Why One American Institution Rebuilt Itself, and Why Another Must
It is the start of another school year, which means many districts are introducing another set of priorities, programs, and expectations. Teachers are asking a reasonable question: Really, one more thing?
For more than thirty years, I have watched education respond to nearly every challenge by adding something new to the existing system. At what point do we stop adding and begin redesigning?
I was thinking about this, and it occurred to me that the place we find ourselves in education is very much like where the car industry found themselves in the early 1970’s and 1980’s. American manufacturers were being thoroughly challenged by Japanese reliability, efficiency, and fuel economy. The 1980 NBC documentary “If Japan Can, Why Can’t We?” helped reintroduce Deming to American executives. Ford brought him in during the early 1980s, as part of a much broader transformation of Ford’s management and manufacturing culture.
Traditional American mass production tended to emphasize volume, quotas, unit costs, and end-of-line inspection. Defects were found after they had already been produced.
The Japanese approach worked to prevent defects inside the process. Workers could identify abnormalities, stop production, investigate causes, and correct the process before thousands of additional defects were created.
The central question changed from: “Who made the defective part?”
to:
“What allowed this process to produce a defective part?”
The shift was not simply, “Let’s inspect cars more carefully.” That would have been the equivalent of giving inspectors brighter flashlights while continuing to build cars, that, by many buyer’s opinions were inferior to American cars.
The person doing the job was not treated merely as a pair of hands following instructions.
The team was expected to notice problems, contribute solutions and stop the assembly line as needed when it appeared that one particular item on the assembly line required additional attention. In a traditional factory, stopping the assembly line was failure. In Toyota’s system, allowing a known defect to continue down the line was failure. Toyota called this concept jidoka—automation with a human touch. The comparison raises an important question for education: What would change if learning problems became visible early, educators had the authority to respond, and the system had the capacity to act before those problems compounded? It sounds like we are ready for our own New Learning Operating System.
Deming’s principles called on leaders to drive fear out of the organization, reduce dependence on end-of-line inspection, break down departmental barriers, and accept responsibility for the system producing the results. Ford increasingly adopted those ideas through a stronger emphasis on quality, teamwork, employee participation, and management accountability.
Principles of our New Learning Operating System
1. The problem has to be visible.
The process only works, if the problem developing on the assembly line is visible to all team members.
In education, we lack this advantage: just as your friend can't observe what's happening in your mind while you discuss a soccer match, you can't see inside theirs either. In education, that is one of our biggest challenges: We can’t see what is actually happening in our students’ minds. The truth is that teaching has always involved a remarkable amount of educated guesswork.
Great teachers become exceptionally skilled at reading facial expressions, body language, participation, written work, and conversations. They learn to sense when a class is engaged, confused, or beginning to drift. But even the very best teachers are still making inferences. They cannot directly observe understanding. They infer it from behavior.
Under the old Teaching Operating System, a student’s learning often becomes visible only after a test has been graded, an assignment evaluated, or a project reviewed. By then, the misunderstanding may already have shaped everything that followed.
W. Edwards Deming described a remarkably similar problem in manufacturing. In Out of the Crisis, his third principle states: “Cease dependence on inspection to achieve quality. Eliminate the need for inspection on a mass basis by building quality into the product in the first place.”[1]
Applied to education, the lesson is clear: we should not depend primarily on tests to discover whether learning occurred after instruction is complete. We should build visibility, feedback, and responsiveness into the learning process itself.
Imagine asking a physician to diagnose a patient without the benefit of X-rays, blood tests, MRIs, or ultrasounds. That physician would rely on experience and observation, but they would also acknowledge that much of the picture remained hidden. For generations, teachers have worked under similar conditions.
Think about it. If learning were visible, teaching would no longer depend so heavily on inference. Teachers could see what each student understands, where misconceptions are forming, which prior knowledge is missing, and who is ready for greater challenge. Small-group instruction could be organized around actual evidence. Feedback could become more timely--teacher attention could become more strategic.
2. Agency was required to immediately resolve the identified issue.
Making a problem visible was not enough. The people closest to the work also needed the authority to act.
When a worker identified a problem on a Deming-inspired assembly line, that worker was empowered to stop production. This was not a minor privilege. Stopping the line could interrupt the progress of every component behind it and potentially affect the many downstream processes dependent upon it. The immediate cost could be substantial, placing production targets and delivery dates at risk. Yet the worker was still expected to act.
The system recognized that allowing a known defect to continue downstream would ultimately be more costly than interrupting production long enough to address it. Quality could not improve if workers were expected to see problems but remain silent because stopping the line was inconvenient.
