Module 2: Working in Conversation
The Practicum opens with the map of the three interaction modes and how to choose among them, then teaches the conversation mode in depth. Work with a model runs on a rising gradient of autonomy, and each step up trades control for leverage.
Session structure
The 30-minute peer share is protected in every module, because peer exchange, not formal instruction, is where the learning compounds.
Capstone build step
Memory decided, and a sample brief from your first brief prompt.
A durable home for context instead of trusting one conversation to hold it. This is where context is king becomes something you can open and read.
Memory: where your context lives
Choose where your context lives, write your first brief prompt, and see a sample brief as a private artifact. No hosting and no code.
Start building ->The materials
Pre-work
What to have done before we meet: the Module 1 assignment.
Open pre-work ->Session guide
The full run of show: the three blocks, the timings, and the facilitator notes. Sign-in required.
Open guide ->Teaching slides
The 16:9 deck for display during the instruction block. Sign-in required.
Open deck ->Homework
You have felt the conversation mode. Next week hands a bounded job to the model instead of steering every turn, so arriv
Open homework ->Debrief
The recap and the session recording, posted after we meet. Sign-in required.
Open debrief ->Reading
The Five Elements Spine: Personal OS Architecture Guide
The primary reading, and the map you will grade your own system against for the rest of the program.
Finding Claude's Voice
A participant technique on tuning how Claude talks with you in conversation. Run the protocol and build your own voice artifact before the session.
Time to read: about 18 minutes
Conversation Architecture
The recovery patterns: when to keep a conversation going, when to start fresh, and how to carry context forward with a handoff summary.
Time to read: about 7 minutes
The Winter Method
Using a model as a genuine thinking partner rather than an answer machine.
Time to read: about 4 minutes
The AI Landscape, Briefly
Narrow, generative, agentic, and an honest way to compare models yourself before Module 6's deeper evaluation.
Time to read: about 1 minute