AI for Entrepreneurs

Why: Nobody should feel dumb on Day 1 because of a word they haven't heard before. What: Self-study reference guide covering 18 terms in plain English — software, app vs. website, server, the cloud, API (the waiter between you and the kitchen), IDE (workspace for building), code, database (spreadsheet with superpowers), terminal/command line, open source, SaaS, framework, version control/Git, GitHub, deploy, domain, DNS, hosting. Each term gets one sentence in plain language plus one real-world analogy. How: Read through before Day 1. Circle any terms you want to discuss. Bring questions. What If: What if you could walk into any tech conversation and understand 80% of what's being said — not because you became technical, but because someone finally explained the words?

Why: OPEN WITH DUSTIN'S STORY — 'Six months ago I didn't know what a database was. I uploaded one HTML file to the internet and then did nothing for 10 months. Then something changed. I started describing problems to an AI and watching it build solutions. In 6 months I went from that to running 10+ production apps with AI agents that work while I sleep. 29 projects. Almost 3,000 commits. Not because I became a programmer — because the tools changed what's possible. And that's why you're here.' Then transition: you're already using AI — spam filters, autocomplete, Netflix recommendations. But nobody's told you what it actually is in plain language. What: AI = pattern matching on massive text, not thinking. The timeline: rule-based systems > machine learning > deep learning > LLMs > agents. What a 'model' is (the trained brain) vs. a 'product' (ChatGPT, Claude — the interface to the brain). The 6-month timeline: Dustin's real progression shown visually — static HTML → first framework → database + auth → mobile → infrastructure → autonomous agents. This is where you're going. How: Group exercise — look at 5 things and guess: is this AI or just software? Discuss what makes the difference. What If: What if you could evaluate every AI claim you hear — from vendors, news, LinkedIn — and immediately know what's real and what's hype? And what if, six months from now, you had your own version of that story?

Why: If you don't understand how the engine works, you can't drive it well. You don't need to be a mechanic — but you need to know it runs on gas, not magic. What: Training = fed the internet, learned patterns. Tokens = how AI reads (chunks, not words). Context window = short-term memory — how much it can hold at once. Why it hallucinates (pattern completion, not fact retrieval). Not live-searching. Not thinking. Not remembering you between conversations. How: Demo — show the same prompt with different context. Show a hallucination live and explain why. Pairs: try to make the AI hallucinate on purpose. What If: What if you never got fooled by an AI confidently saying something wrong — because you understood why it happens?

Why: Walking into a car dealership without knowing brands exist means you're buying whatever the salesman pushes. Same with AI. What: The major players: OpenAI (ChatGPT/GPT), Anthropic (Claude), Google (Gemini), Meta (Llama). Open source vs. closed models — why it matters for business. Free tiers vs. subscriptions ($20/mo) vs. API pricing (pay per use). Which one for what. How: Hands-on — open ChatGPT, Claude, and Gemini side by side. Ask all three the same question. Compare the answers. Discuss differences. What If: What if you picked AI tools the way you pick any business tool — based on fit, cost, and capability instead of whatever your friend mentioned?

Why: Most people treat AI like a search engine — one question, one answer, done. That's like having a brilliant consultant and only asking them one thing. What: The conversation model — it's a dialogue, not Google. You can push back, clarify, redirect, build on. Starting with your real problem, not a test prompt. The difference between asking and collaborating. How: Exercise — bring a real business problem you're facing right now. Spend 20 minutes working through it as a back-and-forth conversation. Partner observes and notes when you could have pushed deeper. What If: What if your first instinct when you hit a business problem was to talk it through with AI before calling a meeting? NESTED LOOPS: Overt = learn conversational AI. Loop 2 = they're doing real business strategy work disguised as AI practice. Loop 3 = partners observe each other, learning prompt patterns from watching someone else's approach. Loop 4 = the experience of AI 'getting it' creates an emotional anchor — AI stops being scary.

Why: 'Garbage in, garbage out' has never been more true. The gap between a bad prompt and a great prompt is the gap between a useless answer and a brilliant one. What: The 5 levers: context (background info), role ('act as a CFO'), format ('give me a table'), examples ('like this, not like that'), iteration ('good, now make it shorter'). Chain prompting — breaking complex asks into steps. How: Exercise — take the same business problem, write a basic prompt, then layer on each lever one at a time. See how the output transforms. Pairs compete: who gets the best output from the same starting question? What If: What if the quality of every AI interaction you had doubled — not because the AI got smarter, but because you got better at asking? NESTED LOOPS: Overt = learn prompt engineering. Loop 2 = the 5 levers (context, role, format, examples, iteration) are actually communication skills — they work on humans too. They're learning how to brief anyone clearly. Loop 3 = competition format means learning from each other's prompts. Loop 4 = they discover clear thinking produces clear prompts — AI becomes a mirror for their own clarity of thought.

