Category: featured

For a long time, one of the privileges of being a professional was being left alone.

You developed expertise. Others trusted you to exercise it. A doctor, lawyer, salesperson, engineer, teacher or other skilled professional didn’t need someone constantly looking over their shoulder. Autonomy was evidence of professional status.

There is something important about that idea worth preserving. But the way professionals exercise that autonomy is beginning to change.

Increasingly, professionals are creating a record of their own work.

Sales calls are recorded and revisited. Product teams preserve customer interviews and research sessions. Meetings are transcribed and turned into decisions and action items. Police officers record interactions. Lawyers increasingly work from searchable digital records. Investors record conversations and use AI to surface information they might otherwise forget.

The reasons vary, but there is a common idea underneath them:

Memory is no longer the best evidence available to us.

Recording can strengthen judgment, not replace it

Historically, recording work could feel like monitoring. Someone else wanted a record so they could determine whether you did your job correctly.

That’s still a legitimate concern.

But something different is happening now. Professionals can increasingly create records for themselves.

A salesperson can revisit a customer conversation rather than rely on hurried notes. A product manager can return to exactly what a customer said. A team can compare what it remembers from a meeting with what actually happened.

The professional is still making the judgment. The record simply gives that judgment better evidence.

That becomes particularly valuable during periods of rapid change.

Education is entering one now.

Teachers are reconsidering what students should do with AI, what they should do without it, how assignments need to change, what should happen inside versus outside the classroom, and what counts as convincing evidence that a student has actually learned something.

There isn’t a settled playbook for many of these questions.

When practice is changing quickly, the ability to revisit reality becomes more valuable.

AI makes context more valuable too

There is another reason professional practice is changing.

AI is only as useful as the context it has.

Ask an AI system a generic question about your work and it can give you a generic answer. Give it the actual conversation, meeting, customer interaction, lesson or other evidence surrounding that work and it can become considerably more useful.

It can summarize what happened. Help identify patterns. Turn conversations into documentation. Compare what happened with an intended practice. Surface moments worth revisiting. Help a professional reflect on decisions or prepare for what comes next.

This doesn’t require surrendering judgment to AI.

In many cases, it does the opposite.

The record provides the context. AI makes the record easier to work with. The professional decides what it means.

That combination can make expertise more powerful.

Education has an unusual evidence problem

Teachers may benefit from this shift more than many professionals because so much of their work disappears.

A teacher can spend six hours interacting with students, making hundreds of small decisions, adapting explanations, noticing confusion, managing relationships, asking questions and watching students attempt to apply what they know.

At the end of the day, remarkably little evidence of that work remains.

There may be assignments, grades, test scores and data from learning platforms. But those are outputs and proxies. They don’t capture much of the professional work that produced them—or the context necessary to understand them.

The teacher is left largely with memory.

So are the students.

So are coaches and leaders trying to help.

Recording changes what a teacher can do with their expertise

Imagine instead that creating a record becomes a normal part of professional practice.

A teacher doesn’t need to carefully watch every student simultaneously. They can revisit moments they missed.

They don’t have to reconstruct a lesson from memory before reflecting on it. They can examine what actually happened.

They can give AI appropriate evidence from their own work instead of asking it to make recommendations without context.

They can turn portions of their work into documentation rather than creating that documentation again from scratch.

They can capture students explaining, creating and applying what they know—not simply the scores produced afterward.

And when useful, they can choose to bring selected evidence into coaching, collaboration, conversations with parents, or other professional work.

Recording can turn an experience that disappears into an evidence base the educator can use.

A new kind of professional autonomy

This requires us to update an old assumption.

Professional autonomy doesn’t have to mean: Trust me because there is no record.

It can increasingly mean: Trust me to create evidence, use it responsibly, improve my judgment with it, and decide when it should inform others.

That is a different kind of professionalism.

It also carries new responsibilities. Recording must be done with tools appropriate for education. Students and colleagues need protection. Personal conversations don’t belong in the record. Access must be limited. Sharing needs a purpose. Evidence shouldn’t become permanent simply because storage is cheap.

The answer to those concerns shouldn’t necessarily be to create less evidence.

It may be to develop stronger professional norms for creating and governing it.

Sales, product development, law enforcement, investing and other fields are already adapting to a world in which more professional work can be captured, revisited and made useful by AI.

Education won’t adopt the same practices in exactly the same way. It shouldn’t.

But the underlying shift is coming to education too.

Professionalism once meant being trusted to work from your expertise. The next version may mean being trusted to create the evidence that makes your expertise even better.

There is a growing debate about how much technology belongs in education.

Some imagine increasingly AI-driven classrooms where students receive highly personalized instruction through technology. Others are moving in the opposite direction—limiting devices, bringing work back into the classroom, and putting more emphasis on discussion, relationships and direct interaction with teachers.

