don’t measure truth, you’re shipping fiction. Here’s how to engineer for reliability.
Read More →
Erik Zettersten · Feb 7, 2026
Most AI roadmaps are expensive theater. Here’s the brutally practical way to build things that survive production.
Read More →
Erik Zettersten · Feb 4, 2026
RAG can be useful. It’s not a strategy, and it definitely isn’t magic.
Read More →
Erik Zettersten · Feb 1, 2026
AI-first sounds bold. In practice, most teams are shipping theater. Here’s the operational playbook that actually works.
Read More →
Erik Zettersten · Jan 28, 2026
Speed without standards creates AI chaos. Here’s how to lead teams that move fast and still ship trustworthy systems.
Read More →
Erik Zettersten · Jan 24, 2026
Most leaders are managing 2025 engineers with 2012 playbooks. Here’s the upgrade.
Read More →
Erik Zettersten · Jan 20, 2026
Many CEOs still hold the title, but boards, investors, and optics now run the playbook.
Read More →
Erik Zettersten · Jan 16, 2026
Top builders are drifting toward smaller teams with real ownership, faster feedback, and less organizational drag.
Read More →
Erik Zettersten · Jan 12, 2026
Frontier model economics are compressing entry-level engineering work faster than most companies are willing to admit.
Read More →
Erik Zettersten · Jan 8, 2026
An enterprise-operator view on OS bloat, reliability drift, and the hidden tax on IT teams.
Read More →
Erik Zettersten · Jan 4, 2026
VR didn’t fail because the tech was impossible. It failed because the strategy ignored behavior.
Read More →
Erik Zettersten · Dec 30, 2025
Polish is not the same as joy. Why modern game design too often optimizes systems over soul.
Read More →
Erik Zettersten · Dec 24, 2025
When a 'normal' enthusiast build starts flirting with luxury pricing, the value equation is broken.
Read More →
Erik Zettersten · Dec 18, 2025
LAN culture taught debugging, systems thinking, and collaborative resilience in a way modern credential pipelines rarely do.
Read More →
Erik Zettersten · Dec 12, 2025
Parenting strategy for an economy where automation changes faster than school systems.
Read More →