Start here
New to AI, or new to this newsletter? These five posts get you from zero to working vocabulary, in order. Each builds on the one before. About 20 minutes total.
01AI/ML Terms Explained Simply - Part 1
The umbrella terms: what AI and ML actually are, and how they show up in your daily tools.
AI/ML Terms Explained Simply - Part 2
Training, overfitting, tokenization, embeddings, inference - the words behind every AI demo.
Look-Alike, Mean-Different: AI's Most Confused Terms
Mistral vs Mixtral, embeddings vs encodings - the sound-alike pairs everyone mixes up in meetings.
Building an App - 101
Vocabulary in action: one small, real app, walked through end to end.
Prompt, Context, Loop, Harness: Four Layers, Four Different Failures
The working framework: when an AI feature misbehaves, find which layer is actually broken.
After that: LLM Evaluation Is a Program Management Problem (10 min) when you start running an AI program, and Frontier models: pick for the job, not the score (14 min) when you're choosing a model.
Most read: See the Popular posts on the homepage →