We think both approaches can miss something important.

Regardless of how much technology is in the classroom, teaching remains deeply human.

And much of what makes it human is barely represented in the data we use to understand schools.

Teaching is more than delivering instruction

It’s tempting to describe the teacher’s human advantage as motivation.

Technology can deliver instruction, the argument goes, but students still need a caring adult to motivate them to do the work.

That’s certainly part of teaching. But it dramatically understates the role.

Teachers exercise judgment constantly.

They decide whether a student is confused or simply hesitant. Whether to explain something again or let a student struggle. Whether a wrong answer reflects a misconception or a careless mistake. Whether a class needs to move forward or stay with an idea. Whether a student needs encouragement, challenge, space, structure or a different explanation entirely.

They interpret context.

They notice relationships between students. They understand what happened yesterday. They recognize when something happening outside the lesson is affecting what happens inside it. They adjust to a room full of people whose needs cannot always be anticipated in advance.

And they do this continuously.

Teaching isn’t simply delivering the right content. It is exercising judgment about people in context.

Our data captures very little of this

Schools have become extraordinarily good at producing data.

We have test scores, grades, attendance, behavior records, learning-platform analytics, engagement metrics, intervention data, surveys and increasingly AI-generated analyses of student work.

All of these can tell us something useful.

But notice what is easiest to capture.

It’s usually the output of the educational process rather than the human work happening inside it.

A test score can tell us how a student performed. It doesn’t tell us what a teacher noticed three weeks earlier that caused them to change their approach.

A platform can tell us how long a student spent on an activity. It doesn’t know why the teacher chose that activity, what they noticed while the student completed it, or why they abandoned the next activity they had planned.

Even increasingly sophisticated AI is usually analyzing whatever evidence we’ve made available to it.

If the evidence doesn’t contain the human work, the human work remains invisible.

What remains invisible eventually matters less

This creates a deeper problem than simply failing to give teachers enough credit.

Organizations naturally manage what they can see.

Researchers study what they can observe. Leaders build systems around what they can measure. Policymakers create accountability around available data. Technology companies optimize the variables their systems can capture.

Over time, the representation of the work can begin shaping the work itself.

If test scores are the strongest evidence available, test scores gain influence.

If platform activity is visible, platform activity gains influence.

If AI can easily analyze student outputs but can’t see the professional judgment surrounding them, the outputs gain influence.

None of this requires anyone to believe that human judgment doesn’t matter.

It simply requires human judgment to remain poorly represented.

What remains invisible risks being increasingly diminished.

AI makes this more important

AI makes the problem particularly urgent because it dramatically increases our ability to act on data.

Systems can analyze thousands of student interactions, recommend interventions, generate lessons, personalize activities and identify patterns no individual could process manually.

That can be enormously useful.

But greater analytical power doesn’t fix an incomplete evidence base.

It amplifies whatever evidence we provide.

If the data represents only part of teaching and learning, increasingly powerful systems may become increasingly effective at optimizing that part.

The answer isn’t necessarily less technology.

It’s better evidence.

We need an evidence base for the human work

For most of education’s history, much of teaching has disappeared the moment it happened.

The teacher remembers some of it. Students remember some. An observer occasionally witnesses a lesson. Everyone else sees the outputs.

Recording changes that.

It can preserve enough of the classroom for educators to revisit the judgments, interactions, adaptations and context that otherwise disappear.

A teacher can examine why a discussion worked. A coach can understand the context behind a decision. A team can look at real examples together. Leaders can build a better understanding of practice without reducing it to a score. AI can work from richer context instead of only the data that happened to be easiest to collect.

Recording doesn’t make human judgment measurable in the same way as a test score.

That’s not the goal.

It makes human judgment visible enough to remain part of the conversation.

This isn’t a choice between humans and technology

A highly technological classroom still needs human judgment.

So does a technology-light classroom.

The important question isn’t which side of that debate wins. It’s whether the evidence we create adequately represents the things we say matter.

If judgment matters, we need evidence of judgment.

If relationships matter, we need enough context to understand them.

If adaptation matters, we need to be able to see it.

If teachers matter for reasons that extend far beyond delivering information and motivating students, those reasons need a place in the evidence base from which schools make decisions.

Otherwise, we risk building the future of education around the parts of teaching that are easiest to count.

Teaching will remain deeply human. The question is whether the evidence shaping education will be human enough to recognize it.

Schools have more data than ever.

Test scores. Grades. Attendance. Behavior incidents. Surveys. Intervention data. Learning-platform analytics. Engagement metrics. Evaluations. And now an expanding layer of AI-generated analysis.

