Introduction
As I put this article together, we are all witnessing a strong backlash against artificial intelligence in all of its forms. Stories are coming out every day about the dangers of this new technology and its impact on almost every part of our lives. Articles are written in newspapers and journals about environmental impacts as well as the cost of energy and its depletion. Then there are the revelations of chatbots infiltrating people’s lives, deceiving them and causing all sorts of psychological stress. Then we have educators at all levels bemoaning the fact that their students are no longer thinking and relying on LLM s to do their work for them. Added to this is the obvious fact that it appears to be growth that has few if any real guardrails. These are scary times and unfortunately unlike those 90 horror movies, this one does not have an ending. This essay touches on many of these concerns and the action we feel must be taken.
The next and equally important part of this essay is that it is an exercise in co-authoring, not between two old grandfathers like Randy and me, but with an artificial intelligence. In the second part of this essay, we show how this essay is in and of itself a model for the proper and effective of Generative Artificial Intelligence. We urge the reader to consider it and to ponder what it really means to be creative and original in this era.
Prologue: Why We Wrote This
Before you read another word, there is something you should know about the two people who wrote this article.
We are not software engineers.
We are not venture capitalists.
We are not employees of OpenAI, Google or Microsoft.
We are not trying to sell you artificial intelligence.
We are two grandfathers.
Between us are grandchildren who still believe tomorrow will be better than today simply because the adults in their lives are doing everything possible to make it so.
One is just beginning to ask endless questions about how the world works. Another is discovering books. One is fascinated by building things. Another still believes that every problem has a solution if you simply keep asking "Why?"
We watch them learn, and like every grandparent before us, we quietly imagine the people they will one day become.
Will one become a physician?
An engineer?
A teacher?
A scientist?
An artist?
Perhaps none of those.
Perhaps something that doesn't even exist today.
That last possibility is precisely why we wrote this article.
Between us, we have spent more than a century working in education.
We have taught in classrooms, led educational organizations, advised governments, worked with universities, spoken to educators around the world and dedicated our professional lives to one simple belief:
Education is society's greatest promise to its children.
Not because it teaches them about the past.
Because it prepares them for the future.
For most of our careers, those two things were largely the same.
Today, they are not.
For the first time in modern history, the world outside our schools is changing faster than the world inside them.
That should concern every parent.
It certainly concerns us.
We have watched artificial intelligence move from a curiosity to a companion, from a tool to a collaborator and now toward becoming something far more profound. We have also watched the debate surrounding it become increasingly emotional, increasingly polarized and, at times, increasingly detached from reality.
Some people insist AI will save the world.
Others insist it will destroy it.
We believe both sides are asking the wrong question.
The question has never been whether artificial intelligence is good or bad.
Fire was neither.
Electricity was neither.
The Internet was neither.
The better question—the one that has echoed throughout every major technological revolution—is whether society will have the wisdom to prepare the next generation for the world those technologies create.
That is the question that keeps us awake.
Not because we fear artificial intelligence.
Because we fear complacency.
We fear institutions that mistake stability for progress.
We fear systems that prepare children brilliantly for a world that quietly disappeared while adults were still debating whether change was necessary.
Most of all, we fear looking into the eyes of our grandchildren twenty years from now and realizing we saw the future approaching...
...and chose to argue about it instead of preparing them for it.
This article is not a defence of artificial intelligence.
Nor is it an attack on education.
It is a plea.
A plea for consistency.
A plea for courage.
A plea for the humility to recognize that every generation inherits a moment when the future arrives before it feels ready.
This is ours.
If, at times, these pages make you uncomfortable, that is intentional.
They made us uncomfortable too.
Because every argument we make here forced us to question our own assumptions before we ever questioned anyone else's.
If you finish this article agreeing with everything we have written, we will probably have failed.
Agreement changes very little.
If, however, you finish it asking yourself questions you hadn't considered before...
If you find yourself looking at your children or grandchildren just a little differently...
If tomorrow morning you begin wondering whether school is preparing them for the world they will actually inherit rather than the one we remember...
Then perhaps this article will have done what education itself has always done at its very best.
Not provide all the answers.
But inspire better questions.
And in an age increasingly defined by artificial intelligence, perhaps asking better questions is the most profoundly human skill we have left.
So, before we begin, we ask only one thing of you.
Read with an open mind.
Challenge us.
Question every argument we make.
Demand evidence.
Demand consistency.
Above all, ask yourself not what kind of future you hope your grandchildren will inherit...
...but what kind of education will prepare them to build it.
