Watch & learn · Hand-picked · Teacher-Checked

Understand AI — in the right order

YouTube has a million AI videos and most parents don't have time for twelve of them. So we built a path instead: the clearest explainers we could find, in the order a beginner actually needs them — what AI is, the different types, how to use it well, how it really works, and how to use it ethically with your kids. Every video is free, and every one carries a teacher's note on why it's here and what to watch for.

Teacher-Checked Video Library

Brand new to AI? Start with Path 1, videos 1–3, in order. In about 25 minutes you'll understand what AI is, the main types of AI out there, and how tools like ChatGPT actually work — more clearly than most adults ever learn it. Then jump to whichever path you need next.

1 · Start Here 2 · Use It Well 3 · Go Deeper 4 · Ethics & Family Rules

Path 1 · Start Here — AI Explained Simply

Beginner

No jargon, no math, no coding. These three videos build the foundation everything else stands on.

1. How Machines Learn

CGP GREY · ~9 MIN · START HERE

The single easiest explanation of machine learning ever made. Using simple animations of "bots" that learn from tests and data, it shows how a computer can learn to do something nobody taught it step by step — and why companies are so hungry for your data.

Teacher's note: This is the video to watch with older kids, too. Afterwards, ask: "What was the 'test' the bots kept taking?" If they can answer that, they understand machine learning better than most adults.

2. The Types of AI, Explained in One Video

IBM TECHNOLOGY · ~10 MIN · TYPES OF AI

AI, machine learning, deep learning, generative AI, foundation models — people use these words like they mean the same thing, and they don't. This clear IBM walkthrough sorts them out, shows how they evolved, and clears up the most common misconceptions (including what chatbots and deepfakes actually are).

Teacher's note: After this one, you'll be able to sort every AI tool you meet: the spam filter, the photo app, and ChatGPT are three different types of AI — and the type tells you what it can mess up.

3. How Chatbots & Large Language Models Work

CODE.ORG · SHORT · FAMILY-FRIENDLY

From Code.org's excellent How AI Works series, made for students and families: how a chatbot like ChatGPT is trained on huge amounts of text and generates answers one prediction at a time. Short, visual, and made to be understood by non-experts.

Teacher's note: The key idea to take away: the AI is predicting likely words, not looking up true facts. That one sentence explains almost every AI mistake you'll ever see — including the wrong answer key in our Bad AI vs. Teacher-Fixed post.
A family homeschooling together at a table Path 1 complete

Finished Path 1? You have the foundation.

You can now explain what AI is, name the main types, and say how a chatbot works. Next: put it to work in Path 2 — or skip straight to the ethics path if family rules are your first concern.

Path 2 · Use It Well — Practical AI Skills

Practical

Now the hands-on part: how to actually use ChatGPT day to day, and the eight-minute video that will permanently change how much you trust its answers.

4. ChatGPT Tutorial for Beginners

STEP-BY-STEP TUTORIAL · BEGINNER → CONFIDENT

A complete, calm walkthrough: creating your free account, the interface, writing and improving text through conversation, uploading documents and images to summarize and question, researching and organizing information, and building plans and checklists. It clearly flags which features are free and which need a paid plan.

Try this while you watch: pause after the basics and run one prompt from our Prompt Library — the Week Builder is a great first one. Doing beats watching.

5. Large Language Models, Explained Briefly

3BLUE1BROWN · ~8 MIN · INTERMEDIATE

Eight minutes, no heavy math, and the clearest picture anywhere of what a large language model actually computes: a probability for every possible next word. Once you see that, "hallucinations" stop being mysterious — a system built to produce plausible words will sometimes produce plausible falsehoods.

Teacher's note: This is the video behind our whole Teacher-Checked approach. AI output isn't retrieved truth; it's a well-educated guess, every single time. Check it like one.

Path 3 · Go Deeper — How AI Really Works

Advanced

For parents (and older students) who want the real mechanics. These are the most beautiful explainers on the internet — visual, honest, and worth every minute. No coding required for any of them.

6. But What Is a Neural Network?

3BLUE1BROWN · ~19 MIN · ADVANCED

The famous first chapter of the Deep Learning series: stunning visuals show what a neural network actually is — layers of numbers that learn patterns, like recognizing handwritten digits. Assumes no background; rewards your full attention.

