Teaching a Machine
By identifying and applying the five Big Ideas of Artificial Intelligence to train and test a machine-learning tool to classify drawings the students will be able to give examples of a bias that can arise in machine learning and understand how biases may arise.
Overview
Students learn how machines are taught and how bias can enter machine learning. The lesson opens with the Intelligent Piece of Paper, a tic-tac-toe activity showing that a paper of instructions works like a computer program. Students complete the AI and Me course from Readyai.org on the five big ideas of AI, brainstorm what a robot is, compare a car factory line with Waymo, and work with a Cozmo robot's face and object recognition. In Google's Teachable Machine they train a cat-dog classifier on a biased dataset and retrain it. They then train an A.I. Bot to recognize fish features and expressions, and record a Flipgrid summary video. Exploration takes 30-40 minutes, explanation 50 minutes plus an optional 15-20, and elaboration 25-40 minutes.
Lesson Objectives
- By identifying and applying the five Big Ideas of Artificial Intelligence to train and test a machine-learning tool to classify drawings the students will be able to give examples of a bias that can arise in machine learning and understand how biases may arise.
Standards
NGSS · 2
- MS-ETS1-1Define the criteria and constraints of a design problem with sufficient precision to ensure a successful solution, taking into account relevant scientific principles and potential impacts on people and the natural environment that may limit possible solutions.
- MS-ETS1-2Evaluate competing design solutions using a systematic process to determine how well they meet the criteria and constraints of the problem