Chapter 1 — The AI Revolution

“Artificial intelligence is the new electricity.” — Andrew Ng

“what is man that you are mindful of him, and the son of man that you care for him?” — Psalm 8:4 (ESV)

Why This Matters

Welcome to the course. My name is Coach, and for the next sixteen weeks I am going to make you something specific: not a programmer, not a data scientist, not a “prompt engineer” — a professional who is fluent with AI. Someone who can walk into any office, clinic, classroom, or shop floor and know which of these tools to reach for, how to get real work out of it, and — this is the part almost nobody teaches — how to check what it hands back before you put your name on it.

You do not need to write a line of code in this book. Not one. The tools we use are the same ones sitting on your phone right now: a browser, a chat window, a microphone. What you do need is judgment, and that is the one thing the machine cannot supply. So that is what we will train.

Here is why the timing is not optional. For most of the last seventy years, artificial intelligence was a lab curiosity and a movie villain. Then, in the space of about eighteen months starting in late 2022, it became a thing that a receptionist, a nurse, a middle-school teacher, and a landscaping-company owner all use before lunch. That is not normal. Technologies usually arrive over decades. This one arrived over a season, and it arrived unevenly — which means the professionals who understand it early get a real, measurable edge, and the ones who wait get managed by the ones who didn’t.

Let me put the whole book on one line, and I want you to feel the weight of it, because everything else hangs here:

You choose the tool. You own the verdict. AI drafts, generates, and accelerates; the professional decides, verifies, and is accountable. AI is an assistant, not an authority.

Read that twice. I will restate it a hundred different ways over sixteen weeks — I will call it the spine rule — but it never changes shape. The person who thrives in this era is not the one who uses AI the most. It is the one who knows when to use it, which tool to reach for, and how to verify what it gives back — keeping human judgment exactly where the stakes are real.

And underneath the practical work runs a deeper question, one I will not preach at you but will not dodge either. We have built machines that write, argue, draw, and converse — machines that produce the outputs of a mind. So it is worth asking, honestly and carefully: what are we, that we can build such things — and what is a made intelligence next to a human being? Hold that question. We will give it its due at the end of the chapter, and it will change how you carry the tool.

Coach’s Note — Here is the house rule, and it is the reason this library works. Learning is a sport. The keyboard is the gym. You do not get fit reading about squats, and you will not get fluent with AI reading about it. Every chapter ends with Reps and a Project. There is an interactive lab embedded right below this page. Do the work with your hands or this book is just decoration.

1.1 — What We Actually Mean by “AI”

Let us define the word before we build a course on it, because “AI” has become a marketing sticker slapped on everything from a thermostat to a toaster.

Artificial intelligence is the broad project of getting computers to do things that, when a human does them, we call intelligent — recognizing a face, understanding a sentence, planning a route, spotting fraud, writing a paragraph. That is it. It is an umbrella term, and it has been around a long time.

The name itself dates to the summer of 1956, when a small group of researchers gathered at Dartmouth College and coined “artificial intelligence” for their new field. Six years before that, the mathematician Alan Turing had already asked the question that still haunts the whole enterprise — can a machine think? — and proposed a famous test: if you could hold a conversation through a screen and not tell whether a human or a machine was answering, on what grounds would you deny it intelligence? Keep that image; we will come back to it, because in 2026 that test suddenly stopped being hypothetical.

For its first few decades, AI mostly meant rules. Humans wrote down expert knowledge as long chains of “if this, then that,” and the computer followed the rules. This worked for narrow, tidy problems — playing checkers, diagnosing a specific machine fault — and fell apart on anything messy, like understanding a photograph or an ordinary sentence. Funding dried up twice, in periods historians of the field now cheerfully call the “AI winters.” For a working professional in 1995, “AI” was a punchline.

What thawed the winter was a change of strategy so important it gets its own section.

1.2 — Machine Learning vs Deep Learning (the nesting dolls)

The breakthrough was to stop writing the rules and start learning them from examples. This is machine learning, and it is the engine under almost everything you will touch in this book.

The difference is worth feeling concretely. Suppose you want a program to tell cats from dogs in photos. The old way: a human tries to write rules — “cats have pointed ears, dogs have longer snouts…” — and drowns, because for every rule there are ten thousand exceptions. The machine-learning way: you show the computer a hundred thousand photos already labeled cat or dog, and it figures out the patterns itself, adjusting as it goes until it gets good at guessing. Nobody wrote the rules. The machine learned them from the examples. That is the whole idea, and it is genuinely different from ordinary software.

