For Teachers

This is our reading of the U.S. News 2027 Best Colleges ranking methodology. See the rankings page →

In the 2027 rankings, Harvard is not first. Neither is Princeton. MIT is.

The 2027 Best Colleges rankings were released September 22, 2026. The top ten national universities:

1. MIT

2. Princeton

3. Harvard

4. Yale

5. Caltech and Stanford, tied

7. University of Pennsylvania

8. Duke

9. Johns Hopkins, Northwestern, and University of Chicago, tied

The top ten national liberal arts colleges:

1. Williams

2. Amherst

3. Bowdoin, Claremont McKenna, Pomona, Swarthmore, and United States Naval Academy, tied

8. Harvey Mudd

9. United States Military Academy

10. Wellesley

Two movements worth naming. MIT took the top spot, ending the run Princeton had held since the 2012 edition. Caltech rose from eleventh to fifth.

The question worth another ten minutes is not which school moved. It is why the list moved. The answer is that U.S. News changed its measuring instrument twice in three years, and almost nobody reported the second change.

The first change, in the 2024 edition, was to the ruler. The second, this year, was to who each school gets held up against.

The 2027 edition rebuilt its peer groups on the 2025 Carnegie Classification, replacing the 2021 version it had been using. The effect: roughly one in five previously ranked schools now sits in a different category than it did last year. Every school is ordered within its own group, so moving groups means facing an entirely different set of competitors.

U.S. News says so itself: 2027 positions should not be compared directly against prior years.

Carry that caution back up to the list above. MIT taking the top spot and ending a fifteen year run is true. It is also more complicated than the headline. When one in five schools changed peer groups and the publisher says year over year comparison does not hold, the question "who passed whom" loses part of its meaning. Knowing this is not a reason to ignore the news. It is a reason to know what the news rests on.

Your students will quote this list at you this month. It is not the ranking you memorized. Here is what moved, why, and what to say.

In the 2024 edition, U.S. News made the largest methodology change in the list's three decades. In the 2027 edition it swapped an indicator and, more consequentially, rebuilt the peer groups schools are ranked within. If you are still advising from the old intuitions, your guidance will run half a step behind, and students can feel it.

A note on the naming, which misleads people every year. U.S. News publishes in the fall and names each edition for the following year. The 2027 edition appeared in September 2026. The overhaul described below carries the 2024 name and actually landed in September 2023.

First, the five that were dropped, and why dropping them was defensible.

Dropped: alumni giving rate, high school class standing, the share of faculty holding terminal degrees, undergraduate class size, and the proportion of graduates who took out federal loans.

Alumni giving rate measures whether the people a school admitted decades ago have money now, and whether they feel warmly toward the place. That is a proxy for prestige and wealth. It is not a proxy for teaching.

Undergraduate class size feels like it measures teaching and actually measures resources. Small classes may well be better, but a school's ability to run many of them first tells you it can afford enough faculty. And whether what happens inside those small classes is any good is a question the metric never asks.

Share of faculty with terminal degrees, same shape. It measures the hiring budget.

Those three, and the other two, share one defect. They measure what a school possesses, not what a school does to a person. Dropping them was a decision that holds up.

Now the four that were added, and how to use them in a session.

Added or heavily upweighted: first-generation graduation rates, graduate earnings compared with high school graduates, Pell Grant recipient graduation rates, and graduation rate performance. In the 2027 edition, earnings by major replaced the median federal loan debt indicator. Outcome measures now carry fifty two percent of the total weight.

In a counseling conversation, the most useful of these is graduation rate performance, because it has a control built in. It subtracts the rate a model predicts from the student profile a school admits. What remains is the closest quantity on the list to "how much did this school's teaching add."

For a student seized up by rank anxiety, that number does something very concrete. It converts the unanswerable question "is this school good" into the answerable question "what does this school do for people like me." The second question a student can actually work on.

Pell Grant graduation rate and first-generation graduation rate are directly relevant information if any of your students fit those categories. Schools that post strong numbers here usually have an explicit support apparatus, rather than assuming every student arrives already knowing the rules.

Now the honest half. You should give students this part too.

Schools have withdrawn. Columbia University withdrew. The law schools at Harvard and Yale withdrew, criticizing the methodology publicly as they left.

The rankings still correlate strongly with family income and with entering test scores. Which is to say the new method did not resolve the oldest criticism: a "top" list is still substantially measuring whom a school admitted, not what it did for them. The higher education scholar Christopher Newfield makes exactly this argument. The methodology changed; the central criticism went unanswered.

And a whole category never included at all: campus climate, arts opportunities, career and technical education, psychological and emotional support. Plus the structural problem no reweighting resolves, four years compressed into a single number.

Here is the thing the source does not say, and it matters most to your work.

Facing the new methodology, the most valuable thing a counselor can do is not help a student rank twenty schools. It is teach that student to read one metric all the way down.

Ranking twenty schools is something the student cannot do and you cannot do either, because the weighting judgment it requires is itself subjective. But "what graduation rate performance means" and "why that number says more than the raw graduation rate" are things a seventeen-year-old can genuinely learn. Once learned, the skill travels with them for life, into every situation where somebody hands them a ranked list.

