The other day, I asked my students to write a song about what they learned that day. In 5 minutes. And yes, they could use AI.
A few minutes later, I had a playlist of AI-generated songs—different styles and lyrics, but quite polished (see here and here, for example). We had fun. And now I have some fun music to play before class!
But did they actually learn anything?
Design and creation are acts of learning. When we make something we’re actively constructing knowledge. It’s about the thinking, the problem-solving, the meaning-making. But this activity? It felt different. Students typed a prompt, hit enter, and out popped a song (well, most of them used an LLM to create the lyrics and then took those to a music program like Suno, but still). Where was the learning?
I could have asked them to spend more time refining, and it would have been a slightly different experience. Maybe they would have to think more about what the lyrics were about and try to match a musical style to go with it. But it is still taking a shortcut. And If AI takes over the hard parts of creating, how does that impact the learning?
We learn through the process of designing or creating something because making a meaningful product requires making meaning ourselves. AI disrupts that. It churns out something polished in seconds, sidestepping the messy, frustrating, but ultimately rewarding act of making. But maybe the issue isn’t just that AI makes creation easy—it’s that it makes shallow learning easier to mask.
AI-generated work can look impressive. But is that representative of a deeper understanding? When students create with AI, they might not need to wrestle with ideas or build connections. They can produce something without actually knowing much at all. So what did they learn?
When Ideas Are Mine but the Words Are Not…Or Are They?
As I think about this in my teaching, I realize I’m facing it in my own work too. Right now. As I write this post.
Writing is creating and learning. When I write, I’m not just producing a blog post; I’m engaging in the same process my students go through when they are creating products of their learning. I’m testing connections, refining perspectives, and making meaning through creation.
This very blog post is a kind of AI-assisted creation. I started with a prompt—just a rough idea of what I wanted to explore about the song making activity. Writing the prompt helped me clarify what ideas I wanted to bring in. This led to the focus on learning through design and creation. But I didn’t want to write a prompt with a long explanation of my own perspective on design and learning, and I didn’t want AI to simply reflect the common conversation around these topics (I’m sure “design thinking” would make an appearance!). So I pulled out some of my old slides—materials rooted in past research and thought—to communicate my paradigm more clearly and integrate those ideas with my reflections on my students’ AI-generated songs.
I must admit, I was amazed at how the ideas from my slides were integrated with my recent reflections!
Now, with each iteration of this post, my ideas get a little bit clearer. The ideas feel like mine; they are built on not just a recent experience but my past work in learning and design.
But. It still doesn’t feel like me, like my voice. I asked the AI to match my blog’s style, but it ‘s struggling with that. Even if it did match my style, would it be my voice? Now, when I read through it, the writing is clear and structured (if a bit cheesy)—so clear that it feels pointless to make changes. I know that what’s here is very typical writing, not me.
So what did I gain? Efficiency. Clarity. A sharpened articulation of my thoughts. But what did I lose? The slow, sometimes painful, process of wrestling words into coherence. Maybe a deeper reflection that only comes from struggling through an idea from scratch. Maybe even a little ownership over the final piece.
What Are We Really Learning?
So where does that leave us? Maybe we need to rethink what it means to create, to design, to learn. What happens when AI doesn’t just help us create—but helps us skip the thinking altogether? And what does that mean for the future?