All of it exists for good reasons. Schools need to know whether students are learning, whether programs are working, and where support is needed.

But something strange has happened as education has become more data driven.

The people inside schools don’t seem to trust each other more. They seem to trust each other less.

Teachers feel less trusted by leaders. Leaders struggle to establish enough trust to create change. Coaches need teachers to open their classrooms and expose uncertainty, yet teachers can reasonably wonder how that information might eventually be used. Parents question whether schools are meeting their children’s needs. Even students increasingly have reasons to question how teachers know whether work represents their actual understanding.

More data was supposed to give us greater certainty.

Instead, it may be contributing to mistrust.

The problem isn’t data. It’s what the data represents.

Most educational data is a proxy.

A test score represents something about learning. An attendance record represents something about participation. A behavior incident represents something that happened between people. A classroom observation represents a small sample of teaching. A platform metric represents an interaction with software.

Each can be useful.

The problem begins when the proxy becomes more influential than the reality it was intended to represent.

A leader looking at a dashboard may have excellent information and still know very little about what happened between a teacher and students that day.

A policymaker can see achievement data across thousands of students without ever seeing the teaching that produced it.

A parent can see grades and test scores while knowing surprisingly little about how their child thinks, participates, struggles or learns.

A teacher can see a completed assignment without knowing how much of it was actually produced by the student.

The people making decisions increasingly have data about classrooms without having much evidence from classrooms.

That distinction matters.

Everyone fills in what the evidence leaves out

When evidence is incomplete, people naturally interpret it.

A disappointing score might mean instruction was ineffective. Or the assessment may not represent what students understand.

A struggling initiative might mean teachers aren’t implementing it. Or implementation may be colliding with realities leaders can’t see.

A classroom observation might reveal a problem. Or it might capture an unusual ten minutes.

A student’s polished assignment might demonstrate mastery. Or AI may have done much of the thinking.

Without better evidence, each person fills in the missing context from their own perspective.

Teachers believe leaders don’t understand classrooms.

Leaders believe teachers may be resisting necessary change.

Parents wonder whether schools are telling them the whole story.

Students learn that the evidence used to judge their learning doesn’t always represent what they actually know.

Nobody necessarily has bad intentions.

They simply aren’t looking at enough of the same reality.

Relationships used to carry more of this burden

Education has traditionally depended heavily on relationships.

A principal knew their teachers. Teachers knew their students. Parents knew the school. Coaches developed credibility over time.

When information was incomplete, relationships supplied trust.

Relationships still matter enormously.

But they are being asked to carry a burden they increasingly cannot carry alone.

Schools are larger and more complex. Staff turnover disrupts relationships. Leaders are expected to demonstrate measurable improvement. Parents have access to more information and more ways to question institutions. State and district accountability systems demand evidence. AI is making it harder to know whether student work demonstrates actual understanding.

Telling everyone to build stronger relationships doesn’t resolve those pressures.

Trust increasingly needs evidence too.

Mistrust makes change harder

This becomes particularly damaging when schools try to change.

A district adopts new curriculum. Leadership wants instruction to shift. Coaches are asked to support implementation. Teachers encounter problems in classrooms.

What happens next depends heavily on trust.

Do teachers believe leaders understand what implementation actually requires?

Do leaders believe teachers are making a genuine effort?

Can coaches see enough real practice to help?

Can teachers admit what isn’t working without worrying that honesty will be interpreted as failure?

When trust is weak, everyone becomes more defensive.

Leaders ask for more documentation.

Teachers protect their autonomy.

Coaches struggle for access.

Organizations respond by collecting still more data.

And the cycle continues.

What if schools had better evidence of reality?

The answer isn’t to eliminate data.

It’s to fill the evidence gap underneath it.

Imagine if educators routinely created evidence of their own work.

Teachers could revisit what actually happened rather than reconstructing it from memory. Students could be observed explaining and applying what they know. Coaches could work from real examples teachers chose to share. Teams could calibrate around actual practice rather than abstract descriptions of it.

Leaders wouldn’t need access to every recording. Parents wouldn’t need a window into every classroom.

The important change would be that reality existed in the evidence base.

When a proxy raised a question, there would be richer evidence available to understand it.

When educators disagreed, they could sometimes return to what actually happened.

When AI analyzed teaching or learning, it could work from better context.

And when professional judgment mattered, educators would have evidence with which to exercise and explain it.

Trust needs something to stand on

For years, schools have tried to become more data driven.

We don’t think that was a mistake.

But if the data shaping decisions doesn’t adequately represent teaching and learning, collecting more of it won’t necessarily produce greater trust.

It may do the opposite.

The next step may be to create an evidence base that begins much closer to the classroom—and gives educators themselves a meaningful role in creating and governing it.