One metric read well beats twenty schools ranked badly, by a wide margin.

Try this in your next session.

Do not open the ranking. Open one school's graduation rate performance number, and ask the student what they think it is saying.

Let them reason it out. They will start with "high graduation rate means good," and then you ask, "what if this school only admits people who were always going to graduate?" They will stall for a moment, and then walk themselves to the point. What they learn in that moment holds far better than the conclusion delivered directly.

One aside, unrelated to methodology but directly relevant to this kind of conversation. Acknowledge before you advise. A student arriving with a ranking is usually not processing an information problem. They are processing an anxiety problem. If you skip the step of saying "this is genuinely hard, and a lot of people get stuck right here" and go straight to "here is what you should do," their body cannot take the advice in. We wrote about that sequencing problem in a separate piece.

Last.

The 2024 change was real progress, because it moved the question from "who did this school admit" toward "what happened to the people this school taught." That direction is right and worth acknowledging.

It is still not the whole question. A family that reads only the ranking has outsourced the most important judgment to a magazine. And you, as the person sitting across from the student, are the only one in the chain with a chance to hand that judgment back to them.

### Questions to sit with

1. The last time you talked through school choice with a student, how much of it went to rank and how much to "what do you want out of these four years"? Was that ratio one you chose, or one the student's anxiety pushed you into?

2. If you had to explain in three sentences, to a seventeen-year-old, why graduation rate performance carries more information than raw graduation rate, could you? Whatever part you cannot say is the part you have not fully thought through yourself.

3. Among the students you have advised, has anyone chosen a school that ranked lower but was clearly a better fit? What made them brave enough to choose that way? Can you give that same thing, more deliberately, to the next student?

This is our reading of Henry Ren's dispatch for Bloomberg Weekend, 2026-10-08: China's Young People Are Turning to an 84-Year-Old Cambridge Professor. Read the original →

The facts first, so we are not arguing in the abstract.

In 2025, an 84-year-old emeritus professor of social anthropology at Cambridge, Alan Macfarlane, began posting five-to-ten-minute videos on Xiaohongshu, a Chinese social media platform. Seventeen months later, more than 2.5 million people were following him. Young Chinese users in their twenties and thirties call this stranger from England Dumbledore. In the comment thread they ask him, in volume: how do I face my parents' expectations? Do I study what they want, or what I want? Is an ordinary life worth living at all?

What does this mean for a working teacher?

It means: your students are listening to someone give lessons. Not in the classroom. In their phones. The person they are listening to does not know more than you do, and probably does not know them better than you do. He is simply speaking to them in an order you can learn to speak in, this week.

What he does.

He calls his method social therapy. Three steps.

Step one, acknowledge the problem is real, and that it is not just this one person's problem. When a young person asks "I do not know how to face my parents' expectations," his first sentence is not advice. His first sentence is a version of "this is presumably quite a widespread anxiety and worry for young people in China." He moves the young person off the island labeled "something is wrong with me" and onto a continent labeled "many people are here with me."

Step two, hand over a measure of hope that is not empty. Not "it will all work out." Just: before you decide this is permanent, give yourself one more step.

Step three, offer a concrete, small, doable-today next step. Not "you should be more resilient." Instead: "next time you eat with your parents, try saying this exact sentence to them: ..."

The order is the point.

Most teachers, hearing a student say "I cannot do this," skip step one and go directly to step three. "Here, let's try again." "Listen, in this step you want to think like this..." The intent is good. The teacher believes they are helping. But in that student's body, in that moment, they are using all their strength to hold off a scarier question: "am I a stupid person?". Skipping step one, from their body's point of view, confirms that "this is something I was supposed to already know." The information they take in is not the step you explained. It is "the teacher also thinks I should already know this." Their body closes the door on this lesson, no matter how clearly you explained it.

Validate first, then teach. Reverse the order and everything changes.

In the classroom, concretely.

Imagine you are teaching a math problem and a student says in the middle of class: "I cannot. I am just bad at math."

The step-one-skipping response is: "Do not say that. You can do this. Here, let us walk through this problem step by step." Good intentions, no effect. What the student hears is "the teacher is pressuring me to keep going." The door at this hallway did not get opened; the energy for the next hallway did not get activated.

The three-step response is:

"That this problem looks hard right now is normal. (Step one, acknowledge the problem is real.) Several people in this class are stuck in the same place in this unit. You are not alone.

I believe you can get through it. (Step two, hand over hope that is not empty.) Not that you will understand it today, but that across the next few practices, we will walk it through.

Right now, we are going to do only one thing: do not solve the whole problem. Just tell me what the first second looks like in your head when you look at this problem." (Step three, a concrete, small, doable-today action.)

On the surface, those three sentences take twenty seconds longer than "here, let us try again." In fact, they move the student out of the inner room labeled "I am stupid" and into another room labeled "this problem is hard." Only in that second room is actual learning possible.

The same order applies at parent-teacher conferences.

A parent sits down and the first sentence is "my child has an attention problem."

