Why Lumina Learning Exists

We have reached a strange moment in education: the quality of a student’s work may tell us almost nothing about the quality of the student’s thinking.

Students are learning that they can use AI to generate the appearance of understanding with ease. Teachers, meanwhile, are using these same tools to create lessons and give feedback.

Education in 2026 has begun to resemble a bizarre loop in which AI generates an assignment, AI completes the assignment, AI evaluates the assignment…and everyone politely continues as though the exchange represents learning.

It is a peculiar situation.

I do not think the answer is to try to ban AI, nor do I think we can (or should) keep children apart from it indefinitely. AI is becoming part of the intellectual environment in which they will grow up.

The key, I believe, is to teach kids to work with AI consciously and with intention.

Its arrival forces us to reconsider what we mean when we say that a student knows something.

For a long time, schools have relied heavily on written products as evidence of understanding: the essay, the worksheet, the reading response, the research report. That made a certain amount of sense when producing those things required the student to do most of the intellectual work involved.

That relationship is changing.

A polished paragraph can now tell us remarkably little about what is happening inside a student’s mind.

Conversation, however, can tell us much more.

Ask a child why he thinks a character made a particular choice. Ask him whether he agrees with a character’s choices. Ask him to explain an idea in his own words, respond to another student’s argument, or sit with a question for a moment before answering.

There is nowhere to copy and paste from.

The purpose is not to catch students doing something wrong. Quite the opposite. Conversation gives us access to something more interesting than the finished product: the process of thought itself.

This is one reason Lumina Learning is built around dialogue.

Students need places where ideas can remain unfinished for a while. They need opportunities to speak, to listen, and to explore ideas. That is the way they can discover what they actually think. They need to become comfortable saying, “I’m not sure yet.”

Reading and writing remain deeply important to me. I have spent much of my life teaching students both skills, and I do not believe their value will disappear.

But they will change. It is happening already. We are watching humanity’s needs and capabilities evolve in real time.

And for the youngest humans, those who will become known as “AI Natives,” it may become more important than ever to know the difference between having an answer and understanding it.

Perhaps we need to become clearer about why we ask students to read and write in the first place.

We do not read simply to extract information from a page. We read because sustained attention changes what we are capable of noticing. We encounter minds unlike our own. We hold complicated ideas long enough for them to work on us.

We do not write only because the world requires written products. Writing can teach precision. It can reveal the gaps in our reasoning. It can slow thought down enough for us to see it.

The value lies partly in what the activity does to the learner.

That distinction becomes important when a machine can perform the visible part of the task almost instantly.

At Lumina, I am interested in what happens when education begins from that question: What kinds of experiences help a young person become more attentive, curious, articulate, and aware of his or her own thinking?

That leads naturally beyond traditional subject boundaries.

Literature matters. So do history, science, philosophy, mythology, rhetoric, art, and the questions that do not fit neatly into any one of them. Children are often far more interested in connections between ideas than school structures allow them to be.

Mindfulness belongs here too, though perhaps not in the way the word is sometimes used in schools.

I am interested in teaching children to notice.

To notice when their attention has wandered. To notice the difference between understanding something and recognizing familiar words. To notice what happens internally when someone disagrees with them. To pause before answering. To become aware of how they are learning, not merely whether they have completed the work.

AI makes this kind of awareness more necessary, not less.

A child growing up with increasingly capable machines will need to make choices that previous generations rarely had to make. When should I ask for help? When should I struggle with something myself? Is this idea mine? Do I understand what I am presenting? Has this tool expanded my thinking, or replaced it?

Those are not questions that can be answered by a plagiarism detector.

They are questions of judgment.

Lumina Learning is still small. I want it to remain small enough, at least for now, that students can actually be known. A seminar should feel like a place where a child's ideas matter, where questions can change direction, and where the teacher can notice when something has clicked or when a polished answer is hiding confusion.

I am especially drawn to students who are curious, capable, and hungry for more than simply moving through the next unit. Not necessarily children who carry a particular label, but those who enjoy ideas and are ready to be challenged by them.

I do not know exactly what education will look like twenty years from now. I am skeptical of anyone who claims that they do.

But I think we can already see enough to ask better questions about what is worth preserving and what needs to change.

The ability to generate an answer is becoming cheap.

Attention is not.

Judgment is not.

Curiosity is not.

The ability to sit across from another person, listen carefully, form a thought, and say something you genuinely mean is not.

Those are capacities worth educating for.

That is why Lumina exists.