AIQ
AIQ
Everything a teacher needs to deliver this lesson — pick your grade's script below once you've read the background.
This is the closing lesson of World 4, and it deliberately steps back from "here's a cool AI application" (the shape of the five lessons before it) to "here's what this means for your life." Keep that shift in mind — the content is shorter than a typical AIQ lesson, but the conversation it opens is bigger. By the end, a student should be able to:
You do not need an economics background to teach this lesson well, but it helps to have one clear mental model going in, because almost every question a student asks will circle back to it: a job is not one thing — it's a bundle of many different tasks, and AI is good at automating some kinds of tasks and bad at others. Once you have that idea solid, every example in the lesson becomes a case of sorting a specific task into one bucket or the other.
The tasks AI is good at tend to be routine and pattern-based: reviewing thousands of similar documents for a specific clause, generating draft images from a text description, scanning data for an anomaly, transcribing speech. These are tasks with a large amount of past example data and a fairly consistent "right answer" — exactly the conditions that pattern-matching systems are built for, which is the same idea AIQ teaches throughout every other world lesson. The tasks AI is bad at tend to be the opposite: they require an original idea nobody has had before, they depend on reading another person's emotional state and responding with genuine care, or they happen in a messy, constantly-changing physical environment where no two instances look quite the same (which is why the lesson uses a plumber, not a factory-line worker, as its "AI can't easily replace this" physical-labor example — a plumber deals with a different, oddly-shaped, decades-old pipe configuration in every house).
The lesson's "new jobs" examples are real and worth being able to describe accurately. Prompt engineering is the practice of writing and refining instructions to get better, more reliable results out of an AI system — it became a distinct skill (and in some organizations a job title) because getting good output from these tools is a skill that takes practice, not something anyone can do perfectly on the first try. AI ethics officer (or "responsible AI" roles, under various titles) is a real and growing function inside companies that build or deploy AI — someone whose job is checking that a system is fair, safe, and doesn't cause foreseeable harm before it ships. AI trainer refers to the very real, very large workforce of people who label data, rate AI outputs, and correct mistakes so a model can improve — most AI systems in use today, including the kind AIQ itself is built to teach about, were shaped by large amounts of this kind of human feedback work.
A specific number worth flagging before it comes up: the lesson's Explorer-through-Builder hook mentions a prompt-engineering salary figure ("$100K+"). Treat this the way you'd treat any single reported salary figure — it reflects some real, high-end reported job postings for this title, not a typical or guaranteed number, and it will read differently to a Philippine classroom than to the market it was likely drawn from. It's fine to use it as "this is a real job people get paid well for," and worth being upfront that pay varies enormously by employer, location, and experience — exactly like any other job title.
For Hacker and Architect classes, two more precise ideas are worth having ready, because the 11–14 and 15–18 quizzes both use this vocabulary directly. The "augmentation hypothesis" is the idea, associated with labor economists studying automation, that AI more often enhances what a human worker can do than replaces the worker outright — a claim clerk doesn't disappear when AI helps process claims faster, but their job shifts toward the parts of the work that still need human judgment. Related to that, economists including David Autor have argued for analyzing automation at the level of individual tasks within a job rather than treating "the job" as one all-or-nothing unit — decomposing a role into its component tasks and asking, task by task, how automatable each one is. This is a genuinely useful lens for a 15-to-18-year-old thinking about their own future career: the right question isn't "will my future job be automated," it's "which parts of it are, and what does that leave for me to be good at."
One more term worth knowing before an Architect-level discussion: the lesson's quiz touches on "entry-level compression" — the observation, discussed in recent labor-market commentary, that AI tools can perform tasks traditionally handed to junior employees (drafting a first version, doing preliminary research, writing routine code), which may shrink the number of true entry-level openings even in fields that aren't shrinking overall. This is a live, debated concern in current economic discussion rather than a settled, long-established finding — present it to older students as "something economists and employers are actively watching and arguing about right now," not as a proven law of how AI affects the job market.
Nothing beyond AIQ's usual requirements: one device per student (or per small group) with a modern browser and an internet connection to open aiq.ph. No login is required, no printouts, and no advance setup — the lesson runs entirely inside the app in one 5–15 minute sitting. If you plan to run the discussion questions as a whole-class conversation, budget an extra 5–10 minutes, and it's genuinely useful to have the board free — several of the discussion prompts below work well as a two-column "jobs AI changes" / "jobs AI mostly can't touch" list built with the class as you go.
For Builder mode and up, the in-app lesson includes a short timed sorting activity where students sort job-related tasks into "AI" or "not AI" as quickly as they can — this runs automatically inside the app after the main scenes, needs no extra prep, and works fine solo or in pairs on one device.
No specialized vocabulary needs pre-teaching for Explorer or Builder. For Hacker and Architect classes, it's worth previewing two terms before starting if you have a spare minute: "automation" (a machine or program doing a task a person used to do) and "task" as distinct from "job" (a job is made of many tasks, and they don't all behave the same way when AI enters the picture) — both quizzes for those age bands build directly on that distinction.