Three terms nest inside each other like Russian dolls, and professionals mix them up constantly, so let us get them straight:

Deep learning is machine learning done with neural networks — software loosely inspired by how brain cells connect, arranged in many stacked layers (hence “deep”). You do not need the math. You need the intuition: each layer learns to spot slightly more complex patterns than the one below it — edges, then shapes, then “that’s an ear,” then “that’s a cat.” The public turning point was 2012, when a deep neural network shattered the record on a giant image-recognition contest and made it undeniable that this approach worked. From there, deep learning ate speech recognition, translation, and eventually language itself.

Coach’s Note — When a vendor says “our product uses AI,” a fair follow-up is “machine learning, or just some if-statements someone’s calling AI?” Both are legitimate; they fail differently. Rule-based systems break loudly and predictably. Learned systems break quietly — they give you a confident, plausible, wrong answer and no error message. The rest of this book is largely about the second failure mode.

1.3 — The Transformer Breakthrough

Everything you have heard about in the last few years — ChatGPT, Claude, Gemini, the whole generative wave — traces back to a single 2017 research paper out of Google with a title that is now a nerd punchline: “Attention Is All You Need.” It introduced a neural-network design called the Transformer.

You do not need the internals. You need to know what the Transformer unlocked, because it explains why AI felt like it went from zero to everywhere overnight.

Older language models read text one word at a time, left to right, and tended to lose the thread over long passages — like reading a contract through a keyhole. The Transformer’s key trick, called attention, lets the model look at all the words at once and learn which words matter to which — that “it” in a sentence refers back to “the dog,” that this clause modifies that noun three lines up. That sounds modest. It was seismic, for two reasons:

1. It scales. Transformers train efficiently on enormous amounts of text using modern hardware. Feed one a large fraction of the public internet, and it gets startlingly good at predicting what word comes next — which, it turns out, is most of what fluent writing is.
2. It generalizes. A model trained just to predict the next word learns, as a side effect, to summarize, translate, answer questions, draft emails, and write code — none of which anyone explicitly programmed.

That is the “T” in GPT: Generative Pre-trained Transformer. When you type into a chatbot in 2026, you are talking, at bottom, to a very large Transformer that has read an enormous amount of human writing and is predicting the next chunk of text one piece at a time. We will open that hood carefully in Chapter 2. For now, hold the headline: one architecture, invented in 2017, is the seed of the entire revolution.

Coach’s Note — File this away for when a confident answer dazzles you: the system is, mechanically, an extraordinarily sophisticated next-word predictor. That it can draft a passable legal summary is genuinely astonishing. That it can invent a court case that never existed, in the same fluent voice, is the exact same machinery doing the exact same thing. Fluency is not knowledge. We will hammer this all term.

1.4 — Why Everything Changed After 2022

The Transformer arrived in 2017, but for five years it stayed mostly inside labs and developer tools. Then, on November 30, 2022, a research lab called OpenAI put a free chat window on the public internet and called it ChatGPT. Anyone could type into it in plain English and get a coherent answer back. No manual, no code, no cost.

That was the hinge of the decade. ChatGPT was, by widely reported accounts, the fastest-growing consumer application in history at the time — reaching an estimated hundred million users in about two months. But the number is not the point. The point is access. For the first time, the frontier of AI was not a research demo or an enterprise API; it was a text box your aunt could use. The technology had existed. The doorway was new.

Since that hinge, the field has moved on several fronts at once. As of mid-2026, these are the trends a professional should be able to name (I am dating this deliberately — every specific below will drift, so learn the direction, not the version number):

• Multiple serious players, not one winner. No single “best AI” — instead a handful of frontier labs each shipping a family of models. We map the whole landscape in Chapter 3; for the working list of who’s who, see the AI Tool Directory (Appendix B).
• From text to everything (multimodal). The tools no longer just read and write text; they see images, hear audio, watch video, and generate all three.
• Reasoning models. Newer models can “think before they answer” — working through a problem in steps — which helps on hard, multi-step tasks at the cost of speed and money.
• Agents. The leap from a tool that answers to a tool that acts — booking, browsing, filling forms, completing multi-step jobs. This is the big story of the second half of our course (Chapter 13).
• Cheaper and smaller. Capable models now run on a laptop, offline, for free — which matters enormously for privacy (Chapter 11).