Why: You think faster than you type. Most people's AI usage is bottlenecked by how fast their fingers move. What: Voice-to-text tools built into phones, browsers, and AI apps. Speak naturally, let AI clean it up. Dictating prompts, brain dumps, meeting notes, emails. The workflow shift: talk > capture > refine. How: Exercise — take a complex prompt and dictate it instead. Compare speed and quality. Try a 2-minute voice brain dump, then ask AI to organize it. What If: What if you could capture every idea the moment it hit you — walking, driving, in the shower — and have it organized by the time you sit down?

Why: Everything before this was learning tools. This is using them. The difference between knowing how a hammer works and building a shelf. What: Pick your highest-value business problem from today. Apply everything: the right AI tool, conversational approach, prompt engineering, voice input. Work through it end to end. How: 60-minute working session. Facilitator floats and coaches. Pairs check in at 30 minutes to share progress and get unstuck. Final 15 minutes: each person shares their result and what surprised them. What If: What if this became your Monday morning — 30 minutes with AI on your biggest problem of the week, every week?

Why: Everything so far has been using tools other people built. The next level is building your own. And the gap is smaller than you think. What: The spectrum: using AI > directing AI > building with AI. Why now is different — AI can write the code for you. You bring the domain knowledge, the business problem, the vision. AI brings the implementation. What 'building' means in 2026 — not computer science, not 4-year degrees. Describing what you want and refining until it works. How: Demo — START BUILDING THE CLASS PORTAL LIVE. This is the cohort's actual resource site: recordings, materials, exercises, links. Build the first page in 10 minutes with AI while the room watches. The demo is real — this is their deliverable taking shape. Group discussion: what just happened? What If: What if the thing you've been waiting to hire a developer for... you could build yourself this week? NESTED LOOPS: Overt = see AI build something live. Loop 2 = the thing being built is THEIR resource — emotional investment in the demo. Loop 3 = they see the full build cycle on something real, not hypothetical. LIVE PORTAL BUILD: This is where the class portal begins — every subsequent demo adds to it.

Why: Before you can build, you need to understand where things live. Most non-technical people have never thought about how software is organized — it's like trying to cook without knowing where the kitchen is. What: Files and folders — structured with purpose. What a 'project' is (a folder with everything needed to run). Types of files: code, config, data, assets. How they reference each other. What happens when you 'run' a project. Dependencies (other people's code your project uses — ingredients in a recipe). How: Exercise — open a real simple project folder, walk through every file. Pairs create a project folder structure on paper for a hypothetical app. Name the files, discuss what goes where. What If: What if looking at a project folder felt as natural as looking at a filing cabinet — because you understood the organization system?

Why: You could write a novel in Notepad. But you use Word because it gives you spell check, formatting, and collaboration. An IDE is the same leap — for building things. What: IDE = Integrated Development Environment. What 'integrated' means: file browser, editor, terminal, AI, all in one place. Cursor and VS Code — the two to know. How AI lives inside the IDE now — it sees your files, suggests changes, writes code, runs commands. The difference between AI in a chat window vs. AI that can see your entire project. How: Hands-on — open Cursor (pre-installed). Open a simple project. Make a change using the AI sidebar. See the file update in real time. Run the project and see the result. The loop: describe > AI writes > you see it > refine. What If: What if you had a workspace where you could describe what you want, watch it get built, and tweak it until it's right — all in one place?

Why: You've learned the concepts, you've seen the demos. Now you build. Knowing how is different from doing — and the first one is always the hardest. What: Scoping: pick something small enough to finish today but meaningful enough to care about. A landing page, a simple calculator, a tool that solves a problem you have. AI does the heavy lifting — you direct, refine, and decide. What 'done' looks like for a first project. How: 60-minute build session. AI-assisted, facilitator floating. Continue building the class portal as a parallel track — participants can contribute features/pages to the portal OR build their own project. Check-in at 30 minutes — share screens, get unstuck, celebrate progress. Final 15 minutes: show what you built. It doesn't have to be perfect. It has to be real. What If: What if this was the moment you stopped saying 'I'm not technical' — because you just built something that works? NESTED LOOPS: Overt = build a project. Loop 2 = the project solves a REAL problem — actual business value. Loop 3 = scoping teaches product thinking (MVP, what's done, what to cut). Loop 4 = the experience of building installs the identity shift (the breakthrough). LIVE PORTAL BUILD: Portal continues growing — participants can add pages or features to the class portal as their build exercise.

Why: You've built something. Now imagine your laptop dies. Or you make a change that breaks everything and can't undo it. Version control means never losing work and always being able to go back. What: Git = version history for your project (like Google Docs history but for everything). GitHub = where your project lives online (backup + collaboration). Commits = save points. Branches = parallel experiments. Push/pull = syncing. You don't need to be an expert — you need 5 commands. How: Exercise — take your project, initialize git, make a commit, push to GitHub. Make a change, commit again. Look at the history. Revert a change. Feel the safety net click into place. What If: What if you could experiment freely on any project — knowing you could always undo everything and get back to a working version?

