Reflections from the Courageous Minority

At the first sub-group meeting of our third Courageous Minority meeting, I began with a question that has become increasingly difficult to avoid: What is worth teaching when artificial intelligence can already do much of what schools teach students to do?

For a long time, discussions about AI in education have treated it mainly as a tool. How can teachers use it to prepare lessons? How can students use it to improve their writing? How can schools prevent cheating? These questions are understandable, but they are becoming too small.

AI is no longer simply helping people perform familiar tasks more efficiently. It can find information, explain complex ideas, generate arguments, write programs, analyze evidence, and guide users through unfamiliar problems. It is becoming capable of doing many of the things for which schools have traditionally rewarded students.

This does not mean AI can do everything. Nor does it mean students no longer need to learn. It means we must reconsider what learning is for.

From Knowing Nothing to Knowing Something

People often say that students need foundational knowledge before they can use AI or undertake serious intellectual work. There is truth in this claim, but it can also conceal an important logical problem.

Before anyone possesses foundational knowledge, that person begins without it. Every learner encounters new domains from a position of partial or almost complete ignorance. The real educational question is therefore not whether foundations matter, but how a person moves from knowing nothing to knowing something.

Must that movement always begin with a teacher transmitting a predetermined body of knowledge? Or can a student begin with curiosity, a meaningful problem, access to other people, and AI as a partner in exploration?

One participant described his 12-year-old son building a binary typewriter in Minecraft. The boy had not first completed a formal course in computer science. He became interested in something, searched for resources, watched videos, used ChatGPT, experimented, failed, and continued. Through the process, he encountered ideas that might ordinarily appear in a university computer science class.

The point is not that children no longer need teachers. Nor is it that every child will independently discover advanced mathematics or computer science. The point is that the traditional sequence—first teach the foundations, then permit meaningful work—is no longer the only possible sequence.

Sometimes meaningful work creates the need for foundations. A problem generates questions. Questions lead to knowledge. Knowledge becomes useful because the learner has a reason to acquire it.

Students Must Own the Learning

This idea also appeared in a short program with students in Kansas and Missouri. Students were invited to identify problems that mattered to them and use AI to investigate possible responses.

Some examined neighborhood safety and asked how communities might create cleaner and more welcoming public spaces. One student developed a website with maps, parks, community events, and possible actions. Another student, anxious about public speaking, researched the relationship between anxiety and excitement and proposed a speaking group designed specifically for students who feared public presentations.

The projects were not perfect. They were completed in a short period of time, and the students were still learning how to investigate problems and develop solutions. But the organizers heard something important from them: this was the first time they felt adults were genuinely listening.

Schools often tell students to take ownership while adults continue to determine the topic, process, product, criteria, and timetable. Students are given choices, but only among options already selected for them. They are told to exercise agency inside a structure designed to prevent them from exercising very much of it.

Authentic ownership begins when students are treated as human beings with interests, concerns, experiences, and lives of their own.

Freedom Is Not Easy

However, giving students freedom does not mean they immediately know what to do with it.

Personalization is a process. Problem finding is a process. Collaboration is a process. Human interdependence is also a process. Students need opportunities to practice all of them.

Several participants described students becoming frozen when asked to choose what they wanted to learn. Some responded with grand ambitions: “I want to change the world.” Others said they had no idea where to begin. Even students who had voluntarily entered self-directed learning environments sometimes felt disoriented when teachers stopped telling them exactly what to do.

One school found that its highest-achieving students struggled most when asked to select their own science topics and establish their own criteria for success. They knew how to earn an A. They knew how to interpret a teacher’s expectations. But without grades and predetermined instructions, they did not know how to judge whether they were succeeding.

Students who had performed less successfully in conventional schooling were sometimes more willing to experiment. They had less to lose because the old system had rewarded them less generously.

This reveals a serious problem. Many students have learned to treat school as work assigned by adults rather than as a place for learning. Their task is to comply, perform, receive a grade, and move on. When that structure disappears, they may initially experience freedom not as liberation but as uncertainty.

Teachers therefore remain essential, but their role changes. They must learn how to coach students through uncertainty, help them identify manageable problems, ask productive questions, make choices, seek assistance, respond to feedback, and recognize progress.

Agency cannot simply be announced. It must be developed.

Schools Still Matter

AI does not make schools unnecessary. It may make some of their most neglected functions more important.

In one classroom, students were asked to redesign school for 2026. During their discussions, they repeatedly emphasized friendship, clubs, sports, belonging, and social interaction. What they valued most about school was not always the formal curriculum. It was the opportunity to be with other people.

