How to Train Your Algorithm Lesson Plan

Artificial intelligence
Internet & Mobile
Privacy

Level: Grades 9-12

Author: Matthew Johnson, Director of Education, MediaSmarts

This lesson is part of USE, UNDERSTAND & ENGAGE: A Digital Media Literacy Framework for Canadian Schools.

Overview

In this lesson, students will learn about algorithms and how they function, particularly recommendation algorithms utilized by popular apps like YouTube, TikTok, Instagram, and Netflix. Students explore the role of optimization goals such as watch time, engagement, and daily active use in shaping the content algorithms prioritize. Through activities like “red teaming,” students will critically analyze the potential downsides and biases of these optimization goals. Students will also discover how to train algorithms by providing both explicit inputs through actions like liking and sharing, and understand the implications of implicit inputs gathered from their online activity. Finally, students design their own algorithm for an app of their choice, identifying which goals it should be optimized for and how they should be weighted as well as what inputs it should use.

Learning Outcomes

Big ideas/key concepts: Students will understand that…

Media are constructions:

  • Algorithms are designed and optimized for a particular purpose

Media have social and political implications:

  • Recommendation algorithms can have negative impacts on people and society due to their optimization goals

Media have commercial implications:

  • Algorithms’ design and optimization serve commercial purposes such as maximizing watch time and engagement

Digital media have unexpected audiences:

  • Many apps, websites and other digital tools collect information while you are using them to use as inputs for their algorithms

Digital media experiences are shaped by the tools we use:

  • How algorithms are designed and optimized influence how we use an app or tool
  • It is possible to consciously “train” algorithms to influence the content they deliver

Key questions:

  • How do algorithms work?
  • What are some of the implications of how they work?
  • How can we use them more consciously and mindfully?

Essential knowledge: Students will know…

Reading media: Operation of recommendation algorithms, including optimization goals and explicit and implicit inputs

Consumer awareness: Commercial goals of algorithms’ optimization goals

Privacy and security: Collection of personal data for use as implicit inputs

Performance tasks: Students will be able to…

Access: Find and use privacy settings

Use: Refresh and customize an app’s recommendation algorithm

Understand: Analyze the implications and potential impacts of algorithms, their optimization goals and their inputs

Engage: Identify potential harms of recommendation algorithms, take steps to use algorithms more mindfully, and design an algorithm that would mitigate those harms

Student-friendly outcomes:

We will learn how algorithms and recommendation systems work, what their optimization goals are, and key terms like “algorithm,” “optimization goal,” and “implicit input.”

 We will think about the social, political, and commercial impacts of algorithms, how they collect and use our data, and how we can use them more mindfully.

We will find and use privacy settings, customize app recommendations, analyze the effects of algorithms, and design ways to reduce potential harms caused by recommendation systems.

This lesson and all associated documents (handouts, overheads, backgrounders) is available in an easy-print, pdf kit version.

Lesson Kit