Teacher's note: Great for a curious parent or a teen interested in how AI works under the hood. Watch it in one sitting; the payoff is a real mental model, not vocabulary.

7. But What Is a GPT? — Transformers, Visualized

3BLUE1BROWN · ~27 MIN · ADVANCED

What "Generative Pre-trained Transformer" actually means, shown step by step: tokens, embeddings, and the attention mechanism that lets a model understand words in context. This is the technology inside ChatGPT, made visible.

Teacher's note: The "attention" idea — words passing information to each other — is why ChatGPT handles context so much better than older chatbots. Worth it for that insight alone.

8. Intro to Large Language Models — The One-Hour Masterclass

ANDREJ KARPATHY · ~1 HOUR · ADVANCED

Andrej Karpathy co-founded OpenAI and led AI at Tesla — and he may be the best AI teacher alive. This general-audience talk (no code) covers how LLMs are trained, why they hallucinate, security risks like prompt injection, and where the technology is heading. If you watch one long video, make it this one.

Teacher's note: His framing — an LLM as a "dreaming" compression of the internet that you then steer — will sharpen every prompt you write and every output you judge.
Planning supplies on a desk Keep going

Want more after Path 3?

Both 3Blue1Brown and Karpathy have full free series on YouTube that continue from these videos. And Code.org's complete How AI Works series is a wonderful watch-together curriculum for older kids.

Path 4 · Ethics & Family Rules — Using AI Well

Ethics

The most important path on this page. AI raises real questions — privacy, bias, honesty, and who does the thinking. These videos help your family answer them on purpose, instead of by accident.

9. 6 Big Ethical Questions About the Future of AI

TED · GENEVIEVE BELL · ETHICS

Ethical AI expert Genevieve Bell shares six framing questions for building and using AI responsibly, safely, and sustainably — and for raising the next generation of critical thinkers. Big-picture, thoughtful, and very discussable.

Family discussion: Pick one of her six questions and talk it over at dinner. There are no grades here — the conversation itself is the lesson.

10. Ethics & AI: Privacy and the Future of Work

CODE.ORG · SHORT · FAMILY-FRIENDLY

From the How AI Works series: how to develop and use AI responsibly and fairly while respecting people's privacy and livelihoods — including what automation changes about work. A gentle, clear entry into ethics for parents and older kids alike.

Teacher's note: Privacy is the rule families can act on today: what you type into an AI tool may be stored and used to train it. First names only — or no names at all.

11. AI & Kids: Raising Responsible Kids

RAISING EAGLES · FOR PARENTS OF OLDER KIDS & TEENS

A parent-focused session on guiding teens' AI use: use AI to learn rather than just to finish, protect attention, work through challenges instead of skipping them, and keep asking "who did the thinking?" — plus why AI should never replace friends, parents, or real help.

Teacher's note: "Who did the thinking?" is the best single question a homeschool parent can ask about any AI-assisted work. Borrow it — we have.
Our 5 family rules for ethical AI use
  1. Parents run the AI. Kids benefit from it; they don't chat with it unsupervised.
  2. No personal data. Never enter a child's full name, school, address, photos with identifying details, or family information.
  3. Think first, AI second. The child attempts, wonders, and drafts before AI helps — AI sharpens thinking, it doesn't replace it.
  4. Verify before you teach it. Facts, dates, and answer keys get checked against a trusted source. AI is a first draft.
  5. Be honest about help. In portfolios and evaluations, note when AI assisted. Modeling honesty is the curriculum.

Coming soon — videos from Brian

The library above is the best of the internet, curated. These are the walk-throughs I'm filming myself, showing our exact prompts and routines at a real kitchen table.

Planning supplies on a desk Coming soon

Plan a Week in 30 Minutes — the Full Walkthrough

Watch the Sunday planning routine from start to finish, prompts included.

A family homeschooling together at a table Coming soon

5 Prompts Every Homeschool Parent Should Know

The greatest hits from the Prompt Library, demonstrated live.

A child reading under a tree Coming soon

Build a Reading List Your Child Will Actually Love

Level-matched book lists built around your child's real interests.

Planning supplies on a desk Coming soon

Portfolios Without the Panic

A simple records system that takes ten minutes a month.

Get new videos first

Join the free list

New videos, posts, and prompts — one email a week, plus the free Starter Kit today.

Free forever · Unsubscribe anytime