Personalized learning should never mean learning alone. Human beings discover themselves through relationships with others. They recognize their strengths through comparison and collaboration. They learn to contribute, receive help, negotiate differences, and participate in communities.

Schools may be among the few organized institutions that bring nearly all children into sustained contact with people beyond their families. In an age when AI can deliver explanations, information, and personalized assistance, the human purposes of school become more—not less—important.

The question is whether schools will cultivate genuine relationships and interdependence or continue replacing them with programs, standards, and slogans.

Preparing for a World in Flux

Participants also returned repeatedly to uncertainty. We do not know what AI will be able to do next year, much less when today’s students become adults. We do not know which occupations will disappear, which new ones will emerge, or which forms of knowledge will remain economically valuable.

Schools cannot prepare students for one clearly defined future because no such future exists.

What they can do is help students become capable of responding to change. Students need to learn how to learn, how to change direction, how to use unfamiliar tools, how to recover from failure, and how to continue when their previous knowledge or plans are no longer sufficient.

One participant described a son who studied physics and engineering, entered a different occupation, taught himself about AI, left a successful position to join an AI startup, and then began considering what he would need to learn next as AI took over more of his current work.

That may become an ordinary life story. Education cannot guarantee students a stable path, but it can help them become people who are able to create another path when the old one disappears.

Contribution, Not Merely Competition

Self-directed learning, however, should not become an education in private ambition. Students should not simply be encouraged to follow their passions without considering the effects of their actions on others.

The question is not only, “What do you want to do?” It is also, “For whom will it matter?”

Schools have long taught students to compete. Grades rank them. Tests compare them. Universities select a few and reject many. Students learn to understand success as being better than someone else.

But human value is not fully realized through ranking. It is realized through contribution.

Students should be encouraged to ask:

What value am I creating?

Who might benefit from my work?

Whose knowledge or assistance do I need?

How could my strengths complement the strengths of others?

Am I changing the world for the better, or simply changing it for my own advantage?

People can be self-interested without being selfish. They can pursue meaningful lives while also improving the lives of others. This is the essence of human interdependence.

Begin With the Willing

A major challenge is that not every teacher, parent, or administrator is ready for this transformation. Some are anxious about AI. Some want to ban it. Others are comfortable with the existing curriculum, grading system, and division of authority.

Trying to persuade everyone before doing anything may ensure that nothing happens.

The Courageous Minority approach begins differently. Find the teachers who are willing. Find the students who want to try. Create a small, protected space—a class, club, camp, project, or school-within-a-school—where a different form of learning can begin.

One school leader in South Africa described introducing AI gradually during staff meetings. He demonstrated practical uses and invited teachers to experiment. Over time, some began using AI in lesson preparation.

Students used Gemini Storybook to write imaginative stories, formulate questions, strengthen reading comprehension, and create posters for a community blanket drive. These were not presented as revolutionary breakthroughs. They were small gains. But transformation often begins with small gains that reveal new possibilities.

The goal of a courageous minority is not to force everyone to change. It is to make change visible.

Building a Different System

During the meeting, we also demonstrated an AI-supported platform being developed around three principles: personalization, problem orientation, and human interdependence.

The platform helps educators design contexts in which students can identify real problems, make choices, pursue different learning journeys, document their work, seek feedback, and contribute to individual and group projects. AI is not used merely to produce answers. It serves as a thinking partner, a design assistant, and a means of capturing evidence of students’ interests, decisions, development, and contributions.

The system is still developing. Like the educational paradigm it supports, it will need experimentation, criticism, and revision.

But its purpose is clear: technology should not make the old grammar of schooling more efficient. It should help us create forms of education that were previously difficult to imagine or sustain.

What Should Schools Do?

The answer will not come from another universal list of knowledge, skills, or competencies. It will not come from banning AI, adding an AI course, or purchasing another packaged program.

Schools should create conditions in which young people can discover what matters to them, learn what they need in order to pursue it, work with others, respond to uncertainty, and create value beyond themselves.

This will not happen everywhere at once. It does not need to.

A few educators can begin. A few students can demonstrate what is possible. Their work can be documented, shared, questioned, and improved. Other courageous minorities can form in other schools and communities.

Education has often tried to prepare students to succeed in the world as it exists. In the age of AI, education must also help them imagine and create a world that does not yet exist—and ensure that the world they create is better for others as well as for themselves.

Based on the third meeting of the Courageous Minority group and the participants’ reflections on AI, student agency, self-directed learning, human interdependence, and educational change.