I let an AI coding agent loose inside a real codebase, and yeah, it was impressive for about five minutes.
Then the weird stuff started.
A clean refactor broke auth, renamed things no one asked for, and quietly added a cost line item nobody mentions in demos.
But it also saved hours on boring edits, test scaffolding, and repo search.
That’s the truth most blo...
Most "AI agent" articles list frameworks for building agents.
This isn't that.
Using an AI coding agent and building one are completely different problems. Using one means file system access, real test suites, actual PRs, production config files. The bar is higher than "it autocompletes well."
I've been watching teams use — and abandon — these tools on r...
90 days of telemetry from an open-source MCP server that drives the actual Excel desktop app. The numbers were not where I thought they would be.
Half a year ago I asked a simple question: why can AI agents write a React app from scratch but choke on Revenue Model v27 final final.xlsx?
The answer turned out to be boring and important.
Agents...
Anthropic just shipped Claude Design, a new Labs product that turns prompts into working designs, interactive prototypes, slide decks, and one-pagers. It's powered by Claude Opus 4.7 and - most interestingly for...
Most conversations around AI still revolve around one fear:
Job loss.
Will roles disappear?
Will automation replace people?
Will entire industries shrink?
It’s a valid concern.
But the more I observe what’s actually happening…
the less convinced I am that this is the real risk.
Breaking the Expectation
We tend to look for visible ...
Hello, I almost forgot I had to post on the 30th. I probably would have forgotten if I hadn’t checked my calendar last night.
(Yeah, not the best sign right at the start… but I’m not giving up this time.)
So, I started learning from The Odin Project, and I’ll be honest — the first few chapters...