Education faces a similar tension. Schools operate within systems of vertical and horizontal alignment. Curriculum maps, pacing guides, grade-level expectations, testing schedules, promotion requirements, and graduation timelines all depend upon students progressing according to an established sequence. In some classrooms, the pacing guide can feel as though it comes complete with a stopwatch, counting down to the moment when every student is expected to be ready for the “big test.”
But once the problem became visible, would the educator closest to the learner have the agency to act? Teachers routinely recognize when a student is confused, lacks essential prerequisite knowledge, or is beginning to fall behind. Yet many feel they do not have the authority or flexibility to interrupt the established instructional sequence for one or two students. The class must continue. The pacing guide must be maintained. The next unit is waiting. The assessment date does not move.
The traditional system is organized around the assumption that the group will move together. Common pacing, whole-class instruction, grade-level schedules, and shared assessment dates depend upon that collective movement. When a teacher exercises meaningful agency on behalf of one learner—slowing down, revisiting prerequisite knowledge, changing the sequence, or allowing additional time—the action can disrupt the alignment upon which the rest of the system depends.
In other words, responding appropriately to an individual learner can appear to “blow up” the group-based structure of pacing and alignment. This creates an impossible tension for educators. They may know that moving forward will deepen a student’s misunderstanding, yet stopping to respond may place the rest of the class behind schedule. They are asked to exercise professional judgment inside a system that leaves little room for the consequences of that judgment.
Educators may therefore be permitted to identify the problem without being genuinely empowered to alter the process producing it.
That is the educational equivalent of installing an Andon cord but designing the factory so that pulling it threatens the entire production schedule.
Giving educators agency does not mean abandoning standards. It means allowing an educator to say:
“This learner is not ready for the next step. We need to respond before the system moves on.”
But genuine agency requires more than permission. It requires a learning system in which responding to one student does not depend upon stopping progress for everyone else.
“The question is no longer simply whether educators can see that learning has broken down.
The question is whether the system allows them to act without destabilizing everything around them.”
3. Capacity was required to resolve the problem
--Visibility made problems impossible to ignore.
--Agency empowered people to act.
--But neither mattered if the organization lacked the capacity to respond.
Stopping an assembly line was never the objective. Restoring quality as quickly as possible was.
When a worker pulled the Andon cord, they were not expected to solve every problem alone. The entire system was designed to support them. Team leaders responded immediately. Engineers were available to diagnose root causes. Standardized procedures guided the investigation. Replacement parts and specialized expertise were readily accessible. The organization had built the capacity to transform information into action.
Without that capacity, stopping the line would have accomplished little more than delaying production.
In education, the word capacity is often associated with additional people, programs, time, and intervention structures. When a student struggles, the response may include MTSS, Response to Intervention, tutoring, specialist support, additional instructional periods, summer school, or other supplemental services. These supports are important, but they generally increase the system’s capacity to respond after a learning problem has already become visible and consequential. Still, the traditional response to intervention comes at a significant cost of time and resources.
That is not how I am defining capacity within a Learning Operating System. Here, capacity means the ability of the learning system itself to identify and address a misunderstanding at the moment it occurs.
A Learning Operating System should be capable of detecting missing prerequisite knowledge, adjusting sequence, offering targeted support, and showing whether the misunderstanding has been resolved.
Rather than waiting for a student to fail an assessment, fall behind peers, or qualify for intervention, an AI-enabled Learning Operating System can detect missing prerequisite knowledge, provide a different explanation, adjust the sequence, offer targeted practice, and monitor whether the misunderstanding has been resolved.
The response occurs within the learning process itself. The student does not have to wait for the next assessment cycle, the next data meeting, an available interventionist, or a scheduled support period. The system can respond immediately, while the learner is still engaged with the concept and before the misunderstanding compounds.
This does not eliminate the need for MTSS or specialized intervention. Some students will continue to require intensive human expertise and additional support. But a Learning Operating System could prevent many ordinary misconceptions from growing into the substantial learning gaps that later require formal intervention.
Traditional Intervention capacity adds support AFTER a student has fallen behind. A Learning Operating System builds responsive capacity directly into the act of learning
The goal is not merely to build a better system for repairing learning later.
The goal is to create a system capable of responding early enough that fewer students require repair.
4. The system had to be redesigned, not patched
Visibility, agency, and capacity only matter if the system is willing to learn from what they reveal. When the same problems recur, the answer cannot always be another inspection, intervention, committee, or reporting requirement. At some point, the process itself must change.