Coach’s Note — Notice what did not change after 2022. The machine still predicts; it still hallucinates; it still cannot be held accountable. What changed is that the doorway opened and hundreds of millions of us walked through it. A revolution in access is exactly the kind that rewards the person who brings judgment — because everyone now has the tool, and almost nobody has been taught to verify it.

1.5 — The Six Types of AI

Walk into any conversation about AI and you will hear these six words used loosely. Let us make them yours. I will give you a plain definition, an everyday example, and — because this is the whole game — what to watch for with each.

One honest caveat first: these are not airtight scientific categories. They are practical lenses. A single product is usually several at once — the assistant on your phone is conversational and multimodal and agentic in the same minute. The cleanest real line is between the first two; the rest are mostly flavors of generative. The reason to name the type is not to win a debate — it is to know what the thing is doing so you know what to verify, because each type fails differently.

1. Predictive AI — it guesses a label or a number. It learns from past data and then forecasts, scores, ranks, or classifies something that already exists. Example: your bank texting “Was this you?” on a suspicious charge; the spam folder; the “you might also like” row. Watch for: bias baked into old data, and a confident score on a case it has never really seen.

2. Generative AI — it makes new content. Text, images, audio, video, code — produced fresh from a prompt. Example: “write a thank-you note to a donor” and you get a finished draft; “a golden retriever as an astronaut” and you get a picture. Watch for: hallucination — fluent, confident, and simply false, with no tell in the tone.

3. Conversational AI — it talks back, in turns. It holds a natural back-and-forth, remembering the last several messages so you can say “shorter” and “now more formal” without re-explaining. Example: the chat mode of ChatGPT, Claude, or Gemini; a customer-service chatbot; a voice assistant. Watch for: it sounds like a colleague, so you trust it like one. It is a superb pattern-matcher, not a person who knows things.

4. Multimodal AI — it reads and makes more than text. It takes in and produces several media types at once. Example: snap a photo of a broken part, a menu, or a rash and ask “what am I looking at?”; point it at a chart and ask what it says. Watch for: it can misread an image as confidently as it misreads a sentence. Verify what it claims to “see.”

5. Agentic AI — it takes multi-step action toward a goal. It plans a sequence, uses tools (browse, click, fill, send), and works toward an outcome with less step-by-step hand-holding. Example: “find three dog-friendly hotels near the venue under $180 and put them in a table” — and it goes and does it. Watch for: it can take real actions that cost money or leave the building. Delegation never transfers accountability; keep a human approval gate on anything that spends or sends.

6. Physical / Robotic AI — it has a body in the world. AI embodied in hardware that senses and acts physically. Example: the robot vacuum, the self-driving feature, the warehouse picker, early humanoid robots. Watch for: failures here can hurt people or property, not just an inbox — the highest stakes and the strictest oversight.

Keep the one-page version handy: code/ai-types-cheatsheet.txt. You will sort real products into these six buckets in this week’s lab, and you will notice fast that the interesting products live in several buckets at once. That is not a flaw in the taxonomy; it is the taxonomy telling you the product has several ways to go wrong.

1.6 — Where AI Already Lives (across your profession, and everyone’s)

Here is the thing most people miss: you were using AI for years before you ever opened a chatbot. It was quietly running your spam filter, your maps ETA, your photo search, and your bank’s fraud alerts. The generative wave of 2022 did not introduce AI to your workday. It made a piece of it conversational — something you talk to instead of something that hums in the background.

So let us look across the professional world, because part of literacy is knowing that this is not a “tech industry” story. It is everywhere already:

Two lessons from this table. First, AI is horizontal, not vertical — it is a general capability showing up in every field, not a product for one industry. Second — and this is the professional’s edge — the value lands where a person applies judgment on top of it. The radiologist still signs the read. The teacher still knows the child. The analyst still owns the recommendation. The AI compressed the grunt work; it did not absorb the accountability. That pattern is the whole book, and it is why your inventory worksheet, code/ai-in-my-work.txt, starts by asking not “what does AI do” but “where does the judgment live in my week.”

1.7 — AI as the New Electricity: A General-Purpose Technology

Now step all the way back, because I want you to size this correctly. AI is not “a useful app.” Economists have a name for a rare kind of invention that does not just do one job but reshapes every job: a general-purpose technology. There are only a handful in all of history — the steam engine, electricity, the computer, the Internet. Each shares a signature: it starts as a specialist novelty, spreads into every industry, and quietly rewires how ordinary work gets done.