Why: A project on your laptop is a project only you can see. Deploying means anyone with a link can use it. That's when it becomes real. What: Deployment = your files go to a server, the server runs them, people visit a URL and see your thing. Hosting platforms: Vercel, Netlify — free tiers, one-click deploy from GitHub. Domains and how to connect one (yourname.com instead of random-url.vercel.app). The deploy loop: change code > push to GitHub > site updates automatically. How: Deploy the CLASS PORTAL live — the thing we've been building all course gets a real URL. Then each person deploys their own project too. Share URLs with the person next to you. This is THE moment. What If: What if every idea you had could be live on the internet within an hour of starting? LIVE PORTAL BUILD: The class portal goes live here — deployed to a real URL with a custom domain. The cohort's deliverable is now on the internet. THE DEPLOY MOMENT = THE BREAKTHROUGH.

Why: You've used AI in a chat window and AI in an IDE. There's a third mode — AI in the terminal, working alongside you like a pair programmer. It doesn't just suggest. It reads your files, runs commands, makes changes, and builds with you. What: Claude Code — what it is, how it works. The terminal as a workspace. How it differs from chat (full project context, can execute actions, iterates in real time). The shift from chatting about code to collaborating on a project. When to use chat, IDE AI, or Claude Code. How: Demo — open Claude Code on the deployed class portal, describe an improvement ('add a resources page,' 'make the recordings section sortable'). Watch it read the codebase, make changes, redeploy. Each person tries it on their own project. What If: What if your workflow was: describe what you want, refine through conversation, and deploy — all from a single terminal? LIVE PORTAL BUILD: Portal gets improved live with Claude Code — the audience sees iteration on their own resource in real time.

Why: You've been building with AI. But the AI landscape is far bigger than code — there's an entire toolshed most people don't know exists. What: The ecosystem: writing (Jasper, copy.ai), images (Midjourney, DALL-E, Gemini), video (Runway, HeyGen), transcription (Otter, Granola), research (Perplexity). AI already inside your tools (Google Workspace, Microsoft Copilot, Notion AI). Automation platforms: Zapier, Make — 'when this happens, do that' (no code). When to use chat vs. specialized tools vs. automation. How: Demo tour — show 5 specialized tools live. Then build a simple Zapier automation live. Pairs design one automation for their business on paper. What If: What if you had the right AI tool for every type of task — chat for thinking, specialized tools for creating, automation for the repetitive stuff — and you knew which one to reach for?

Why: You now know how to use AI, build with AI, and deploy. The question is: what's the strategy? Most businesses adopt AI randomly. You're going to be intentional. What: The adoption framework: (1) Audit — where do you spend time on repeatable tasks? (2) Match — which AI tools fit those tasks? (3) Sequence — what to adopt first (high impact, low risk), what to wait on. The 'AI stack' for a small business. Build vs. buy decisions. When AI is the wrong answer. How: Exercise — map your weekly tasks, identify the top 5 candidates for AI, build a 30-day adoption plan. Pairs review each other's plans for blind spots. What If: What if you had a clear, prioritized plan for AI in your business — not someday, starting next Monday? NESTED LOOPS: Overt = build an AI adoption plan. Loop 2 = the audit exercise is actually a full business process audit — most entrepreneurs have never mapped their own workflows, they leave with operational clarity regardless of AI. Loop 3 = pairs reviewing each other's plans means learning about other businesses, building network, seeing patterns across industries.

Why: This is the capstone. You didn't come here to learn about AI — you came here to become someone who builds. Now prove it. What: Every participant presents their deployed project to the group — live URL, what it does, what problem it solves, what they learned building it. 3-5 minutes each. The class portal is also presented as the collective deliverable. This is public speaking practice (nested loop), social proof (watching 10+ 'non-technical' people all show working projects destroys the limiting belief collectively), and celebration. How: Each person presents: show the live URL, walk through what it does, share one thing that surprised you. Group feedback: what impressed you? What ideas does this give you? Facilitator captures key moments. What If: What if this was the moment you realized you'll never go back to who you were before Wednesday? NESTED LOOPS: Overt = share your project. Loop 2 = public speaking practice (stacks with speaking course). Loop 3 = learning from each other's projects ('oh, AI can do THAT?'). Loop 4 = collective social proof — watching everyone succeed destroys the limiting belief at a group level. Loop 5 = the felt experience of presenting something you built to applause.

Why: Demo Day showed you what's possible. The question is: what do you do Monday morning? What: Review the full spectrum you've traveled: jargon > understanding > using > building > deploying > presenting. The portfolio approach — don't pick one project, take 20 shots on goal (connect to fail-fast nested metaphor). What to learn next based on where your energy pulled. Communities, resources, how to keep learning. The rate of change: what's possible today will be trivial in 6 months. How: Exercise — write your 30-day AI action plan: (1) which AI tools to adopt daily, (2) one automation to build, (3) one project to start. Share with a partner for accountability. Exchange contact info. What If: What if a year from now, you looked back at this week as the moment everything changed — the moment you stopped being a consumer of technology and became a builder? NESTED LOOPS: Overt = create an action plan. Loop 2 = NLP goal-setting structure applied to AI. Loop 3 = accountability pairing builds a real relationship — they leave with a network, not just knowledge. Loop 4 = the fail-fast/portfolio frame changes their entrepreneurial operating system, not just their AI usage.