In the 1980s, American manufacturers confronted an uncomfortable truth. Their quality problem was not primarily the result of careless workers or inadequate inspections. It was the predictable outcome of the system they had designed. Japanese manufacturers had built a different operating model, one that made problems visible early, empowered the people closest to the work, and treated quality as something created throughout the process rather than inspected at the end.
American manufacturers could have responded to Japanese competition. By:
· hiring more inspectors;
· imposing tougher quotas;
· demanding that workers try harder;
· adding another quality committee
· creating a new defect-reporting form;
Some manufacturers did, and many continued losing market share. Many of those companies are no longer in business. Education’s response to poor results frequently resembles the weaker industrial response:
· more assessments;
· more interventions;
· more specialists;
· more programs;
· more documentation;
· more professional development;
We continue to preserve the core architecture of the existing Teaching Operating System: the same age-based groups, fixed schedules, common pacing, teacher-centered information delivery, delayed feedback, and advancement by calendar. Instead of stopping and reflecting on what we have created, we add another bolt!
Children are not products, and learning is not manufacturing. But the analogy is useful because public education has often organized students as though common pacing and standardized movement were more important than individual readiness.
But, in many cases, isn’t that what we have largely done as a system?
The remarkable thing about the American manufacturing transformation was not that companies became better at inspection. They became better at designing systems that required less inspection.
Education has an opportunity to ask the same question.
Rather than asking, "How do we become better at remediation?" perhaps we should ask, "How do we design a learning system in which remediation becomes less necessary because misconceptions are identified and addressed before they become learning debt?"
Our New Learning Operating System
What would it look like if we had access to research on learning strategies, tools, and technology that we did not have available to us a century and a half ago? If we were starting today, starting with everything we now know about learning, technology, cognition, and how human development works together, what would we build?
We cannot create a New Learning Operating System by adding more features to the Teaching Operating System.
I would submit that our new learning operating system would be built around a few principles:
1. Learning is visible—no more “guessing” or “assuming”
Today we infer learning from quizzes, tests, homework, and observations. Our new Learning Operating System will continuously reveal current understanding, misconceptions, confidence, effort, pace, readiness, and mastery. Not to judge or measure students—but to guide learning.
2. Progress will replace pacing
Time becomes flexible. Mastery becomes our guiding light and becomes “fixed.” Students move because they understand, not because the calendar says October has arrived.
3. Support becomes embedded
Not the teacher, not the curriculum—Infrastructure. Just like electricity. Like electricity, AI should operate quietly beneath the learning experience, expanding what students and teachers are able to do.
It can help identify misunderstandings, recommend the next appropriate step, provide immediate feedback, support practice, surface patterns, and show teachers where their attention is most needed.
Its value is not that it replaces human judgment. Its value is that it gives human judgment better information, greater reach, and more time to act.
Teachers should no more “teach AI” than they teach electricity. They should use it as an underlying capability that makes a more responsive, visible, and personalized learning system possible.
4. Teacher attention becomes strategic rather than scattered
This may be the biggest shift.
Today, teachers must constantly divide their attention across an entire classroom, often relying on observation, raised hands, incomplete work, or assessment results to determine who needs help.
In our new Learning Operating System, the system would help direct that attention with far greater precision. Right now, teachers are asking themselves, “Who needs me?” The new question will be: “These five students need assistance right now, and here is why.” One student may be missing a prerequisite concept. Another may be ready to accelerate. A third may understand the content but lack the confidence. The teacher is no longer forced to search blindly for the signal within a room where learning is largely invisible.
5. Assessment becomes navigation for EVERY student
· Assessment stops being the report card. It becomes GPS.
· Its purpose is no longer ranking students. Its purpose is directing learning.
· A GPS doesn't tell you you're lost. It tells you where you are, where you're trying to go, and what the best next step is from your current location.
Learning should work the same way.
This shift changes assessment from an event that happens after learning to a process that continuously guides learning. Every student benefits from this navigation, not just those who are struggling. A student who is ready to accelerate receives a different route than one who needs additional support, but both are moving toward the same destination: mastery.
In our new Learning Operating System, assessment becomes a continuous source of guidance. It helps teachers know where to intervene, students understand what comes next, and parents see meaningful progress over time. Instead of asking, "How did I do?" assessment begins answering the far more useful question, "What should I do next?"
Just as we would never use a GPS only after taking a wrong turn, we should not reserve assessment for discovering who has fallen behind. Its greatest value is helping every learner make informed decisions before they drift off course.
So, Where Do We Begin?