Step-one-skipping response: "You can try these methods..." Good intentions, no effect. What the parent is hearing is "the teacher has already decided this is a problem with our family." She cannot hear the methods.

Three-step response: "Attention drifting in class is something several children in my class are navigating right now. (Step one.) I believe he will build it up over time. (Step two.) The one thing I want to ask you today is: in the last week, when he sits down for dinner at home, what state is he in? Calm, or already very tired? (Step three.)"

When that parent walks out of the meeting room, what her body felt is not that she and her child were evaluated, but that she and her child were seen. And that she was seen is the deciding factor in whether she will walk the second mile with you.

Why this matters especially for this generation.

The students you are teaching were born into a different moment than you were. The reality they walk into on the day they leave school is urban youth unemployment at 19%, entry-level jobs they trained for already taken by AI, and a particular shape of exhaustion we are calling neijuan in Chinese, in which the whole group works harder and gets less. Many of them, deep underneath, are already asking the questions Macfarlane's comment thread is asking: "is there something wrong with me, as a person?"

A teacher who walks all three steps, every time, leaves inside the student's body an underlying experience: there really are adults who are willing to acknowledge "this is hard" first, and then walk me through it a step at a time. That experience stays with the student longer than any specific lesson's content.

What strikes me about Macfarlane is that he understands the weight of what he is doing is not "transmitting knowledge", he already published the content in his books , but "telling another person that they are not an island." His translator Qin Yuchen said it very precisely in the Bloomberg piece:

> "He has a rare ability to discuss large and sometimes difficult questions in a warm, personal and accessible way." , Qin Yuchen to Henry Ren, Bloomberg Weekend, 2026-10-08

That ability is not reserved for Cambridge professors. You have an entrance to it every day. All you need to do is reverse the order once, then again, then again, until it becomes the first thing out of your mouth.

As a note, this "acknowledge first, then advise" pattern is not only for teenagers. It is the same order used in the Fraga piece on tantrums for parents, with Juli Fraga's "be a detective, not a CEO" formulation. Different age, different setting, same shape.

One last reminder.

Walking all three steps will not pay off visibly every time. Some students need three months. Some need three years. Some of them will think back to the first sentence of that math lesson only after they themselves have become teachers, or parents. But that one sentence is in them. The sentence you skipped is also in them, in another form.

That decision, skip or walk all three, is not actually about those twenty seconds. It is about which of your moves that child will replay, decades later, in front of a child of their own.

### Questions to sit with

1. The last time a student said to you "I cannot," did you skip step one and go to step three, or did you validate first and help second? Looking back, if you had the chance again, how would you open?

2. In the half-second before you respond to a student (or a parent), what are you most often preparing? A method? A correction? Or a sentence that says "I see you"?

3. In your own life, has any teacher or elder ever spoken to you in the order "acknowledge first, help second"? What did that relationship leave with you? Which part of it are you now passing on?

This is Luminis reading Juli Fraga's article in CNBC Make It, published September 6, 2026, moved from the living room into the classroom. Original (English) is here →

You do not have one problem. You have two.

One student is melting down. Crying, throwing something, or silently pulling their head into their backpack. That is the first problem.

Your thirty students will not remember which chapter you taught today. They will remember the day they saw an adult meet a child at eye level.

Photo: Adam Winger / Unsplash

The second is the other twenty-nine. They are watching. Their pens may still be in their hands, but their eyes are no longer on the page. In this moment, you are the deepest lesson of their semester. Not about grades. About how a grown-up meets a meltdown.

Juli Fraga wrote her five steps for parents. Move them into the classroom without changing a word.

Step one: settle. But you do not have five breaths. You have two.

At home, Fraga suggests five belly breaths. You do not get five. You get about two seconds. Your work in those two seconds is one small honest sentence, said only to yourself: "My chest is tight right now." Naming it means you will not speak on its behalf. Skip the naming, and the first sentence out of your mouth is already carrying a small knife.

Two seconds. That is the foundation of everything you do next.

Step two: detective, not CEO.

The CEO posture grows especially easily in a classroom. You are already standing. You are the room's center of attention. You are trained to solve. Let the word "detective" pull you back for one beat. Behind every child labeled "disruptive" is a story you do not fully know. The detective's first sentence is not "not this again." The detective's first sentence is a quiet question, asked inside you: "Why is he shaped like this today?"

This is not softness. You are not letting go of order. You are looking before you manage.

Your goal is not to make the crying stop right now. It is to let those minutes feel less like being abandoned.

Photo: Jordan Whitt / Unsplash

Step three: help name feelings, but not in the middle of the room.

At home you can sit and take your time. In the classroom you cannot. Thirty pairs of eyes is a pressure system. If you ask a child in front of everyone whether they are sad, they are more likely to feel shamed than seen.

So move this step off to the side. Walk over. Lower your body. Bring your voice down until only they can hear it. One line is enough: "Go get some water. I will find you in a minute." The real work of naming happens in the hallway, or after class, in the two minutes you two get alone. This is not letting them off the hook. It is respect. You are telling this child, quietly, that they are not a live demonstration.

Step four: look at yourself.