Think about electricity. When it arrived, it did not merely replace the candle. It let the factory abandon the water wheel, moved the assembly line, put a motor in the kitchen, and lit the night — changes nobody predicted from watching the first light bulb. The factory owners who thrived were not the ones who bolted an electric motor onto their old steam-shaped factory. They were the ones who rethought the work around what electricity made newly possible.

That is the claim behind Andrew Ng’s line at the top of this chapter — “AI is the new electricity.” Not that AI is magic, but that it is a general capability you will pipe into nearly everything you do, the way you pipe electricity into every room without thinking about the power plant. And the professional lesson is the same one the factory owners learned: the edge does not go to whoever merely adopts it. It goes to whoever rethinks the work around it — and keeps a hand on the switch.

Coach’s Note — Every general-purpose technology also arrived with a con man attached. Electricity brought quack “electric” health cures; the Internet brought the dot-com bubble and the phishing email. AI’s version is the confident-and-wrong answer and the vendor selling “AI-powered” vapor. A general-purpose technology rewards adoption and punishes credulity. Both halves are real. Your job is to hold both.

1.8 — The Spine Rule: You Choose the Tool, You Own the Verdict

I told you I would restate this all term, so here it is with its full weight, because it is the difference between a professional who is empowered by AI and one who is quietly replaced by their own carelessness.

You choose the tool. You own the verdict.

Unpack both halves. You choose the tool — which means fluency is a real skill. There is no single “best AI”; there is the right one for this task, and knowing the difference between reaching for a cheap fast model, a careful reasoning model, an image tool, or a research agent is a competence you will build over sixteen weeks. Reaching for “the one everybody mentions” for every task is the mark of the amateur.

You own the verdict — which means the accountability never moved. AI drafts; you decide. AI generates; you verify. AI accelerates; you answer for the result. When the email goes out, the diagnosis is filed, the report is submitted, the number lands in the budget — your name is on it, not the model’s. “The AI said so” is not a defense a professional gets to use, because a machine cannot be held accountable and you can. That is not a limitation of today’s models that next year’s version will fix. It is a fact about responsibility, and responsibility does not run on silicon.

So the sentence to carry out of Week 1 is short: AI is an assistant, not an authority. An assistant you would be foolish not to use, and foolish to obey without checking. We will practice that stance in every chapter — drafting then verifying, generating then deciding, delegating the typing and keeping the judgment.

1.x — Interactive Lab: Type-of-AI Sorter

Below this chapter on the website you will find an interactive panel called the Type-of-AI Sorter. Go use it now — it is not decoration, it is the rep that wires this lesson into your hands.

The Sorter shows you nine real, recognizable products — Netflix’s “you might also like” row, a bank fraud alert, GPT Image 2 (ChatGPT’s image maker), ChatGPT in plain chat mode, Siri, a snap-a-photo identifier app, an autonomous research agent, a self-driving car, and a warehouse picking robot — and asks you to drop each into the right bucket: predictive, generative, conversational, multimodal, agentic, or physical. When you commit, it scores you and, for each item, explains why it lands where it does — including the ones that honestly belong in two or three buckets at once.

What the Sorter teaches is not a list of six definitions you could memorize from the cheat sheet. It teaches the reflex of looking at any new AI product and immediately asking the right question: what is this thing actually doing, and therefore what would I have to verify before I trust it? A predictive system and a generative one fail in completely different ways; a physical one fails at completely different stakes. By the time you have sorted all nine products, you will do this instinctively — which is exactly the instinct that keeps a professional out of trouble for the rest of the course.

Run it twice. The first time, sort cold. The second time, keep code/ai-types-cheatsheet.txt open beside you and notice which products you had to think hardest about — those “belongs in three buckets” tools are the ones you will meet most in your real work.

1.9 — What Are We, That We Build Minds?

Now the week’s question, given its due. We have built machines that write, argue, draw, and converse — that produce the outputs of a mind. That is a genuinely new thing under the sun, and it presses an old question back onto us with fresh force: what are we, that we can make such things — and what is a made intelligence next to a human being?

