Week 1 · Lesson 3 · 45-60 minutes

Week 1 Device Investigation

Investigate the devices and apps around you, map their inputs and outputs, and uncover the human decisions behind the AI features you rely on.

Materials: This lesson in a web browser · Paper and pencil, or a notes app
This lesson: In progress
Auto-saves on this device

What you'll be able to do

Investigate real devices and apps to find AI features.
Map the input and output of an AI feature you use.
Identify the human decisions behind an everyday AI.

Think about it

You probably used several AIs before lunch without noticing. Which apps on a phone or laptop do you think have a hidden AI feature inside them?

Make a prediction

Predict how many AI features you'll be able to find across the apps and devices you use most.

Vocabulary

See the full course vocabulary

AI feature: A part of an app or device that uses AI, like photo search, autocomplete, or recommendations.
Human decision: A choice a person made when building an AI, such as which examples to use or what counts as a correct answer.
Everyday AI: AI you meet in normal daily life, often without noticing it.

AI is hiding in plain sight

Many apps use AI quietly. Your keyboard suggests the next word, your camera brightens faces, your map predicts arrival times, and your video app lines up the next clip. These are all AI features working in the background.

Being able to spot them is a real skill. Once you look for 'learned patterns', you start noticing AI everywhere.

Keyboard word suggestions
Photo search by what's in the picture ('dog', 'beach')
Estimated arrival time in a maps app

Every AI feature has an input and an output

You can understand any AI feature by naming its input and output. Autocomplete: input is the letters you've typed, output is a suggested next word. Photo search: input is your photos and a search word, output is the matching pictures.

Mapping input and output turns a mysterious feature into something you can explain and question.

People made decisions behind every AI

An AI feature did not appear on its own. People decided what it should do, which examples to train it on, and what a 'correct' answer looks like. Those human decisions shape how the AI behaves — and whether it works well for everyone.

When an AI does something surprising or unfair, it often traces back to a human decision, like which examples were collected.

Deciding that autocorrect should suggest common words
Choosing which photos to show a photo-search model while it learns

Worked example

Investigating autocomplete

Pick the feature: the keyboard suggesting your next word.
Input: the letters and words you've typed so far.
Output: one or more suggested next words.
Learned or ruled? It learned common word patterns from huge amounts of text — that's AI.
Human decision: people chose the text it learned from, which is why it suggests some words and not others.

Takeaway: Any AI feature becomes clear once you name its input, its output, and a human decision behind it.

Anatomy of an everyday AI feature

A labeled diagram of an AI feature. On the left: 'Input' (what you give it). In the middle: 'Learned pattern' (trained from examples chosen by people). On the right: 'Output' (what it gives back). A note points to the middle: 'Human decisions: which examples, what counts as correct.'

Input and output are visible; the learned pattern and human decisions are hidden inside.

Activity

Check your understanding

Check that you can investigate and explain everyday AI.

Question 1. What are the two things you should name to understand any AI feature?
Question 2. Decide if the statement is true or false.
The way an AI behaves can be traced back to human decisions, like which examples it learned from.
Question 3. Which is an example of an everyday AI feature?
On the bus, Leo's phone suggests the word 'library' as he types 'lib', and his headphones pause when he takes them off.
Question 4. What is the most honest classification here?
A friend says a weather app 'is definitely AI.' But you don't know whether it uses physics equations, machine learning from past weather, or both.

0 of 4 answered

Try it yourself

AI feature field guide

Make a short 'field guide' of five AI features you found, like a nature guide but for AI.

Choose five AI features from your investigation.
For each, write a name, its input, its output, and one human decision behind it.
Rank them from 'most helpful' to 'least helpful' for you, with a reason.

Success looks like

Five AI features, each with input, output, and a human decision.
A ranking with at least one clear reason.

Reflection

Saved on this device as you type

Please don't include private information. Your answers are saved only in this browser and are never sent anywhere.

Which AI feature do you rely on most without thinking about it?

Now that you can spot the human decisions behind AI, what questions do you want to ask about the AI you use?

Recap

Everyday AI is everywhere; you can understand any feature by naming its input, output, and the human decisions behind it.

AI features often work quietly in the background of apps you use.
Every AI feature has an input and an output you can name.
Human decisions — especially which examples were used — shape how AI behaves.

Grades 7–8 extension

Design a better everyday AI

Pick one everyday AI feature you found and imagine you could improve it. What extra examples would help it work better for more people?

Write a short proposal: the feature, one problem it has, and the examples or human decisions that could fix it.

Finish this lesson

Attempt the knowledge check — not done
Save a reflection (recommended) — not done

Attempt the knowledge check to finish this lesson.