Module 2
Surface what happened with Perplexity and Shape of Meaning over the past week.
Wins and losses are both useful. If someone hit a significant failure - bad information, an embarrassing output - use it. "That is the verification lesson from last week made real. What would have caught it?"
Walk through the landscape page together. Give participants a map that stays useful even as the tools change.
Pause on the agentic AI accountability callout. In regulated industries, executives will be making deployment decisions about agentic AI within 18 months. The governance question is not theoretical.
Share your screen and walk through jayfontanini.com/accelerator/curriculum/module-2/ai-landscape/ together. The taxonomy diagram anchors the discussion. Participants can read the full page on their own after the session.
Before participants build, show them what done looks like. Share the AI Coaching project - the workspace where this program was built.
The line that lands: "Every page you read in this curriculum, every session structure you are following - this is where it was built. The tool built the program that teaches the tool."
Build a Claude Project live with the group watching. Ask participants for a real problem first - then build around what they offer.
If the room goes quiet, have a fallback project ready to build - something concrete enough to demonstrate the structure but simple enough to set up in five minutes. The point is showing the decisions: project name, instructions, first file upload, first real question.
Once the project is running, let participants drive the first few questions. Your job is to narrate why certain questions land better than others when Claude already knows the context.
Send participants off with a clear assignment and genuine motivation to do it.
The encouragement to give: "The first project you build will not be perfect. The instructions will be too vague or too specific. That is fine - you will know what to fix after one real conversation. The goal is to start."