before you build

You didn't get replaced because you weren't good

You got replaced because the math changed. A model that costs a fraction of a cent per call now does the parts of your job that scale — reading, summarizing, classifying, drafting, routing. The parts that don't scale — taste, judgment, verification, accountability, owning the relationship — got cheaper too, because the model offloaded the bulk and made the remaining work look small.

If you're here because you got cut, you're not behind. You're early. Most of the workers who'll be cut in the next 18 months haven't been cut yet. This chapter is for you.

What this chapter is, and what it isn't

This is not a "gentle intro to coding." It's not a TED talk about how exciting AI is. It's not a pep rally. The course this chapter introduces runs 30 chapters and ends with you engineering production agent harnesses. Ch00 is what makes ch01 possible: identity repair, mental model installation, and the honest naming of where this is all going.

You'll learn:

• What an LLM actually is — not a search engine, not a brain, not a database. A next-token probability function over text, with frozen weights and no memory between sessions. Every later chapter assumes you know this. Most people don't.
• How to talk to one — prompting is briefing. If you've ever written a creative brief, a customer-service script, a contract clause, or an SOP, you've already done 80% of this. Ch03 names the other 20%.
• Where this course goes — and which path through it fits your background. The displaced engineer skips different chapters than the displaced copywriter. We'll map both.

What AI specifically gets wrong about teaching this

• Treats ch00 as a softer ch01. Engineers instinctively frontload syntax. For displaced mid-career adults, syntax-first is catastrophic — it confirms the fear ("I'm too old / too dumb") in five minutes. Ch00 ends with code, but doesn't open with it.
• Calls you "non-technical." You wrote brand voice for a national ad agency. You ran a 12-person support team. You did doc review for a top-50 law firm. You are not non-technical. You are differently technical, and your existing skills are exactly the scarce ones.
• Sells a hallucination as the fix. "Anyone can code now! AI makes it easy!" casually celebrates the mechanism that ended your career and assumes your pain point was difficulty rather than displacement. We're going to do this honestly or not at all.

What you'll be able to do at the end

By the end of ch00 you'll have:

• Said out loud, in writing, what skills from your old work transfer into this one. (Spoiler: more than you think.)
• A correct enough mental model of what an LLM is that ch01-30 stops feeling like magic.
• Written your first prompt with the structure professional prompt engineers use, and seen why structure beats vibes.
• A 30-day plan for which chapters to read in what order, calibrated to where you came from.

After this chapter, you stop being "someone who lost their job to AI" and become "someone who builds with it." The grief is real and the course will not pretend otherwise. But the leverage is also real, and that's what we're here for.