Because relationships still matter.

Professional judgment still matters.

Data still matters.

But when consequential decisions are being made everywhere, trust increasingly requires shared evidence of reality.

Education has never had more data about learning.

Every platform produces it. Students generate scores, levels, completion rates, response times, assignments, assessments, transcripts, engagement measures and increasingly AI-generated analyses of their performance.

The assumption underneath much of this is simple: if we collect enough signals, we can know whether a student is learning.

AI is exposing a problem with that assumption.

Students can now produce remarkably sophisticated work without necessarily possessing the understanding that work appears to demonstrate. They can generate an essay, solve a problem, summarize a reading, write code, answer questions and complete assignments with assistance that is increasingly difficult to detect.

This isn’t simply a new cheating problem.

It’s an evidence problem.

If the work we measure is increasingly disconnected from what a student can actually do, then measuring it more precisely won’t tell us whether learning occurred.

Learning needs to be verified.

All learning data is a proxy

A score isn’t learning.

Neither is a grade, a completed assignment, a platform level, a transcript or an AI analysis.

They are evidence from which we make an inference about learning.

Usually, that is useful. Sometimes it is extremely useful.

But the distinction matters.

A student can earn a high score without retaining much of what they learned. They can complete an assignment with extensive help. They can perform well on familiar problems and struggle when the same idea appears in a new context.

And now AI can help produce many of the artifacts we’ve traditionally treated as evidence of understanding.

That doesn’t make the artifacts worthless.

It makes the inference less certain.

AI makes the appearance of learning abundant

For years, schools have worried about students finding answers online.

Generative AI is different.

It doesn’t just provide the answer. It can participate in the intellectual work.

It can brainstorm, explain, outline, revise, calculate, translate, summarize, write and reason alongside the student.

Used well, that can be enormously valuable for learning.

But it also makes the finished product increasingly ambiguous.

Did the student write that paragraph?

Perhaps that’s no longer even the right question.

The more important question may be:

What can the student understand and do without the artifact speaking for them?

That’s much harder to determine from the artifact itself.

Learning has to be observed

At some point, educators need to see students demonstrate what they know.

Can they explain an idea in their own words?

Can they apply it when the problem changes?

Can they defend a conclusion?

Can they connect concepts?

Can they create something new?

Can they recognize when their approach isn’t working and adapt?

Can they transfer what they learned into a situation they haven’t seen before?

These aren’t perfect measures of learning either.

But they bring us closer to the thing we’re actually trying to understand: what capabilities now exist in the student?

That means assessment can’t happen entirely through products submitted after the fact.

More demonstrations of learning may need to move back into the classroom, where educators can observe the process as well as the result.

But one teacher can’t observe everyone

That creates an obvious practical problem.

Put 30 students in a classroom and ask them to demonstrate understanding through discussion, explanation, problem solving, creation or application.

A teacher can watch a few closely.

They cannot watch everyone.

Small groups make this even harder. Some of the richest evidence of learning may be happening simultaneously around the room while the teacher is working with one group.

Historically, most of those moments simply disappeared.

The teacher saw what they could see.

Everyone else eventually submitted something that could be scored.

Recording changes that constraint.

Students can demonstrate learning while the teacher is somewhere else in the room. Educators can revisit selected moments afterward. They can sample evidence, investigate uncertainty, compare performance over time, or return to something that deserves another look.

Recording doesn’t replace teacher observation.

It extends it.

That changes what we should collect

If our goal is verification, schools may need a different evidence base.

Instead of collecting primarily the outputs that are easiest to score, we can collect more evidence of students actually doing the things we hope they are learning to do.

Not just the essay, but the student explaining its argument.

Not just the answer, but the student working through an unfamiliar problem.

Not just the presentation, but the discussion that tests whether the ideas hold up.

Not just an AI-generated analysis of performance, but evidence a teacher can examine and interpret for themselves.

That evidence doesn’t all need to become permanent data.

Much of it may only need to exist long enough for an educator to make a better judgment.

The goal isn’t to build a permanent record of every student.

The goal is to give teachers enough evidence to know when learning is real.

Verification makes teacher judgment more important

There is an irony in all of this.

AI may automate more instruction. It may generate more assessment. It may produce more data and analyze that data more effectively than any teacher could.

Yet that may make teacher judgment more important.

Someone still has to determine whether the signals correspond to reality.

Someone has to look beyond the polished artifact, score or dashboard and ask whether the student can actually understand, explain, create, apply and transfer what they know.

AI can help educators examine the evidence.

Recording can help them capture it.

Data can help identify where to look.

But ultimately, learning isn’t something we can simply measure into existence.

It has to become visible in what a student can actually do.

And increasingly, educators will need the evidence to verify it.