King David asked the first half of it three thousand years before a Transformer existed. Lying under a night sky, staggered by the scale of it, he wrote: “what is man that you are mindful of him, and the son of man that you care for him?” (Psalm 8:4, ESV). Notice what he is astonished by. Not human greatness — human smallness. Against the moon and the stars, man is a speck. The wonder of the psalm is not that we are impressive; it is that the God who made the galaxies is mindful of us anyway, and — David goes on — has crowned this speck with glory and honor and given him dominion over the works of God’s hands. Our dignity in Psalm 8 is not achieved. It is conferred. It is a gift, from outside us.

Hold that against what we have just built. A large language model is, with no disrespect intended, the work of human hands — data, electricity, and mathematics, arranged by people. It produces sentences, and until recently producing sentences was something only a mind could do, so it is tempting to conclude we have made a mind. But look closely. The model predicts the next word. It does not know what the words mean, in the way you know that “your mother is ill” is grief and not grammar. It does not will, or love, or repent, or hope. It cannot stand before God. It cannot be held accountable, which is why — remember the spine rule — you still own the verdict. As of mid-2026 it is a mirror of human writing so polished that we keep mistaking our own reflection for a face.

This is where the Christian confession says something the culture around you cannot. Scripture teaches that the human being is made “in the image of God” (Genesis 1:27) — a creature, body and soul, addressed by God, answerable to God, capable of faith and love and vocation. That image is not a score on a benchmark. It is not fluency, or reasoning power, or the ability to draft a passable email — because if it were, then a sharper model would make you less of a person, and a stroke that dimmed your memory would make you less an image-bearer, and the confession will not allow either. The image of God is a status conferred by God, held by the newborn and the dying and the disabled and the ordinary, and it does not flicker when a machine out-writes you. Your worth was never in being the only thing on earth that could produce a good paragraph. It was in being known and loved by the One who made you — which is exactly what David could not get over.

So the honest answer to “what are we, that we build minds?” is bracing in both directions. We are astonishing — image-bearers given real dominion, and building intelligences is a staggering, God-imaging exercise of that dominion. And we are not God — what we make is a tool, not a soul; a very good mirror, not a neighbor. The ancient warning about idols was that people carve something with their own hands and then bow to it, and the tragedy was always that they served the work of their own hands as if it could see and know and save. The modern version is subtler, because this work of our hands really can talk back. Which makes the discipline of this whole course quietly theological: do not bow to the thing you made. Use it — wield it well, gratefully, as a gift of a good creation. But keep the verdict, the accountability, and the worship where they belong. That is what it means to be a person holding a tool, and not the other way around.

We will return to this question, in a different key, in every chapter. This week, just let it reframe the whole enterprise: you are not competing with the machine for the right to matter. Your mattering was settled before the machine existed.

1.10 — Common Pitfalls

Pitfall: Trusting an AI answer because it sounds authoritative. Example: You ask a chatbot for the return policy in a regulation and it cites a specific, real-sounding section number that turns out not to exist. You quote it to a client. Fix: Tone is not evidence. A generative model produces true and false sentences in the exact same confident voice. Verify every fact, name, number, and citation against a real source before you rely on it. We build this reflex all term; it is the point of Chapter 8.

Pitfall: Saying “I use AI” as if that were a skill. Example: You reach for the same general chatbot for a legal summary, a photorealistic image, and a private client record — and get a mediocre result on the first, a weak one on the second, and a privacy problem on the third. Fix: You choose the tool. There is no single best AI. Match the tool to the task — a skill you will build across Chapters 3 and 5 and the Tool Directory (Appendix B).

Pitfall: Assuming the type of AI, and therefore missing how it fails. Example: You treat an agentic tool like a harmless chatbot and let it “just handle” a task that quietly spends money or sends an email on your behalf. Fix: Name the type before you trust it (predictive / generative / conversational / multimodal / agentic / physical). Each fails differently and at different stakes. Keep a human gate on anything agentic that spends or sends.

Pitfall: Pasting confidential or personal information into a consumer AI tool. Example: You drop a client’s medical details, a student’s record, or an unreleased financial figure into a free chatbot to “just summarize it.” Fix: Never paste secrets, PII, or confidential material into a consumer AI tool without knowing your organization’s policy and the vendor’s data terms. This is serious enough to have its own chapter (14) and its own appendix — read Responsible & Ethical AI Use (Appendix C) before Project 1.

