Computing · Year 11

Active learning ideas

Artificial Intelligence and Machine Learning

Active learning works for AI and ML because these topics demand concrete, hands-on experience to move beyond abstract concepts. Students need to see how algorithms process data in real time, debate trade-offs, and test ethical limits to grasp both technical and social dimensions.

• National Curriculum Attainment Targets
• GCSE: Computing - Ethical, Legal and Cultural Impacts

• 30–50 min
• Pairs → Whole Class
• 4 activities

Activity 01

• Socratic Seminar
• 40 min · Pairs

Debate Pairs: AI Benefits vs Risks

Pair students to prepare arguments for or against statements like 'AI will create more jobs than it destroys.' Each pair presents for 3 minutes, followed by whole-class rebuttals and vote. Conclude with a shared mind map of key points.

Analyze the potential benefits and risks of widespread AI adoption in society.

Facilitation Tip

During Debate Pairs, provide clear timekeepers and a visible scoring rubric so students focus on evidence rather than volume.

What to look for

Pose the question: 'Imagine a city council is considering using AI to predict crime hotspots. What are the potential benefits for public safety, and what are the ethical risks regarding privacy and potential bias in the data?' Facilitate a debate, asking students to support their points with evidence from the lesson.

• Analyze
• Evaluate
• Create
• Social Awareness
• Relationship Skills

Activity 02

• Socratic Seminar
• 35 min · Small Groups

Small Groups: ML Training Simulation

Provide datasets and free tools like Teachable Machine. Groups train models to classify objects via webcam, then test accuracy and tweak parameters. Discuss what 'learning' means based on error rates.

Explain how machine learning algorithms 'learn' from data.

Facilitation Tip

In ML Training Simulation, circulate with printed dataset samples so students can compare how small changes affect model performance.

What to look for

Provide students with a short scenario, e.g., 'An AI system is trained to identify loan applicants likely to default. The training data shows a historical pattern of fewer loans approved for a specific demographic group.' Ask students to identify: 1. What is the potential problem with this AI system? 2. What is this problem called? 3. How might this bias have occurred?

• Analyze
• Evaluate
• Create
• Social Awareness
• Relationship Skills

Activity 03

• Socratic Seminar
• 50 min · Whole Class

Whole Class: Ethical Dilemma Role-Play

Assign roles like developer, citizen, regulator in scenarios such as biased hiring AI. Students act out decisions, then debrief in circle to vote on outcomes and justify choices.

Predict future applications of AI that could transform various industries.

Facilitation Tip

For Ethical Dilemma Role-Play, assign roles in advance and give each student a one-page scenario to anchor their arguments.

What to look for

On a slip of paper, ask students to write: 1. One way machine learning algorithms 'learn' from data. 2. One specific future application of AI that excites them and why.

• Analyze
• Evaluate
• Create
• Social Awareness
• Relationship Skills

Activity 04

• Socratic Seminar
• 30 min · Individual

Individual: Future AI Predictions

Students list three industry transformations by AI in 2030, with pros, cons, and ethics. Share in gallery walk for peer feedback and class synthesis.

Analyze the potential benefits and risks of widespread AI adoption in society.

Facilitation Tip

During Future AI Predictions, ask students to cite current trends and cite at least one source in their notes.

What to look for

Pose the question: 'Imagine a city council is considering using AI to predict crime hotspots. What are the potential benefits for public safety, and what are the ethical risks regarding privacy and potential bias in the data?' Facilitate a debate, asking students to support their points with evidence from the lesson.

• Analyze
• Evaluate
• Create
• Social Awareness
• Relationship Skills

A few notes on teaching this unit

Experienced teachers approach AI/ML by balancing concrete demonstrations with critical discussion. They avoid letting students rely on black-box explanations by showing raw data and simple models first. Research shows that students grasp bias better when they manipulate skewed datasets themselves rather than reading about it.

Successful learning looks like students explaining how ML models generalise from training data, critiquing ethical trade-offs with evidence, and applying technical insights to real cases. They should articulate limits of AI systems and justify their reasoning with concrete examples.

---

Watch Out for These Misconceptions

• During Debate Pairs, watch for students attributing human-like reasoning to AI.

  Pause the debate to ask pairs to point to the exact pattern in their training data that leads to a decision, forcing them to ground claims in evidence.

• During Ethical Dilemma Role-Play, watch for students assuming ML outputs are always neutral.

  After role-play, display the biased training data snippet used in the simulation and ask students to revise the scenario to address this flaw.

• During ML Training Simulation, watch for students believing more data always improves fairness.

  Have students try training with 20 items, then 10 identical duplicates, to observe how repetition can amplify bias rather than reduce it.

---

Methods used in this brief

• Socratic Seminar