In the classroom this step sharpens into one particular question: why does this student in particular set me off?

You have a student whose behavior is not the worst in the class, but somehow it is his behavior that spikes your blood pressure. Admit it, and then follow it one more question: who does he remind me of? Was there a version of him next to me when I was ten? Is he touching a wound my own teacher left in me that I have not yet closed?

Asking this does not weaken you. It sets you free. Because once you know that some of the fire is not his to answer for, you can lift it off of him and see, for the first time all week, the eleven-year-old actually sitting in front of you: stubborn, silly, small.

Step five: H.A.L.T., plus an O.

Fraga's acronym: Hungry, Angry, Lonely, Tired. In a classroom, add one letter. O for Overstimulated. Flickering fluorescent lights. Thirty bodies of sound. Period after period after period with no natural break. Sometimes a meltdown is not a moral failure. It is a design failure. This body cannot metabolize what this room is pouring into it.

Once you see this, three-thirty stops being the coincidental hour when "the bad kids" act up. It becomes the predictable hour when tired, hungry, unmoved bodies finally break their seams. You cannot rebuild the schedule alone. But when the hour arrives, you can remember: this is a biological problem before it is a discipline problem. Meet it with that understanding.

Same fifteen minutes, parent conference.

You may say: I have thirty students, I cannot do this for each one. You do not have to. You need one body memory. When a meltdown hits you, first hear your own heartbeat, then look at the child, and only then speak. The other four steps grow out of that one move on their own.

One step back: this move has a partner.

Two weeks ago we published Bowe's "answer-first," about being heard in meetings, parent conferences, hallways. Today's Fraga is its other half: how to let the room settle before anything can be heard. One move makes your words land clearly. This other move lets other people's words be spoken at all.

Your thirty students will not remember which chapter you taught today. They will remember the day they saw an adult, in front of a meltdown, not panic, not shame anyone, put a hand quietly to their own chest, and get low.

### Questions to sit with

1. Which student in your current class draws the strongest reaction from you? What might that reaction be about him, and what might it be about you?

2. The last time a child melted down in your class, what did the other students learn from you? How would they describe you that evening at home?

3. If a colleague were watching you through the doorway during a hard classroom moment, what would they see in the first ten seconds? What part of yourself do you want them to see there?

This is our reading of John Bowe's piece for CNBC Make It, published 2026-09-02: To "make people want to listen to you," do 2 simple things. Read the original →

Companion reading: when society talks down to teaching as "just a teacher," keeping the profession's dignity → Dolly Parton and the "label upgrade"

Start with a scene you know well.

Parent conferences. Twenty families tonight. Each one fifteen minutes, back to back, laid out neatly in a spreadsheet.

The first parent walks into your classroom. You smile, invite her to sit. You open your notebook and start in your usual order: first the shape of the semester's curriculum, then the general classroom dynamic, then where her child fits in that general picture, and then...

You glance at the clock. Twelve minutes are gone.

The one thing you most wanted to tell her, you have not said. The one thing she most wanted to ask, she has not asked.

The bell rings. Next parent is already at the door. You both smile, say "let's follow up by email," and she leaves.

You have been doing conferences this way for seven years.

The problem is not you. The problem is the order.

John Bowe coaches executives on how to speak in high-stakes rooms. In his 2026-09-02 CNBC piece he teaches two moves you can use tomorrow. The second one is especially costly for teachers, because you have been trained your whole professional life to do the opposite. Give the answer first.

Translated into your classroom:

Weak version (your current opening): "Hi, thanks for coming in. This semester our curriculum has been focused on... The class in general has been... Emma's position in the class is roughly..."

Strong version (your first sentence): "Emma is thriving socially this semester and struggling on written expression. There are two things you can do at home."

That one sentence. Nineteen words.

The moment you finish it, the character of the next fourteen minutes changes. It is no longer your one-way report. It becomes her Q and A. She asks "struggling how?" "what are the two things?" "does she know?" You have time to answer. Time to give examples. Time to hear what she sees at home.

This is what a parent conference is supposed to be.

Why teachers especially struggle with this.

Because you have been trained into a "context first" person.

In lesson design, you write the learning goal, then the pedagogy, then the assessment.

In a unit plan, you write the background knowledge, then the new concept, then the practice problems.

In parent emails, you write a warm opening, then the setup, then the point.

This training, in your profession, is right. It makes your teaching and your written communication rigorous.

But in spoken conversation, it becomes a trap.

In parent conferences, faculty meetings, or a three-minute hallway conversation with the principal, the physical length of time is short. Every second you spend on setup is a second subtracted from the thing you actually want to convey.

Bowe's move, translated to three teacher scenes.

Parent conference opening:

Weak: 5 minutes of curriculum overview and class context.

Strong: "Emma is thriving socially, struggling on written expression, and two things at home will help."

Faculty meeting proposal:

Weak: 3 minutes of industry context + your classroom observations + some considerations + then the actual proposal.

Strong: "I want us to move the science fair from May to March. Three reasons."

A one-minute conversation with the principal:

Weak: "Hi, sorry to bother you, I know you are busy, so I have been thinking about..."

