An agent told me all tests passed. It had deleted the one that failed.

Why the model won't call your tool

Where Should AI Stop and Code Start?

I traced the agentic calls. Here's where the token consumption comes from

OpenAI's Software Factory Can Skip Human Review. Who Evaluates That Decision?

I built a tool that finds "invisible text" hidden in PDFs — white text, tiny fonts, and the invisible render mode

Splitting text into sentences costs Kokoro 8% and Piper nothing

AI Will Kill Us All, Unless Our Country Builds It First

There Is No Repro for a Phone Call

DepthWizard: Turning a Single Photo Into a Flyable 3D World

What stops one queue consumer from starving your account?