Because in the end, this article was never really about artificial intelligence.
It was always about the children who will one day inherit the consequences of the decisions we make today.
1. The AI Backlash Says More About Us Than It Does About AI
Artificial intelligence has become the latest villain in society's never-ending search for something to fear.
Depending on which headline flashes across your screen today, AI is destroying the environment, stealing jobs, making students intellectually lazy, spreading misinformation, violating copyright, threatening democracy and, if Hollywood has its way, quietly plotting humanity's extinction somewhere between software updates.
Some of those concerns are legitimate.
Some are exaggerated.
All deserve to be examined.
What deserves even closer examination, however, is us.
The real story may not be artificial intelligence at all. It may be how remarkably predictable human beings become whenever the world changes faster than they do.
We like to believe we are rational creatures. We imagine ourselves carefully weighing evidence before arriving at thoughtful conclusions. It is a comforting belief.
It is also one that decades of psychological research have repeatedly challenged.
Researchers such as Paul Slovic have demonstrated that our perception of risk is influenced as much by emotion as by evidence. Daniel Kahneman showed that our intuitive, fast-thinking minds often reach conclusions long before our slower, analytical minds begin evaluating the facts. Jonathan Haidt argues that reasoning frequently behaves less like an impartial judge seeking truth than a lawyer defending a client already presumed innocent.
Translated into everyday language, the process looks something like this.
We react.
We decide.
Then we become remarkably skilled at explaining why we were right all along.
Artificial intelligence has become the perfect target for this very human tendency.
It is new.
It is complicated.
Few people truly understand how it works.
Every week seems to bring another dramatic headline warning of catastrophe, while Hollywood has spent decades teaching us that intelligent machines inevitably develop an unhealthy interest in eliminating their creators.
Under those circumstances, fear is hardly surprising.
What is surprising is how quickly many discussions about AI stop being discussions at all.
They become declarations.
"AI will destroy education."
"AI will eliminate thinking."
"AI will replace everyone."
The certainty is fascinating.
Not because certainty is evidence.
Because certainty is often emotion wearing the disguise of evidence.
None of this means the concerns should be dismissed.
Artificial intelligence consumes enormous amounts of computing power. Training frontier models requires significant energy. Hallucinations remain a serious limitation. Bias exists. Privacy matters. Intellectual property matters. The concentration of technological power in the hands of relatively few companies deserves careful scrutiny. Researchers, governments and international organizations have all raised legitimate questions about transparency, accountability and governance.
Those concerns are real.
Ignoring them would be irresponsible.
But acknowledging genuine risks is very different from concluding that the technology itself should therefore be feared or resisted.
History suggests those are two very different conversations.
In fact, history suggests we have had this conversation before.
Many times.
The names change.
The technology changes.
The arguments barely do.
That observation leads to a question that may make some readers uncomfortable.
If your objection to artificial intelligence is that it consumes too much energy, changes the nature of work, encourages dependency or transforms society, then ask yourself a simple question.
IF THAT IS YOUR POSITION, THEN BE CONSISTENT.
Apply exactly the same standard to every technology you already depend upon.
Not just the new one.
All of them.
That challenge is not intended to silence criticism.
Quite the opposite.
It is an invitation to make our reasoning more rigorous.
If artificial intelligence truly deserves to be treated differently from every transformative technology that preceded it, then make the case.
Show the evidence.
Explain why.
But if the arguments against AI sound remarkably similar to the arguments once made against the printing press, the railway, the automobile, the calculator, the personal computer and the Internet, then perhaps the technology is not the first thing that deserves closer examination.
Perhaps we should begin by examining ourselves.
Because the question that matters most is not whether artificial intelligence will change the world.
The question is whether we are reacting to the evidence before us—or to the uncertainty we feel within us.
2. History Isn't Repeating Itself. We Are.
If there is one lesson history teaches with remarkable consistency, it is this:
Human beings possess an extraordinary ability to believe that this time is different.
Every generation seems convinced that it is witnessing the first technological revolution capable of fundamentally changing society for the worse. Every generation is equally convinced that the fears of previous generations were irrational while its own concerns are firmly grounded in evidence.
History is less charitable.
When Johannes Gutenberg introduced the printing press in the fifteenth century, many feared the unrestricted spread of dangerous ideas. For centuries, knowledge had been controlled by a privileged few. Suddenly, ordinary people could read, question and think for themselves. To many, this wasn't progress. It was chaos.
They were partly right.
The printing press did spread dangerous ideas.
