Before we combine AI with external data, we first learn to read that data cleanly. AI returns in Week 2. This week is five small functions that all use one primitive: search(query, count=n) — a synchronous call that returns a list of result dicts with title, url, and snippet keys.
No agents at all for five lessons? I thought this was an AI track.
Retrieval is half of retrieval-augmented AI. If you can't query, count, and reshape search results fluently, stacking an agent on top only hides the mess. By Friday you'll own the search primitive completely — then Week 2 wires an agent onto it and suddenly every Week 1 function has intelligence layered on.
So Week 1 is the foundation, and Week 2 is when the AI shows up?
Exactly. Read search cleanly this week; reason over it next week. The functions you write now become the input layer for every agent call you make for the rest of the track.
Goal: fluent reading of search() results — the raw material Week 2 reasons over.
Create a free account to get started. Paid plans unlock all tracks.
Before we combine AI with external data, we first learn to read that data cleanly. AI returns in Week 2. This week is five small functions that all use one primitive: search(query, count=n) — a synchronous call that returns a list of result dicts with title, url, and snippet keys.
No agents at all for five lessons? I thought this was an AI track.
Retrieval is half of retrieval-augmented AI. If you can't query, count, and reshape search results fluently, stacking an agent on top only hides the mess. By Friday you'll own the search primitive completely — then Week 2 wires an agent onto it and suddenly every Week 1 function has intelligence layered on.
So Week 1 is the foundation, and Week 2 is when the AI shows up?
Exactly. Read search cleanly this week; reason over it next week. The functions you write now become the input layer for every agent call you make for the rest of the track.
Goal: fluent reading of search() results — the raw material Week 2 reasons over.