Strong: "One thing I want your read on." + a single sentence that says it.

Notice the structure of each.

You have not thrown away your pedagogical thinking. You have moved it to second position.

Answer first. Context after. If the parent asks, you have it. If she does not, the second thing becomes one sentence in an email she reads at home.

You will notice parents leave with a different face.

Before, they used to smile politely, thank you, and walk out with a vague impression.

With the new opening, they leave with one thing they can actually do, two things they now notice about their child, and a feeling that this teacher has actually seen my kid.

Same fifteen minutes. Your first sentence decides whether it is a briefing or a conversation.

One more step back.

This move is one half of a pair. The other half is Dorie Clark on Dolly Parton, which we published last week.

When society talks down to teaching as "just a teacher," you need Dolly's lesson. You need a quiet, non-fighting move that rewrites what "teacher" actually means. "You are right, I do only one thing. Which is why I have gotten pretty good at it."

When you are in front of a parent, a principal, or a colleague and you want to actually be heard, you need Bowe's lesson. You need a first sentence that puts the thing on the table. "Emma is thriving socially, struggling on written expression, two things at home will help."

One move to hold the dignity of the profession. One move to be heard cleanly in every fifteen-minute window. Both are for you and for your students.

### Questions to sit with

1. If you had to say the single most important thing about one of your students in one sentence to their parent, what would you say? Have you ever actually said it that way?

2. Recall your last meeting or parent conversation where your first sentence was context, and you never quite got to the point. What did you lose?

3. Whose first sentence, in your career as a teacher, has stayed with you? What made it land?

This is our reading of Dorie Clark's piece for CNBC Make It, published 2026-08-31. Read the original →

Companion reading: how to make people want to listen to you → John Bowe and "give the answer first"

Family dinner.

Your cousin, the tech founder, raises his glass and says, to the whole table, "You know how it goes: those who can, do. Those who can't, teach."

Everyone laughs. You pull the corner of your mouth up, too. You do not push back. It feels like just a joke, and pushing back would seem thin-skinned. You take a sip of water and steer the conversation elsewhere.

This is not the first time. It will not be the last. Somewhere in the last generation, no one is quite sure which year, "teacher" moved from "a respected identity" to "a profession you can casually joke about at dinner."

You can keep letting it pass. Most people do. But every one of those quiet passes lays the next brick for the pass after that.

There is a move worth learning.

Dorie Clark, who taught communication for over a decade at Columbia Business School and Duke Fuqua, gave a very old move a very new name: the "label upgrade." She teaches it using a moment from a 1970s Barbara Walters interview of a twenty-two-year-old Dolly Parton. Walters used the word "hillbilly" to file Parton away, meaning to embarrass her. Parton did not push back and did not get mad. She smiled and said, "I think you probably are. But we were very proud people, people with a lot of class. It was country class, but it was a great deal of class."

One, she owned it. She did not say "I am not." Two, she rewrote what it meant. She moved "hillbilly" from "backward" to "proud, grounded, wise in a way books do not always teach."

Two steps. Works at dinner. Works at parent-teacher night. Works in social settings. Works anywhere someone tries to press "teacher" into a lower chair.

In teacher-shaped situations, it looks like this.

"You just teach a few classes." → "Yes, that is what I do all day. Teach a few classes. Maybe four forty-five-minute periods. But whether any of those periods pulls a kid who has drifted for a year back into sitting up straight, whether I make one or two of those hours mean something, decides whether that kid keeps reading next year. I do not know how many of those I actually pull off. That is what I think about on the walk in each morning."

"Teaching is what people do when they cannot do anything else." → "Some teachers, sure. Others, the opposite. I know one who walked away from a marketing role at double the salary to teach English at a rural middle school. He told me: in marketing, he could 'influence' hundreds of thousands of ad views a year. In the school, he could 'stand next to' thirty-something people a year, one by one. He decided the second was harder work, not softer."

"Two months off in summer, must be nice." → "A lot of teachers spend that time prepping the fall, or coaching, or making home visits, or training. But even the real time off is part of the job, not a loophole in it. Here is what most people do not see: for every forty minutes in front of a class, there are two hours in the back. Kids see the forty. Parents see the forty. What the summer actually holds is the space to put two hundred kids back into your head, one by one, so each one is still themselves next year. That is where the job actually burns."

"Public? Not private?" → "Because half the kids I teach will not meet another adult in their lives who is willing to care about them without being paid. I might be one of the very few grown-ups who love them for free. That is not generosity. That is my post."

Look at the shape of each one. No raised voice, no eye-roll, no wounded face. You are not proving yourself. You are, in the air over that table, quietly redrawing the two words "teacher" and "teaching."

Why this is worth the small effort.

Clark presents this move as self-protection. When you place it in a teacher's hands, it protects more. Every teacher who swallows a joke at dinner is a small brick in the wall of why the profession bleeds people every year. The young people around you, the former students of yours, the child at the table who has thought about being a teacher, they hear those jokes too, and they see you swallow them. The signal they receive is: "even the people inside the profession are quietly embarrassed to defend it."

