AI Literacy for Year 5-6

The Big AI Project

Supporting Organisation:

Reminder - AI rules - for you

There are some important things that you need to know before you use AI.

AI tools often have age restrictions - many are only for age 13+ or age 18+. Your teacher will help you with what is appropriate for your age.

AI should not be seen as a person - if you want advice speak to someone you trust.

Your school will help you to understand their rules on AI use - you must follow these.

​

AI - Ethics & Bias

Lesson 3

Supporting Organisation:

Last lesson we were in the Lyfta world!

​

In lessons 1 and 2 in Lyfta:

​

• We travelled to San Francisco to meet an AI engineer!
• We learned what AI is and how it is used in everyday life.
• We learned how AI has been developed over time.

Recap:

To learn new vocabulary - like ‘stereotypes’ and ‘bias’�

​

To recognise examples of unfair or biased information

​

​

To understand that AI is created by people and can reflect human bias�

​

01

02

03

Learning Objectives:

Discussion:

​

In pairs think about these questions then share with the class.

​

What do you think AI is?

Do you think the information from AI is always right? Why or why not?

Use the sentence starters to answer the questions:

Partner A: I think AI is…

Partner B: I agree/disagree with you because…

A/B: In my opinion, AI does/doesn’t get things right because…

​

​

Star Words

Assumption

Bias

Fairness

Stereotype

Representation

What do they mean?

Bias

Fairness

Stereotype

Representation

Assumption

A guess based on what you already know or have seen.�

An unfair preference or belief.�

Treating people equally and kindly.�

A common idea or belief about a group of people based only on looks or labels.

�

Making sure everyone is included and seen.

Look at these AI-generated images I got when I asked for a picture of a dancer and a doctor.

1. What do you notice about these pictures?�
2. What kind of person have they chosen?

​

1. Does this tell us about all doctors and all dancers?

​

Click to reveal statements:

Are these two statements true or false?

“All doctors are men.”�“All dancers are women.”

Activity 1: Who is missing?

Story time: The Robot That ‘Forgot’ Me

Lina loved drawing, especially drawing scientists.

​

One day, she asked her AI drawing tool:

​

"Can you show me a picture of a scientist?"

�The AI showed her five pictures, all of them were men in white coats.

​

None looked like Lina. None were women.

​

None had brown skin like hers.

​

Lina frowned.

​

“What about me?” she thought.

​

“Can’t girls be scientists too?”

​

(Read aloud)

Story time: The Robot That ‘Forgot’ Me

Later, her teacher explained:

​

"AI tools learn from pictures on the internet. If most pictures show one type of person, the AI presents that as the only way."

​

Lina raised her hand.

​

“Then we need to show everyone

that anyone can be a scientist.

​

Aha!

​

I’ll draw a whole team of scientists:

boys and girls,

all kinds of people who look unique.”

​

Story time: The Robot That ‘Forgot’ Me

The next day, Lina’s drawing was added to the school’s “Aspirations Wall”. Now, her whole scientist team was part of the picture, too.

​

Lina said: “I think the people who design AI should include everyone.”

Aspirations Wall

My Future

Use the sentence stem to answer the questions:

Partner A: I think AI… because…

Partner B: I think Lina felt… because…

A/B: I would like to build on that… I agree/disagree…

​

​

Story time: The Robot That ‘Forgot’ Me

Who did AI ‘forget’?

How did Lina feel?

What did she do to fix it?

Why is AI biased?

AI tools are trained on pictures, videos and words from the internet.�

If most of the images show only white men as doctors or only women as nurses, that’s what the AI’s data is showing as “normal”.�

This leads to stereotypes in its answers.

​

​

​

For example: Facial recognition tools might not work well for people with darker skin if they were mostly trained on light-skinned faces.

​

AI learns from data made by humans. If the data is unfair or only shows one type of person, AI becomes unfair or biased too.

Instructions:

1. One student is going to build a team using a secret rule given to them on a slip of paper (for example, only picking team members with brown eyes, or who are wearing a hair clip). They don’t tell anyone what the rule is until the end of the game. �
2. Watch carefully: Who do they pick?�
3. Your job: Figure out the rule they’re using!�
4. When you think you know the rule, raise your hand.

Activity 2: Fair or unfair. Can you guess the rule?

Purpose: To explore how AI can make choices using hidden rules that may seem unfair.

After the Round:

​

Was the rule fair or unfair?�

Was it easy or hard to guess the rule?�

Do you think we can always see the rules that have been used to generate an answer in AI?

​

​

Reflection

Let’s remind ourselves - what have we explored today?

AI is powerful, but it only knows what we teach it.

We need to check if what AI has shown as an answer could be biased.

Bias and stereotypes can be hurtful for people.

👍= Confident

👎= Not sure yet

👋 = Curious