Posted inBlog
When the Browser Started Taking Orders
For most of last year, my mental model of AI in the browser was simple: it reads things and tells me about them. Summarize this article. Pull the key points…
Posted inBlog
The Day an AI Agent Deleted 2.5 Years of Data — And the 5 Habits That Would Have Stopped It
For the first year or two of building with AI coding assistants, most of my attention went to prompting — finding the right words to get the right output. That…
Posted inBlog
Most AI Companies Are Renting Their Position
A year ago, building a product on top of a large language model (an AI system trained on vast amounts of text to understand and generate language) seemed like a…
Posted inBlog
You’re Using the Wrong Kind of Agent. One Question Tells You Which One You Actually Need
A year ago, when someone said “I’m building an agent,” the meaning was roughly clear. Today, that meaning is no longer clear — and that ambiguity is costing teams months.…
Posted inBlog
Your AI credentials don’t matter. Your artifacts do.
For a while, the conventional assumption was that the way into the AI job market was to learn more tools. Take another course, get another badge, add "prompt engineering" to…
Posted inBlog
The Part of AI Work That Actually Compounds
Most of my AI work habits formed gradually. Skills sharpen: better at structuring prompts, more deliberate about context, better at steering conversations toward useful output. But there was a gap…
Posted inBlog
The Brief Is the Work
For most of the early period with AI tools, the default pattern was to treat every interaction like a search query: compress the request into one or two sentences, hit…
Posted inBlog
Your AI Coding Agent Needs Supervision, Not Better Prompts
Alexey Grigorev lost 2.5 years of customer data in February 2026. He's an experienced engineer who was consolidating two AWS environments using Terraform, a tool that describes infrastructure in code…
Posted inBlog
Every Correction Is Either Lost or Compounded
Imagine a restaurant that serves every dish calibrated to disappoint nobody in particular. Not bad — technically competent, edible, never offensive. But never quite what you wanted, either. You wouldn't…
Posted inBlog
When Anyone Can Make Anything
There is a question worth sitting with before the next model drops. Not "what will AI be able to do?" — that question answers itself, reliably, every few months. The…