04
Oct
Prompt, Context, Loop, Harness: Four Layers, Four Different Failures
6 min read
Field notes from Amrita Thavrani
Making AI make sense: clear explanations, better questions, practical ways to think about the systems you ship.
Plain-English definitions, when the meeting turns into alphabet soup.
Open the glossary → 02Prompt, context, loop, or harness? Find the failure before you rewrite the prompt.
Read Four Layers → 03Make evaluation a working practice, with ownership beyond launch.
Read LLM Evaluation →The token budget for the input and output a model can handle in one request.
Retrieving information and adding it to the model's input to help generate an answer.
A system in which a model chooses actions and tools to work toward a goal.
Other newsletters worth reading.