It also spread science, literacy, democracy and the Renaissance.
The railway was no different.
When trains first appeared, respected physicians warned that travelling at thirty miles an hour would damage the human body. Some believed the human lungs were never designed to withstand such unimaginable speed. Others worried that the countryside flashing past the windows would cause mental illness.
It sounds ridiculous today.
It didn't then.
The telephone was accused of destroying real conversation.
Radio was going to destroy reading.
Television would destroy families.
Calculators would destroy mathematics.
Computers would destroy employment.
The Internet would destroy civilization.
To be fair, the Internet has certainly done a respectable job of making conspiracy theories easier to find.
The point, however, isn't that the critics were foolish.
The point is that they were human.
Every one of those technologies created problems.
Automobiles polluted cities.
Television changed childhood.
Social media amplified misinformation.
Smartphones quietly rewired our attention spans.
Technology has never arrived carrying only gifts.
It has always arrived carrying consequences.
The remarkable part is not that problems emerged.
The remarkable part is how consistently society adapted.
We improved.
We regulated.
We educated.
We learned.
And eventually, we reached a point where the technology became so woven into everyday life that we forgot it had ever been controversial.
That is the part history rarely tells us.
It remembers the invention.
It forgets the panic.
Which brings us back to artificial intelligence.
Today we hear that AI consumes too much electricity.
It may.
IF THAT IS YOUR POSITION, THEN BE CONSISTENT.
Are you equally concerned about cryptocurrency mining?
About the billions of videos streamed every day?
About the explosion of cloud computing?
About the massive energy demands of data centres that already power virtually every digital service you use?
Or has electricity suddenly become a moral issue only because the technology now happens to be called artificial intelligence?
We hear that AI will make people intellectually lazy.
Perhaps.
IF THAT IS YOUR POSITION, THEN BE CONSISTENT.
Throw away your GPS.
Stop using Google.
Disable spellcheck.
Turn off predictive text.
Drive with paper maps.
Memorize every phone number.
After all, each of those technologies replaced cognitive work that people once performed for themselves.
If they didn't diminish human intelligence, why is AI uniquely incapable of doing anything else?
Then comes the argument that AI will eliminate jobs.
History suggests that one deserves serious consideration.
Technology has always displaced work.
The Industrial Revolution replaced artisans.
The tractor transformed agriculture.
The ATM reduced the need for bank tellers.
Spreadsheet software eliminated countless hours of bookkeeping.
The Internet redefined retail.
Artificial intelligence will almost certainly eliminate some jobs.
It will almost certainly create others.
The real issue has never been whether technology changes employment.
It always has.
The issue is whether societies prepare people for the jobs that follow.
History offers a rather uncomfortable observation.
The greatest mistakes were rarely made by inventors.
They were made by institutions that assumed tomorrow would look remarkably like yesterday.
Kodak invented the digital camera but failed to embrace it.
Blockbuster laughed at streaming.
Nokia underestimated smartphones.
None disappeared because they lacked intelligent people.
They disappeared because they recognized change too late.
That should concern us far more than artificial intelligence itself.
Because every technological revolution eventually reaches a tipping point where the debate quietly changes.
At first, people ask,
"Should we use this technology?"
Later, they ask,
"How did we ever live without it?"
The transition happens so gradually that few people notice it until it is already behind them.
Perhaps that is where we are with artificial intelligence.
Many people still believe we are debating whether AI belongs in society.
Businesses have already answered.
Hospitals have already answered.
Scientists have already answered.
Engineers have already answered.
Financial institutions have already answered.
The only significant debate seems to be taking place in the one institution responsible for preparing young people for the future.
Education.
History isn't the story of technologies repeating themselves. It is the story of human beings repeating themselves.
The uncomfortable question is whether, once again, we are spending so much time arguing about yesterday's fears that we fail to notice tomorrow quietly arriving without us.
3. The Debate Is Already Outdated
A curious thing is happening.
Much of society is still arguing about ChatGPT.
The technology industry has already moved on.
That should concern you.
The public debate is largely about chatbots that answer questions, summarize reports and occasionally write surprisingly good emails.
The companies investing hundreds of billions of dollars into artificial intelligence aren't building better chatbots.
They're building digital colleagues.
There is an enormous difference.
Imagine it's seven-thirty on a Monday morning.
Dr. Patel walks into her clinic carrying the same travel mug she's carried for years. Before she reaches her office, the waiting room is already full. A child with a persistent cough. An elderly gentleman whose blood pressure has become increasingly erratic. A woman whose fatigue has baffled three specialists. Twelve minutes have been allocated for each appointment.