When you answer with one label upgrade, you do something else. You tell the table that this profession is not something you get to casually joke about. You did not insult anyone, you did not embarrass anyone, you did not ruin the meal. But you are no longer the teacher who quietly gives the profession's dignity away.

In this work, one teacher's silence may be harmless. An entire generation of teachers' silence is a disaster.

One level deeper.

Clark writes that what this move really protects is not the comeback. It is your right to define who you are. When someone uses their words to try to tell you what you do, if you do not respond, you have handed them that definition. If you respond in your words, you have taken it back.

For a teacher, that right is not only your dignity. It is the ground you stand on when you walk into a classroom the next morning, whether you believe what you are doing is worth a group of children spending some of the most valuable years of their lives sitting in front of you. The one sentence you leave at the table, seen from a distance, is also the one sentence you leave for yourself, for the next time you walk in.

One last thing.

Write, in advance, the one sentence you most want people around you to hear about your work. Keep it short. Do not make it sharp. Make it true.

The next time that dinner-table joke comes flying, take it out. Set it down. Say it evenly, no heat. Then keep eating.

You will notice the air at that table has shifted.

### Questions to sit with

1. The last time someone waved "teaching" away in front of you, what was the sentence? Did you stay quiet because you did not want to ruin the mood, or because you had not thought through what to say?

2. If I gave you three minutes right now to write the label upgrade you most want the people around you to hear about your profession, what would you write? Read it out loud to yourself three times. Do you believe it as you hear it?

3. Do you have a student who talks about wanting to teach someday? Do they know whether you defend the profession in the rooms they are not in? The next time they ask, how do you want to be able to answer?

This is our reading of Bill Gates's 2023 essay The risks of AI are real but manageable (Gates Notes). Read the original →

You have probably been in this staff-room conversation more than once.

Half your colleagues say: we have to ban it, otherwise these kids will never learn to write. The other half say: you cannot ban it, the tide is turning, we have to teach them to use it. Nobody convinces anybody. Everyone goes back to grading the same essays, pretending it will sort itself out.

In July 2023, Gates offered a third path in his essay. Not as theory. He named a real teacher who was already walking it.

Her name is Cherie Shields.

In January 2023 she wrote a piece for Education Week about how she uses ChatGPT in her English class. She did not ban it. She did not pretend it did not exist. She redesigned her entire class around the tool.

Gates quoted her: "Teachers will have to embrace AI technology as another tool students have access to. Just like we once taught students how to do a proper Google search, teachers should design clear lessons around how the ChatGPT bot can assist with essay writing. Acknowledging AI's existence and helping students work with it could revolutionize how we teach."

What did she actually do? Two things, worth stealing.

One: turn AI into a whetstone, not a ghost-writer.

Her assignment stopped being "write an 800-word argumentative essay." It became a four-step assignment.

Step 1: get ChatGPT to write an 800-word draft. Turn in the prompt you used, so I can see it. How you ordered the food is itself a skill.

Step 2: read the draft aloud and mark three sentences that "do not sound like something you would ever say." Write a short note on why.

Step 3: rewrite those three passages in your own words.

Step 4: submit the AI's draft, your final version, and a 200-word reflection: "what did I learn from the AI, what did I cut, why."

When she graded, she weighted steps 2, 3, and 4. Step 1 was raw material. The result: her students did not learn to write less. They learned to write better. Because every assignment now forced them to practice discern, critique, rewrite: three abilities AI will never build in you.

Two: turn AI's biggest weakness into your best teaching material.

In the same essay Gates writes: "educators can have AI generate articles and then work with their students to check the facts."

This sentence is gold. It takes AI's most painful problem, hallucinations, biases, made-up citations, and converts it directly into a lesson.

You can assign it like this. Have AI write a short piece on some historical event, some scientific topic, some figure. It does not matter which subject you teach. Then require students to pull every specific number, name, date, and "studies show" phrase, and verify each one. Getting the right ones right earns no points; catching the wrong ones earns real points.

Do this a few times and your students will: one, never blindly copy AI output again; two, learn a genuinely scarce skill for the AI era, fact-checking; three, develop muscle memory for what "trustworthy" even means.

Gates also brings up an older piece of history worth mentioning in your next staff meeting.

"It reminds me of the time when electronic calculators became widespread in the 1970s and 1980s. Some math teachers worried that students would stop learning how to do basic arithmetic, but others embraced the new technology and focused on the thinking skills behind the arithmetic."

You may remember, or your older colleagues may. The first teachers who worked with the calculator and the last teachers who refused it produced students separated by a generation. Your choice now, facing ChatGPT, is the same shape. And the same size.

One last thing, for you personally.

Cherie Shields did not do this because she was a technology expert. She did it because she decided: instead of waiting for the school to hand down a policy, she would try it herself. Then she wrote it down so others could see.

You do not have to wait for your school. You can, next assignment, pull one class period and try one version of "AI writes the draft, students take it apart." It does not have to work the first time. Tell your students you are experimenting. Tell them why. A teacher who is willing to learn alongside their students is more powerful than a teacher who pretends AI is not there.

