Use of the AI for Active Learning

Like a lecture passing from the instructor’s notes to the student’s notes without passing through the mind of either.

It's an old saying - My first encounter was a YouTube lecture posted by Dr. Eric Mazur . It had a profound impact on my teaching.

https://youtu.be/wont2v_LZ1E

From that point forward, we spent at least half the class in active groups. To be sure, it wasn't perfect, but without question it was better than the passive lecture. On the best days, students lost track of time and I had to escort them out of the room. They were awake! They were excited to learn, they asked many questions, and they showed a tremendous respect and compassion to each other as they wrestled with challenges. They were even excited to work the complex math shown as this article's opening image.

Active learning with the AI as the More Knowledgeable Other

Don't be too quick to dismiss the AI in education. Personally, I view it as Vygotsky’s?More Knowledgeable Other (MKO). If the learner is honest and not satisfied with just finishing the assignment, the AI has much to offer as a partner in an active learning environment.

While I am no longer teaching in a traditional classroom, my passion has not diminished. Today, my lessons are written for an international audience on the DigiKey TechForum with teasers here on LinkedIn.

Teaching with AI is something that I've been thinking about for a long time. Today is my first attempt to organize my thoughts and share them with the world.

Let's start with a recognition that the AI is generally aware of many highly technical topics far more than my finite mind will ever know. However, our objective is not to be like the AI. Instead, we need to focus on specific topics and applications. If we can focus the AI into that small corner of the world, it is a capable learning tool.

A Focused AI Learning Example

As an example, let's focus the AI on a highly specialized topic. If you will allow some DigiKey promotion I will point you to this article: https://forum.digikey.com/t/guide-to-selecting-and-controlling-a-mosfet-for-3-3-vdc-logic-applications/42606. If I've done my job correctly, this article is written for Junior EE students, with lessons extending to Seniors exploring a related capstone project.

We begin with a simple prompt to focus the AI on a specific topic and to place it into "MKO mode." Note that GPT 4o is used in this example:

Read this article. Test me by asking related questions. One question at a time: https://forum.digikey.com/t/guide-to-selecting-and-controlling-a-mosfet-for-3-3-vdc-logic-applications/42606

The AI will dutifully ask questions related to the article. It's generally smart enough to understand your written responses and point out when you are missing a critical point.

It's not perfect, but neither is a real MKO. Over time, the AI tends to drift back to a global focus, requiring a periodic memory flush and reload of the page. It's been my experience that the AI is getting smarter. I wouldn't have written this article for GPT 3.5. Personally, I can't wait for the next version.

Expanding Examples

We could continue our learning examples by asking the AI to summarize the article, write T/F questions, or make flash cards. It is certainly capable of this and more.

However, it's not the only AI. Your students may be better served using a purpose-built AI. An example is https://notebooklm.google.com/. This tool is advertised as a "Your Personalized AI Research Assistant." It is purpose-built to present the key points. It can even convert a page into a podcast allowing you to listen to the host and guest explore the content.

Parting Thoughts

The AI makes a reasonable stand-in MKO for the learner willing to take the time to explore the material. It a worthy tool, but it lacks that respect, compassion, and enthusiasm associated with a well-run active learning session. It can't (yet) read the tone of a classroom like a skilled teacher.

Please share your thoughts on how to integrate the AI into your classroom.

Sincerely,

Aaron