Twelve minutes.
She hangs her coat, opens her laptop and pauses.
Every patient's file has already been reviewed.
Every blood test compared.
Every prescription checked against every other prescription.
A newly published clinical study from Sweden, released only hours earlier, has been flagged because it may change the treatment plan for one of her patients. A potentially dangerous interaction between two medications has been highlighted. One patient's symptoms, insignificant on their own, have been connected to thousands of similar cases around the world.
The summary takes less than two minutes to read.
What would have taken an experienced physician hours—if those hours even existed—has already been done.
The AI didn't diagnose the patient.
It didn't replace the doctor.
It simply gave the doctor something that modern medicine has been quietly stealing for years.
Time.
Now leave the clinic.
It's almost eleven o'clock at night.
The lights are still on in a suburban kitchen.
Mrs. Tremblay, a Grade 5 teacher, sits surrounded by stacks of notebooks. Tomorrow's science lesson isn't finished. Twenty-eight math assignments still need correcting. Three parents have emailed. One student has quietly stopped participating in class. Another is brilliant but hopelessly bored. Report cards are due on Friday.
She rubs her eyes and reaches for another notebook.
Now imagine a different evening.
The routine marking has already been completed.
Patterns in student errors have been identified.
The struggling student has been quietly flagged before disappearing completely.
Tomorrow's lesson has been adapted for three different learning levels.
Draft responses to the parents are waiting for review.
The report cards have already been assembled, leaving only the professional judgment that no machine should ever replace.
The AI hasn't become the teacher.
It has removed the work that prevented the teacher from teaching.
Now picture someone else.
A small business owner.
Sunday afternoon.
The office is silent except for the hum of fluorescent lights.
Outside, families are at soccer games, barbecues and birthday parties.
Inside, she's trying to reconcile invoices, answer customer emails, prepare payroll and determine whether there will be enough cash flow to make it through another month.
She didn't open her business because she dreamed of spending weekends buried in paperwork.
She opened it because she loved building something.
An AI agent doesn't replace her.
It gives her Sunday afternoon back.
Notice something.
None of these stories are really about artificial intelligence.
They're about people.
They're about reclaiming time.
They're about allowing human beings to spend less of their lives completing repetitive tasks and more of their lives exercising judgment, creativity, empathy and imagination.
That is what makes the current public debate feel strangely disconnected from reality.
Most people still imagine AI as an assistant that waits politely for instructions.
Agentic AI doesn't wait.
It works.
Give today's chatbot a question and it gives you an answer.
Give an AI agent an objective and something entirely different begins to happen.
Suppose you say:
"I'd like to take my family to Italy next spring. We have a budget of $8,000, we'd like to spend a few days in Tuscany, we'd prefer boutique hotels, and my daughter has a peanut allergy."
A chatbot gives you suggestions.
An AI agent quietly begins comparing hundreds of flights while you sleep. It monitors prices, books tickets when they fall within your budget, reserves hotels matching your preferences, identifies restaurants that accommodate food allergies, builds an itinerary around local events, places everything in your calendar, alerts you when a rail strike threatens your connections and rearranges your plans before you've even finished breakfast.
You didn't ask twenty questions.
You gave it one objective.
Now multiply that capability.
By medicine.
By engineering.
By scientific research.
By law.
By manufacturing.
By logistics.
By finance.
By government.
By education.
That is why the world's largest technology companies are investing extraordinary sums in agentic AI. They are not trying to build better search engines. They are trying to redefine knowledge work itself. Research from organizations including McKinsey, the Stanford Institute for Human-Centered AI and major industry analyses increasingly points toward AI evolving from a tool that assists individuals to one that coordinates increasingly complex workflows across entire organizations.
The implications are difficult to overstate.
For most of human history, productivity depended upon how much work one person could personally accomplish.
Increasingly, productivity will depend upon how effectively one person can direct work they never actually perform themselves.
That changes what it means to be productive.
It changes what it means to be employable.
It changes what it means to be educated.
Yet while all of this unfolds at breathtaking speed, much of society continues arguing about whether students should be allowed to use ChatGPT to write essays.
It is rather like debating the educational value of the horse at the very moment someone is inventing the automobile.
One discussion isn't necessarily wrong.
It has simply become irrelevant.
Perhaps that is the most unsettling realization of all.
We are holding passionate debates about the first generation of AI while the second generation is already beginning to reshape the world around us.
Committee meetings continue.