### Questions to sit with

1. The last time you were grading and knew, halfway through, that a piece had been written by AI. What was your first reaction? Does that reaction match the kind of teacher you want to be for that student?

2. In the subject you teach, how is "source trust" being taught right now, citations, sources, fact-checking? If it was already weak before AI arrived, you have one chance now to move it to the center. Which assignment do you start with?

3. Does your school currently have any written policy on AI? If not, and if you decide not to wait, which class period next week do you plan to run the first open experiment with your students?

This is our take on Bill Gates's essay The turbulent AI era is here (Gates Notes, August 2026). Read the original →

First, name a thing you are already seeing in the room.

The same homework, this term, comes back to you in three different shapes:

One, written by the student. Rough, uneven, mistakes only that particular child makes.

Two, written by the AI. Smooth, complete, tone consistent, possibly more polished than anything you would write on a good day.

Three, the tricky one. Written with the AI. It is no longer the child's sentence, but you do not have proof.

This is the new normal you are teaching into.

Gates devotes a whole passage of his newest essay to this.

On one side he is genuinely optimistic: "AI can free teachers up to spend more time working with students one on one or in small groups and give them a clearer view of where the whole class is struggling." This is real. Grading, drafting exercises, first-pass feedback, answering routine questions, most of that is work AI can already carry, and it was work that used to eat your evenings and your weekends.

Then, without a pause, he writes the sentence you should hold on to:

"One preliminary survey suggested that heavier AI use was associated with less critical thinking. The effect was stronger for younger people." And then, one line later: "This would be the worst possible time for humans to lose their critical thinking skills."

Translation:

The same tool that gives you time back with your students can quietly remove from your students the very thing you are trying to build in them. The line between those two effects is yours to draw.

Gates hands you one key phrase: productive struggle.

The exact sentence: an AI used well "preserves what researchers call 'productive struggle', the cognitive work that builds understanding." Concretely: when the student meets a new idea for the first time, the AI can give substantive explanations and offer both questions and answers. When it is checking that they have understood, it holds the answer back and helps them arrive at it on their own.

That one sentence is worth taping to your desk. It resolves the whole "is AI good or bad in the classroom" argument in one move:

Scaffolding. AI struggles with the student and makes the struggle more effective. It passes them the next question, an intermediate step, a counterexample, an analogy.

Shortcut. AI struggles for the student. The cognitive work that was supposed to happen in the student, happens somewhere else and is delivered pre-packaged.

On the surface they look almost the same. The difference is whether the student ends the assignment stronger.

Two questions you can use in the next essay you grade.

Every time an assignment shows the fingerprints of AI, ask yourself only two things:

One: Did the AI struggle for them, or did it help them struggle more effectively?

Two: Without the AI, can they solve the next problem in this sequence on their own?

Two yeses, this was scaffolding. Two nos, this was a shortcut. Somewhere in between, you owe that student a five-minute conversation.

Two moves for your teaching system.

First: do not choose between "AI allowed" and "AI banned" as your whole policy. Let your policy read: "On this stage of the assignment AI is scaffolding, on that stage AI is forbidden." For example: for pre-reading questions, AI is welcome. For the first draft, no AI. For revision, AI can give question-shaped feedback but cannot rewrite for the student. On exams, obviously not. When students see that you have your own layered system, they know you are neither in panic nor asleep at the wheel.

Second: explicitly protect a set of assignments as "must-struggle." In math, the derivations they have to walk through themselves. In language, the pieces where they have to name their own title, find their own example, arrive at their own thesis. In history, the essays where they have to judge cause and consequence for themselves. Tell them, out loud: "AI would ruin what this assignment is for. Please write this one yourself." Students hear this better than you expect them to.

One thing for you personally.

The whole thrust of Gates's essay carries one signal for you: this is not the era where teachers matter less. This is the era where teachers matter more than they have in a long time. The routine, the grading, the drafting, the answering, can be handed off. But the judgment call, what to hand off, and what must stay in the student's own hands, that is not a workflow, it is a decision. The person who makes that decision is you.

Half of what you did last semester will be quietly optimized away by AI in the next two or three. The other half, the half where you decide what this assignment is for, and what your student is supposed to grow, that is the part the era cannot take. Please put more of your strength into that half.

### Questions to sit with

1. Recall a moment when a student struggled toward the answer instead of being told it. What did they gain that a shortcut would have removed?

2. In the subject you teach, where is "productive struggle" most fragile, the exact place a well-meaning student would most easily let AI do the wrong work for them?

3. What is one small policy in your classroom you could add this month that keeps AI as scaffolding, not as shortcut?

What made us stop, in the Harvard reporting, wasn't the 20% cap. It wasn't the student protests. It wasn't Wall Street's algorithm on GPA. It was a literature professor, Karen Thornber, redesigning her general-education course on literature and mental health, quietly saying:

"The course is also becoming more demanding. We didn't feel empowered to do that. We're under tremendous pressure from students to give A's."

