My Code, My Test, and My Prompt All Agreed. All Three Were Wrong.
I tried to replace my receipt model with one 18× cheaper. The bug was in my benchmark, and the real lesson was about trusting AI to verify itself.
Jul 8, 202612 min read5
Command Palette
Search for a command to run...
#llm
Articles tagged with #llm
The Job Isn't Writing Code. It's Knowing When the AI Is Wrong.
Three examples from my AI-powered job board pipeline where the output looked right but wasn't.
Feb 19, 20264 min read77I Benchmarked 6 LLMs to Automate My Job Board for $0.35/Month
Not all models are good at the same thing. Here's the data.
Feb 10, 20269 min read25