Book 1 · Ch 1: What AI Actually Is (and Isn't)
Book 1 · Chapter 1: What AI Actually Is (and Isn't)
This page is an interactive demo of how AI pattern-matching works, plus the chapter's four-terms card and a set of starter prompts to try.
This page assumes you've read the chapter. It hands you the tools, not the lessons.
ML Pattern Demo
Pick a sentence and press play. Watch the model guess the next word, one at a time, from the words before it. That guessing, scaled up, is the whole trick.
The glowing words are the ones it leaned on.
The doctor examined the patient.
Scale this to hundreds of billions of words. That's the engine.
Download · PDF
Four AI Terms
- Machine learning
- How AI learns. Instead of programming in rules, you show the system thousands of examples and let it figure out the pattern on its own. Like learning to ride a bike: try, fail, adjust, repeat, get better.
- Large Language Models (LLMs)
- The engine behind Claude, ChatGPT, and Gemini. “Large” because they trained on a ridiculous amount of text, “language models” because text is their native language. They also work with images, data, and code.
- Generative AI
- AI that creates new stuff instead of just sorting existing stuff. Every email it drafts, every report it writes, every image it generates.
- Agentic AI (agents)
- When AI stops waiting for you to hold its hand through every step and handles multi-step tasks on its own. You say “research these three vendors, compare their pricing, and draft a recommendation,” and it does the research, builds the comparison, and writes the draft.
From the Jargon (60 Seconds, Then I Promise We're Done) section.
Try It: the prompts that work
- The Accountant
I'm a CPA at a 12-person firm. I need an engagement letter for a new audit client, a mid-size nonprofit, $5M revenue. They've never been audited and they're nervous. The letter needs to be professional but reassuring, cover scope and fees ($15K), and include our standard liability language.
- The Marketing Manager
I run marketing for a B2B SaaS company that sells project management software to construction firms. LinkedIn and Twitter. Audience: ops managers and project leads, not C-suite. Need a content calendar for next month, 3 posts per week. Big push right now: new mobile app feature. Tone: practical, not corporate.
- The Operations Manager
I'm a regional ops manager at a logistics company, three warehouses, 200 employees. Order-to-delivery time went up 15% this quarter because of picking inefficiencies at our biggest facility. I need a proposal for my VP. She likes data-first with clear ROI projections. Give me the bottleneck analysis, two or three quick wins for the next 30 days, and cost/time savings estimates.
- The HR Professional
I need a mid-year review for a junior graphic designer. She's been here 8 months. Great on creativity and deadlines, but she gets defensive when she gets feedback. We use a 1-5 scale with specific examples required for each competency. I want it encouraging but about the feedback issue.
- The Real Estate Agent
I'm a residential real estate agent listing a 3-bedroom, 2-bath colonial in a walkable neighborhood near downtown. Built in 1985, recently renovated kitchen with quartz countertops and stainless appliances, original hardwood floors throughout, fenced backyard, two-car garage. Target buyers are young families relocating from out of state. The neighborhood has a top-rated elementary school two blocks away and a weekend farmers market. Write a compelling MLS listing description under 250 words that leads with lifestyle, not just specs.
- The Weekend Trip Planner
My partner and I want a three-day weekend trip in September. We're driving from Austin, max four hours. Budget around $1,500 total. We like hiking, good food, and local breweries, but we don't want a city trip, we want to disconnect. We've already done Fredericksburg twice. Give us three options with a rough itinerary and cost breakdown for each.