Teaching Students How AI Works: Moving From Consumers to Creators

Summary for Educators | School Leadership 2.0

Based on:

Source: Ross Hartley, Edutopia, October 29, 2026
Original Article: Teaching Middle and High School Students How AI Works

THE BIG IDEA

Students shouldn't just learn how to use artificial intelligence. They should understand how it works.

As schools wrestle with AI policies, academic integrity, and the growing presence of ChatGPT and other intelligent technologies, computer science teacher Ross Hartley offers an important shift in perspective: AI literacy should extend beyond prompting tools and evaluating their responses.

Students need opportunities to explore how machines learn, recognize patterns, interpret information, and make decisions.

In his middle school classroom, Hartley uses a practical instructional framework called Watch, Read, Create. Students encounter real-world AI applications, build background knowledge through informational reading, and complete hands-on projects that demonstrate how machine learning operates.

The approach transforms AI from a mysterious technology into something students can investigate, question, and build.

For school leaders, the message is particularly relevant: Preparing students for an AI-driven world requires more than teaching them to use technology responsibly. It means helping them understand the technology itself. Edutopia

KEY TAKEAWAYS

1. A Simple Framework: Watch, Read, Create

Hartley's instructional model combines three complementary learning experiences:

• WATCH: Students view brief videos illustrating real-world applications of artificial intelligence.
• READ: Students examine informational texts, answer comprehension questions, and strengthen literacy skills.
• CREATE: Students complete hands-on activities exploring how AI models are developed and trained.

This predictable sequence builds engagement, conceptual understanding, and practical skills.

It also demonstrates that AI instruction can reinforce traditional academic competencies rather than compete with them.

2. Teaching AI Through Environmental Problems

Students explore how artificial intelligence can help identify plastic pollution in oceans.

Using Code.org's AI for Oceans, they train a machine learning model to distinguish fish and other marine creatures from underwater trash.

As students classify images, they discover that machine learning depends heavily on the examples used during training.

Important lesson: An AI system's performance depends partly on the quality and representativeness of its training data.

This activity provides a natural introduction to discussions about inaccurate predictions, incomplete information, and algorithmic bias.

3. Designing AI-Powered Assistive Technology

In another project, students investigate smart glasses designed to help people with color blindness.

Using Scratch programming and video sensing, students create a model that recognizes colors through a camera.

They then develop a virtual pair of glasses capable of identifying and announcing those colors.

The experience combines computer science, creativity, and human-centered problem-solving.

Students begin to understand that AI can be more than a convenience or entertainment tool. It can also address meaningful challenges involving accessibility and independence.

4. Understanding Autonomous Vehicles

Self-driving cars provide another engaging introduction to machine learning.

Students examine how autonomous systems interpret their surroundings and make decisions.

Using robotics equipment, they construct miniature vehicles that respond to colored traffic signals.

Their programs produce three behaviors:

• Red light: Stop.
• Yellow light: Reduce speed.
• Green light: Proceed at full programmed speed.

Students must train the vision sensor before programming the corresponding actions.

The project reveals an important relationship between data, recognition, programming, and decision-making.

5. Creating a Personal AI Assistant

For an optional culminating project, students develop their own voice-responsive assistant using Scratch.

Their programs respond to requests involving the time, date, jokes, facts, and translations.

Students also program an appropriate response when the system cannot understand a request.

The activity helps students investigate the concepts behind familiar digital assistants such as Siri and Alexa.

Although a Scratch prototype does not necessarily replicate the sophisticated AI models that power commercial assistants, it provides an accessible introduction to how conversational interfaces operate.

6. Free Resources Make AI Instruction More Accessible

Teachers do not necessarily need expensive software to introduce students to AI concepts.

Hartley recommends resources from organizations including:

• Code.org and its AI learning activities.
• The Raspberry Pi Foundation.
• The Scratch Foundation.
• The Computer Science Teachers Association.
• The Hour of AI.

These resources offer potential starting points for teachers who want to introduce AI concepts without having to develop an entire curriculum independently.

WHY IT MATTERS FOR SCHOOL LEADERS

Much of the educational conversation about artificial intelligence focuses on what students should be permitted to do with AI.

Should they use ChatGPT for homework? How should teachers detect AI-generated assignments? What constitutes acceptable assistance?

Hartley's approach raises a different question:

Are schools teaching students enough about artificial intelligence to become informed users, critical evaluators, and capable creators?

Understanding machine learning can help students recognize why AI sometimes produces inaccurate results, why training data matters, and why human oversight remains essential.

These lessons also support interdisciplinary learning involving science, mathematics, technology, reading, and ethics.

AI literacy should not be limited to computer science electives. Its underlying concepts have relevance across the secondary curriculum.

LEADERSHIP ACTION STEPS

1. Expand the Definition of AI Literacy

Review district AI initiatives to determine whether they address how AI systems function, not simply how students should use them.

2. Introduce Watch, Read, Create

Encourage teachers to experiment with Hartley's instructional framework in science, technology, mathematics, and other appropriate subjects.

3. Start With Accessible Activities

Pilot a lesson using Code.org's AI for Oceans or introductory Scratch activities before investing in specialized robotics equipment.

4. Connect AI Learning to Real-World Problems

Encourage students to investigate applications involving environmental protection, accessibility, transportation, and healthcare.

5. Teach AI Limitations Alongside Capabilities

Ensure that students investigate inaccurate predictions, algorithmic bias, privacy, and the importance of human judgment.

LEADER REFLECTION

If we visited classrooms across our district tomorrow, would we find students simply using artificial intelligence—or would we find them learning to understand, question, and create it?

The distinction may determine whether schools prepare students to become passive consumers of emerging technologies or informed participants in shaping their future.

Adapted and summarized for School Leadership 2.0 from Ross Hartley's October 29, 2026, Edutopia article.

Original Article

------------------------------ 

Prepared with the assistance of OpenAI's AI software. (2026). ChatGPT (5.2) [Large language model]. https://chat.openai.com