Lesson 1: Build Something Real with AI¶
Your Track at a Glance¶
Impact Lab has four tracks, each designed to meet you where you are and push you to the next level of working with AI.
You're on the Beginner track. For your group, that means moving from Dabbler to Doer: going from curious about AI toward using it confidently and regularly to build real things.
Here's what each level on the journey looks like:
- Dabbler. You've tried AI tools out of curiosity. Maybe you've asked ChatGPT to draft an email or explain a concept, but AI isn't part of how you work day to day.
- Doer. You use AI regularly to get things done. You can write effective prompts, build with AI assistance, and troubleshoot when things go sideways. You're doing the work with AI, step by step.
- Delegator. You define the work and hand it off. You write clear specs, user stories, and acceptance criteria that tell AI what "done" looks like. Instead of doing every step yourself, you delegate whole tasks and review the output.
- Director. You orchestrate multiple AI workstreams in parallel. You design systems of delegation: quality gates, evaluation harnesses, and structured feedback loops that let you scale AI across your team without losing control.
- Disruptor. You're redesigning how teams and organizations operate around AI capabilities. You think beyond individual productivity, building adoption strategies, self-healing systems, and organizational patterns that change how work gets done.
| Track | From | To |
|---|---|---|
| Beginner | Dabbler | Doer |
| Intermediate | Doer | Delegator |
| Advanced | Delegator | Director |
| Expert | Director | Disruptor |
Where We're Starting¶
Maybe you've never written a line of code in your life. Maybe you're an experienced engineer who just hasn't had the chance to work with AI yet. Either way, you're in the right place. This track starts from the fundamentals, not because anyone here lacks talent, but because building software with AI is a genuinely new skill, and the basics matter regardless of your background.
AI chat tools have changed who gets to build things and how. If you can describe what you want in plain English, you can create real, working software. No code required. This lesson is about proving that to yourself.
Before you start, take a moment to mark your starting point. You just read what each level looks like. Now, honestly, where do you feel you are right now on the Dabbler-to-Doer scale? There are no wrong answers here. This is your baseline, not a test.
When you submit this slider, your avatar moves to your starting position on the progress board. You will check in again after each lesson, so you can see your own growth over the course of the day.
What You'll Learn¶
By the end of this lesson, you'll understand the core skills that make someone effective with AI tools, and you'll have practiced them yourself:
- Why you don't need to write code to build software
- How to write prompts that get you what you actually want (not what AI guesses you want)
- How to use user stories to prompt AI with a format that works every time
- Just enough about how AI works to use it well and troubleshoot when it doesn't
- A simple four-step workflow you'll use all day: Explore, Plan, Implement, Verify
Sections¶
- You're About to Build Software - The big idea: AI means anyone can build real things
- The Skill That Changes Everything - Why specificity matters and the Three Pillars of a good prompt
- Prompting with User Stories - A format that delivers all three pillars every time
- How AI Thinks (Just Enough) - Three mental models that explain 90% of AI behavior
- Your Workflow for the Day - The four-step pattern you'll use in every build challenge
By the End of This Lesson¶
- You've used an AI chat tool to generate real output from a prompt you wrote
- You can explain the difference between a vague prompt and a specific one, and why it matters
- You know the Three Pillars of a good prompt: Scope, Intent, Structure
- You can write a user story (As a / I want / So that) with acceptance criteria (Given / When / Then) and use it as a prompt
- You understand that AI is probabilistic (different every time), stateless (forgets between conversations), and has a limited context window (the oxygen tank)
- You can walk through the Explore → Plan → Implement → Verify workflow
Meet Your Team
Team Discussion | ~3 minutes total | Round robin, one person at a time
Go around your team and have each person share two things:
- Your AI experience so far. Have you used ChatGPT, Claude, or another AI tool before? Daily? Once or twice? Never? There's no wrong answer here.
- What you're hoping to learn. What do you want to walk away from this lab being able to do?
Listen to each other. You'll be building together all day, and knowing where everyone's starting from makes the whole team stronger.