Long story short, it was a great success! I tested the lesson a grade 3/4/5 Montessori class at my school. They were approaching a natural transition where they were finishing up a unit, so they seemed like the perfect candidates to try this lesson. They were a little puzzled at first when they realized that there were no devices at their workstations, but they were so excited to talk the moment I introduced the topic. During the first part of the lesson with the slide deck, we had such a rich discussion about what they knew, what they were wondering, and looking at what AI can and cannot do at that point in time. They loved looking at all of the images I had created with AI and had a lot to say about what they noticed

Bias in AI – the AI knows “teacher”, but there isn’t a frame of reference for “computer teacher”
Even simple outputs need to be examined carefully

Every student had something to say about the visual examples, or could give me a thumbs up or a thumbs down letting me know if they thought it was ethical to use AI in different situations. There was not a single student who was disengaged with the conversations. I was amazed by the judgement calls they made, and their points on where the grey area might lie.

For the activity, I was initially worried that the game would be difficult to explain and get the students to play. At that point, I was running out of class time because the conversation was too important. I made the call to let them know that we wouldn’t have time to play the game a lot, but I would extend the learning to their next class with me. This play period would be a “head start”. Again, the students took like fish to water. They got into pairs quickly, picked up all of their materials, and started playing. I felt great pride that this lesson was going off without a hitch, aside from timing. The students asked some really interesting questions regarding what happens when all the beads are removed from a box, and what if the algorithm they see on the card is the mirror image of what they see on the board. By the end of the class, we created little “grab and go” packages of their games and reflective worksheets, so the next class could be a straight to work session. In that next class, all of the student pairings got back to play, and finished their reflective worksheets with really thoughtful responses. I was so pleased that the students not only engaged with the topic so intensely, but they retained the teachings and made sense of it for themselves within the context of the game and worksheet. I gave the students who finished early access to the online version of Hexapawn to use as a survival game and they had a blast trying to avoid losing to the AI. All of the students were super excited to learn that the game would be linked on my website and available as a “free choice” activity from now on.

I thought about all the years I had been teaching them all and how digital literacy and digital citizenship has always been at the forefront of my practice. We talk about privacy, being kind online, and knowing how things work before you use them all the time. My students take all of that very seriously and are able to have the difficult conversations with me when they are unsure of how to use a program, or they suspect their privacy has been compromised. I’d like to think that all of these talks in the past have given them the foundation to make good judgement calls on how to use new technology in responsible and kind ways, while still pushing their understanding.

Because I was so proud of how much my students were learning and how much they enjoyed engaging with the material, I gave the students a final follow-up lesson through code.org’s “AI for Oceans“, where they clean up the ocean by organizing fish from garbage using AI. The videos interspersed through the activity reinforced some of the learning we had already explored, while showing them how sophisticated AI can get once it gets the basics down. The class was abuzz discussing with each other why their AI system made certain mistakes. Some noted that they didn’t give their AI enough information to make good judgement calls, while others shared that their AI’s learned that in their ocean, all fish were red, so everything that is red belongs in the sea. Again, I was in awe of what the class were sharing and learning. Again, I made this activity available as a “free choice” if students wanted to experiment with changing how an AI might learn.

Through the introduction of one lesson, it snowballed into a 2 week learning period that not only piggybacked on our previous conversations, but had the students wanting to learn more. It made the hours of cutting, gluing, laminating, organizing, and shopping for materials worth it. It has been a highlight in my school year, and I intend to bring this lesson back next year for my new crop of intermediate learners when we do our larger digital literacy unit. I am looking forward to the final meeting with the rest of the committee to share out our success, obstacles, and next steps.