If you teach, that sentence probably knocked the wind out of you. You may no longer remember which semester it started, the semester you quietly dropped the passing bar a notch, swapped out the harder reading, deleted the deep prompt from the final. You have probably been telling yourself one of the reasonable versions: I'm being kind, they're exhausted. Or: if I grade honestly they won't grow anyway, they'll just complain. Or: end-of-term reviews are what my tenure runs on.

All of those are true. You aren't faking it. What you're up against is a real, structural pressure. That the Harvard faculty admitted this out loud is a big deal, because it means even at the school with more applicants than it can turn away, teachers have been giving ground to give-more-A's. You aren't the exception. You are the norm of your generation of teachers.

But let us look with you at another line in the same reporting. A 2025 internal Harvard report found that students are picking classes by perceived easiness, avoiding hard prompts, not attending class, not doing the assigned reading, expecting a level of "flexibility" they've grown used to since Covid. Which means, every requirement you lowered, they felt. They didn't become happier. They became hollower.

There is something we especially want to say to you inside that: giving an A isn't kindness. Making the course easier isn't care. Real care is trusting the student to hold something hard and staying next to them until they do. When you lowered the bar, what you were quietly telling them was I don't think you can do it. That isn't what you meant. But that is what their gut received.

So here is something you may not have been given in a long time: permission to be demanding again. Not colder. Not harsher. Not more punishing. Permission to give an honest grade at the end of the term. Permission to put back into the syllabus the book that once shook you. Permission, at the end of one class, to ask the harder question, the one you originally wrote in and later deleted because you were afraid of the silence.

You will lose some rating points. You will get some complaints. One or two students will be angry at you, fewer than you expect. And then some of them, years later, will write to you on their own to tell you that you were one of the few adults on their whole path who didn't back down. Most people accumulate maybe three such teachers in a lifetime. You have a chance to be their third.

This coming semester, you don't have to become a different person. You just have to, right next to the "easier" version you were about to run, prepare the version that is as demanding as it should be, and pick the second one. Once. See what happens.

Every teacher makes hundreds of "where does this go next" judgments in a day, when to ask, when to state, when to just wait.

Child development specialist Siggie Cohen's observation about parents holds true in a classroom too: dressing a required action as "can you?" invites unnecessary pushback. "Can we quiet down?" "How many times do I have to remind you?", those aren't questions; they're commands we didn't state.

Her principle: Say What You Mean. Before you speak, ask: what am I trying to communicate? Then say it plainly.

Starting a lesson: not "Are we ready?", "We're starting. Please turn to page thirty-four."

After a fight: not "Why did you hit him?", "In our class, we don't hit. You're angry right now, I know, let's find another way to tell him."

Roughhousing in the hall: not "Are you at it again?", "The hallway is a quiet place. Please come back and try that walk again."

Then when do you ask?, When it's actually worth asking. "What did that story remind you of?" "For that problem, what's the first step you thought of?" "Is there an answer you're not sure of yourself?", that is where a question does its real work.

Classroom order isn't built on authority. It's built on clarity. A child who knows where the edges are has room left over to be curious, to try, to be wrong.

When you take your energy back from nagging, interrogating, and reprimanding, you can give it to teaching, watching, being with them. That, in itself, is the tenderness the students feel.

Reem Raouda, a parenting coach who has worked with over 200 children, recently wrote a piece for parents naming seven signs a child is building emotional intelligence. We want to put those same seven things back into the classroom, because the side of a child you see is the side most parents cannot.

1. They can name what they're feeling. Watch for the student who can actually say "I'm tired today" or "something's off." Putting a feeling into words is already taking responsibility for it. When a child blows up, throws something, slams a door, they aren't "difficult." They simply don't have the words yet. Teaching the words beats demanding silence.

2. They come to you when something is wrong. The child who slips in at break to tell you "I had a fight" or "I don't have friends" has decided you are a safe harbor. The opposite is the warning: the child who never comes to you may be the one who needs you most.

3. They can be disappointed without falling apart. Lost the match, got the lower grade, missed the part, they may cry, they may go quiet, but they recover. Not because they don't care, but because they aren't defined by a single bad day. Your job is to give them time, not to rush them to an answer.

4. They notice how other people feel. "Teacher, are you OK today?", that student is rare. They are looking at the world with the eyes of a future Bole. Tell them so. Let them know that noticing is a gift, not nosiness.

5. They can actually apologize. Not the forced "sorry" to avoid trouble, but the moment they see they've hurt someone and want to repair it. That kind of self-awareness is rare even in adults. When you see it, remember the child, they are not just "good"; they are already beginning to be their own teacher.

6. They can ask for what they need. "I need a minute alone." "Can I sit next to you?" A child who can ask is a child who has been answered. If they can't ask, ask back: "What do you need right now?" That sentence is its own lesson.

7. They don't feel like they have to perform around you. This one is the easiest to miss. The emotionally intelligent child is not constantly managing the look on your face, not suppressing themselves to stay in your good graces. They can be a little tired in front of you. A little funny. A little wrong. That is the highest form of trust.

Emotional intelligence is not a subject. Children learn it from the adults around them, and you may be the most important adult in the building. Their parents may be too busy. The world outside may be sanding them down. But you saw them. That alone changes what is possible for them.