Pitfall: Believing the hype cycle in both directions. Example: Either “AI will do my whole job by Friday” or “it’s just autocomplete, a fad.” Both are postures, not assessments. Fix: Hold the middle. AI is a genuine general-purpose technology and a confident fabricator. Learn the specific capability and the specific failure mode of the specific tool in front of you — that is what literacy is.

Pitfall: Chasing version numbers instead of learning the durable skill. Example: You memorize which model is “the best” this month and feel lost the moment a new one ships. Fix: Learn the categories and the judgment, not the SKU. Models turn over weekly; the disciplines — choose the tool, verify the output, own the verdict — outlast every version number in this book.

1.11 — Reps

The work is in the exercises. The keyboard is the gym, and this is where Week 1 actually gets into your hands. Most of Week 1’s reps are about seeing the AI already in your life and learning to name what it is doing — no sign-ups required. For the two reps that do reach a live assistant, you already have one: sign in to BoodleBox — the AI workspace Concordia licenses for the whole campus — at box.boodle.ai with your Concordia (EDU) account. A preview of what is waiting:

• Inventory the AI already in your workday with code/ai-in-my-work.txt — most people find six or seven systems they never counted.
• Sort ten real products into the six types, and defend the ones that belong in more than one bucket.
• Catch the confident-but-wrong answer in a first chat with BoodleBot — your first, gentle taste of hallucination.
• Find where the judgment lives in your own week — the one task whose failure would cost the most.
• Explain “AI is the new electricity” to a skeptical friend in five sentences, then poke holes in your own analogy.

A short Check Your Reps quiz is embedded on this page, right under the chapter — five questions grounded in exactly what you just read. Take it before you move on.

1.12 — This Week’s Project

Your first project is P1 — “First Contact,” specified in Project 1. You will make contact with the three major assistants — this term, all inside BoodleBox — run the same real work task on all three, journal the differences with a cold and honest eye, and classify five AI tools you already use by type. It is a small project on purpose — this week is about making contact, forming your own first impressions, and starting the habits (compare, verify, decide) that every later week will lean on.

In BoodleBox: first contact is one door. Go to box.boodle.ai, choose “Sign in with your EDU account,” and use your Concordia login — no new password, no three separate sign-ups. You land in a chat with BoodleBot, the default assistant; type @ to open the bot picker and pull ChatGPT, Claude, and Gemini into that same chat to compare them side by side. The setup on-ramp is Setting Up Your AI Toolkit (Appendix A), and the safety rules — what never to paste — are in Appendix C. (No campus access? The same task runs on the free public versions of the three tools.)

The Hard tier asks for the one thing no AI can do for you: a written recommendation — which assistant you will adopt as your daily default, and why — defended from your own evidence. That judgment call is where the thesis gets graded.

1.13 — Coach’s Final Word

Here is what I want you to carry out of Week 1. Something genuinely rare happened in your working lifetime: a general-purpose technology — the kind that shows up only a few times in all of history — arrived in a season instead of a generation, and it landed in a text box that anyone can open. That is not hype. That is the actual shape of the moment you are standing in.

But the moment does not reward the loudest adopter. It rewards the one with judgment. The machine took over the typing — the first draft, the rough summary, the quick sort — and in doing so it pushed the human work up the stack, to exactly where it always should have lived: choosing wisely, verifying honestly, deciding and answering for the result. That is harder than the typing was, and it is far more durable. Half the model names in this chapter will carry a higher version number by the time you read it twice. The disciplines will not. Choose the tool. Verify the output. Own the verdict.

And underneath it all, the quiet thing worth more than any of the tools: you were never in a contest with the machine for the right to matter. A made intelligence is a remarkable work of human hands and a genuine gift to use well. It is not a soul, not a neighbor, and not an authority. You are the one made in the image of God, entrusted with the judgment — which is exactly why you keep the verdict. Wield the tool. Do not bow to it.

Now go do the reps. The Sorter is waiting right below this page, the worksheet is in code/, and Project 1 is where it all comes together.

See you on Monday.

Up next: Read the exercises and do all of Week 1’s reps, then build Project 1 — Project P1: First Contact. Sign in to BoodleBox with your Concordia account (setup help in Appendix A), keep the Tool Directory (Appendix B) and the Glossary (Appendix D) open as references, and read Responsible & Ethical AI Use (Appendix C) before you paste anything into a chatbot. Then Chapter 2 — How AI Models Actually Work.