The ideas described here may sound radical.
A new Learning Operating System cannot run inside the old Teaching Operating System
Learning becomes continuously visible. Progress is based on mastery rather than calendar time. Assessment guides every learner’s next step. Intervention is embedded in the learning process. Teachers direct their attention where it can make the greatest difference. AI quietly expands the capacity of the entire system.
But these ideas are not radical because they ignore what we know about learning. They are radical because they cannot be fully implemented inside an operating system designed primarily around teaching.
Our current Teaching Operating System begins with the instructional structure:
· a teacher,
· a group of similarly aged students,
· a common lesson, a fixed period, a pacing guide,
· a calendar that determines when the class moves forward.
Our new Learning Operating System begins somewhere else. It begins with the learner.
· What does this student understand?
· What prior knowledge is missing?
· What should come next?
· What support is needed now?
· What evidence will demonstrate readiness to progress?
Teaching does not disappear in this new system. It becomes more precise, more responsive, and more valuable. But teaching is no longer the organizing principle around which every learner must move at the same speed and along the same route.
That is why we cannot simply bolt these ideas onto the current system and expect transformation.
A new Learning Operating System cannot operate effectively when it is forced to obey the rules of the Teaching Operating System it was intended to replace.
Before you “swipe left” and disregard these important ideas, recognize that we do not need to replace the entire Teaching Operating System to begin building the next one.
None of this suggests that a district should dismantle its existing instructional model overnight. That would not be bold. It would be reckless.
But neither should the difficulty of system-wide change become an excuse for doing nothing.
The practical next step is to create a protected prototype within the current system: one grade, one subject, one group of educators, and a clearly defined period of time.
Within that prototype, the district can temporarily relax selected rules of the Teaching Operating System and test the principles of a new Learning Operating System.
· Can learning become more visible?
· Can students progress from their actual starting points?
· Can teacher attention become more targeted?
· Can assessment guide the next step for every learner?
· Can AI expand capacity without diminishing the teacher’s role?
· Can small misunderstandings be resolved before they become accumulated learning debt?
These questions cannot be answered through another committee, conference, or strategic plan.
They must be tested under real conditions, with real teachers and real students.
That is the purpose of a carefully designed proof-of-practice prototype: not to bet the future of the district, but to begin gathering evidence about what that future could become.
The question isn’t whether these ideas are possible
· The research exists.
· The technology exists.
· The need has never been greater.
The only remaining question is whether we are willing to continue optimizing a teaching operating system designed for another century...or begin building a new Learning Operating System designed for this one.
The automobile industry did not transform all at once. It learned through disciplined experiments, working models, and the gradual expansion of practices that produced better results. The factory still had to produce cars while the new system was being established.
Education faces the same practical reality. We cannot close our schools for several years while we design their replacements. Students will arrive tomorrow morning, and teachers must be ready to serve them.
But we can create protected places inside the current system where different operating rules can be tested.
· One grade.
· One subject.
· One team.
· One carefully designed prototype.
The goal is not to tear down the schoolhouse.
It is to discover what becomes possible when the people inside it are no longer constrained by an operating system built for another era.
This isn't a finished idea.
...It's an invitation.
For more than thirty years I have worked alongside educators trying to improve teaching.
This article represents a different question:
What if it is time to redesign the operating system that supports learning?
I don't pretend to have every answer. In fact, I hope I don't.
The best operating systems are not designed in isolation. They are built by practitioners, refined through evidence, challenged by thoughtful critics, and improved through real-world testing.
“...Success is progress toward a worthy goal...”
Continue the Conversation.
If this article sparked your curiosity—even if you disagree with parts of it—I would welcome the conversation. Because I suspect the next chapter in education won't be written by one organization.
It will be written by people willing to build it together. I'm actively looking for educators, researchers, technology partners, architects, and philanthropic organizations interested in helping shape what a new Learning Operating System might become.
Explore a Prototype
If your district is interested in exploring these ideas through a carefully designed Proof of Practice prototype, I'd be happy to discuss what that could look like.
The goal isn't another initiative--It's building evidence.
One classroom, one grade, one subject at a time.
Bring the Conversation to Your District
Many district leadership teams are asking the same questions explored in this article.
If your administrative retreat, leadership meeting, board workshop, or professional learning day would benefit from a conversation about Learning Operating Systems, AI, mastery, and the future of school design, I'd welcome the opportunity to be part of that discussion.
[1] W. Edwards Deming, Out of the Crisis, 2nd ed., MIT Press, 2000, p. 23. ©Copyright 2026, Easter Consulting Group