Archive
Applied AI Digest: Volume 3
I wrote a profiling tool for agents
Applied AI Digest: Volume 2
Stop designing chatbots
Artifact-driven AI creation
Using MCPs with evals
Applied AI Digest: Volume 1
The "quick fix" that works: runtime patching in Python
If I could turn back time
Your evals have a Rotten Tomatoes problem
Tiny OCR
The Virtues of Showing Your Work: Do LLM Explanations Actually Help?
On the new larger context windows
Let computers talk to computers
Do what I mean
Your Questions on Building with AI
Using historical codebase to build an eval suite
Customizing Cursor's code generation and personality
Insisting On Known Knowns: Using Evaluators to Drive Reliability
Classification w/Confidence Scores Using Logprobs
Taming LLM Responses: Dynamic Pydantic Models for Flexible Structured Output
Fine Tuning Quirk
OpenAI Structured Output + Pydantic: Adding Support for Default Values

