Hey, y’all!
This week from Austin Scholar...
Why learn to write in the age of AI?
Scholar’s Sources: What I’ve been thinking about…
World School is officially launched! This past week has been the craziest, most exhausting, yet fulfilling of my life and I am so beyond excited to see how these high school students transform education for millions of students in Africa. They just started vlogging, journaling, and reflecting on their experience and learnings. I will share their journey over the coming weeks and keep y’all up to date on everything these amazing students are working towards, so stay tuned for that.
Why learn to write in the age of AI?
Last week, I talked about how you can use generative AI to learn how to write. But I think a lot of teenagers today aren’t sold on the value of writing at all. It doesn’t seem obvious what the point is anymore: if a generative AI tool can produce an essay or story that’s better than a human in seconds, what’s the point in bothering to learn how to write in the first place?
That’s a completely valid question, particularly given the role writing currently plays in our educational system. For most teenagers, writing feels like a task to check off of a to-do list, mostly for the teacher and not really for themselves. Schools often treat writing as the final step of learning: first you understand it, then you write it up. But in practice, writing is often the very thing that creates understanding.
When you write, you have to select the important idea, organize it into a structure, and express it with enough detail and clarity that someone else can follow your logic and engage with your arguments. This process turns vague impressions and understandings into clear, structured ideas. That process of engagement is what actually forges original insights and creative thoughts. Often you don’t actually understand an idea until you’ve written about it, and often you make arguments by the end of your piece that you didn’t even see when you started.
Like in my newsletter, for example, I take my memories, experiences, and half-baked takeaways, put them in a Google Doc, and organize them until it’s interesting. I often don’t know the exact point I’m going to make or how all of the pieces and insights fit together until I’ve gone through the full writing process. My experience of this isn’t unique. In fact, there’s a whole body of “writing-to-learn” research, including a meta-analysis of 48 studies showing that writing-to-learn programs reliably improve academic achievement. And in one widely cited study on writing and critical thinking, college biology students who regularly wrote to explain and argue about course content significantly improved their critical thinking performance, while students in the non-writing group did not.
So if a teenager asked me why they should bother to learn to write in the age of AI, I would tell them this: writing is thinking. Writing still matters because it’s one of the best tools in the world for clarifying ideas, deepening understanding, and generating original thought. AI can produce text. It cannot create connections in your mind between disparate topics, use your personal experience to come up with examples or reasoning, or figure out for you how new knowledge fits into your existing knowledge map.
Why can writing do those things, you might be wondering?
Well, the act of writing as a learning tool is backed by learning science:
Retrieval practice
When you write about a topic from memory, you’re forced to pull information out of your head rather than simply recognize it on a page.This kind of active retrieval strengthens memory significantly more than simply rereading or highlighting. Every sentence composed without looking at your notes is, essentially, a small form of self-testing. That effortful retrieval makes the memory more accessible in the future.
Elaboration
Writing an explanation forces you to answer the questions “why?” and “how does this connect?” You can’t write a coherent paragraph about a concept without relating it to examples, causes, consequences, or other parts of your existing knowledge map. This embeds new material in a web of existing knowledge rather than leaving it as an isolated fact, giving that new material the context that comes through different connections and associations.
Organization
To write an argument, you must decide what comes first, what depends on what, which points are central, and which are supporting. That act of organizing that knowledge builds a schema. (I talked about schemas a few weeks ago, but if you missed it, a schema is a mental model of how the pieces of a domain fit together.) Schemas make your current knowledge more coherent, and they act as scaffolding that makes future learning in the same domain faster, because new information has a structure to attach to. A learner with a well-organized mental model of a topic can absorb a new fact in seconds, while a learner without one has to hold it as a disconnected item.
Metacognition
If you try to explain a half-understood concept in simple, understandable sentences, it will become obvious very quickly whether you actually know the content. Psychologists have shown that most of us walk around with an “illusion of explanatory depth“, which means that we feel like we understand things far better than we actually do, until we’re asked to explain them. Writing gives you an accurate map of what you actually know, which is necessary if you want to fix what you don’t.
Desirable difficulty
Desirable difficulty, which describes conditions that slow learning down in the moment but make it stick far better in the long run, is the underlying idea behind all of these learning science concepts. Writing is slower and more effortful than reviewing, which is exactly why it works better. The effort is the learning; writing is thinking.
Writing also does more than clarify your existing knowledge and thoughts; it actually improves them. Writing is linked to gains in memory, verbal skill, critical thinking, and creativity. One study of a ten-week creative writing program found that students’ critical thinking scores rose 25% while a matched control group barely moved, and other creative writing interventions have produced measurable gains in cognitive flexibility. This suggests that writing helps people explore alternative solutions, sit in ambiguity, and arrive at more original thoughts.
All of this is why I think of writing as a direct path to Depth of Knowledge levels three and four (strategic and extended thinking, respectively). The first two levels of Depth of Knowledge (recall and recognition) are increasingly easy to outsource, since AI can summarize, restate, and recombine information quickly. But strong writing requires analysis, justification, synthesis, and creation: the very skills that make up those higher levels of thought.
When I’m writing my newsletter, I’m synthesizing my personal experiences and the research I’ve done to create new ideas, examples, and methods to communicate innovation in education. I’m able to form new connections between concepts and themes as I’m writing, developing insights and (hopefully) coming up with cool ideas to share with y’all.
Putting all of that together: writing is not becoming obsolete in the age of AI. Its role in education just needs to transform.
Comments
Nothing yet. Say the first thing.
Sign in to join the conversation.