Curriculum reviews continue.
Policy papers continue.
Meanwhile, artificial intelligence becomes more capable every month.
History has never punished societies for asking difficult questions.
It has punished them for asking yesterday's questions long after the world had moved on.
We are not standing at the beginning of this revolution. We are standing in the middle of it while pretending it hasn't started.
4. The First Parent Who Asks the Right Question Changes Everything
It didn't happen in a government office.
It didn't begin with a minister of education.
It wasn't announced in a policy document, debated in parliament or unveiled at an international conference.
Like so many revolutions, it began at a kitchen table.
It was nearly nine o'clock.
The dishes had been cleared away, the dishwasher hummed quietly in the background, and fourteen-year-old Ethan sat staring at a history assignment that should have taken thirty minutes but was entering its third hour.
His mother, Lisa, glanced over from the living room.
"Everything okay?"
"I guess."
Every parent recognizes that answer.
"What's wrong?"
"I've got to write an essay on climate change."
"That doesn't sound too bad."
"It isn't."
He hesitated.
"I'm just not allowed to use AI."
She looked up.
"Not even to help you research it?"
"No."
"To organize your ideas?"
"No."
"To explain something you don't understand?"
"No."
"The teacher said she wants to know what I know."
Lisa nodded, but something about the answer unsettled her.
Just then her husband walked in, dropped his briefcase beside the kitchen counter and smiled.
"Long day?"
"Actually, no."
"You've been working on that board presentation all week."
"I was."
"Were?"
He opened his laptop.
"My AI agent finished the first draft this afternoon."
She frowned.
"It finished it?"
"It analyzed six months of sales data, compared our performance against competitors, generated the charts, summarized the market research and even suggested the questions the board is likely to ask tomorrow."
He shrugged.
"My job wasn't writing the report."
"My job was deciding whether the report was right."
Ethan looked from one parent to the other.
"So Dad gets to use AI..."
"...to do his job," his father replied.
"But I can't use it..."
"...to learn yours."
No one spoke.
The silence lasted only a few seconds.
It felt much longer.
Later that evening, unable to shake the conversation, Lisa opened her laptop.
She wasn't looking for confirmation.
She was looking for reassurance.
Instead, she found UNESCO.
One report after another made essentially the same point. Artificial intelligence should not simply be added to existing curricula; it demands that education rethink what students need to learn in a world where intelligent systems are becoming part of everyday life. The question is no longer whether students will encounter AI. The question is whether schools will prepare them to use it wisely, ethically and effectively.
She opened another report.
The OECD.
Different authors.
Different countries.
The same conclusion.
Routine knowledge is becoming less valuable. Human judgment, creativity, problem solving, collaboration and the ability to work effectively alongside intelligent technologies are becoming more important. The workplace is changing far more quickly than curriculum.
She leaned back in her chair.
Then she laughed.
Not because it was amusing.
Because for the first time she saw the contradiction so clearly it was almost absurd.
Tomorrow morning she would visit a physician who depended upon AI to identify patterns hidden in thousands of medical records.
The pharmacist filling her prescription would use AI to detect dangerous drug interactions.
Her financial advisor relied on AI to analyze markets.
The architect designing the community recreation centre used AI-assisted modelling.
The mechanic servicing her car depended on AI diagnostics.
The grocery store somehow knew she was almost out of coffee before she did.
Every profession touching her family's life had quietly accepted artificial intelligence as another tool of the trade.
Every profession except one.
The next morning, on the drive to school, she couldn't let it go.
"Ethan."
"Yeah?"
"If your teacher had allowed AI yesterday, would you have learned less?"
He looked genuinely puzzled.
"No."
"What would you have done?"
"I would've asked it to explain the stuff I didn't understand."
"And then?"
"I would've argued with it."
She glanced over.
"Argued?"
"Sure."
"I ask it questions until I think it's wrong."
"And?"
"And then it explains why."
He smiled.
"Sometimes I'm right."
"Usually I'm not."
She couldn't help smiling.
"Your teacher knows you do that?"
He shook his head.
"I don't think she'd like it."
"Why?"
"Because she thinks AI does the thinking for you."
He stared out the window for a moment.
"It doesn't."
"It just makes me think about different things."
That sentence stayed with her long after he disappeared through the school doors.
It was more insightful than many of the policy papers she had spent the previous evening reading.
Later that week, she found another study. This one examined teacher preparation. Across much of North America, researchers were reaching an increasingly uncomfortable conclusion. While AI was transforming medicine, engineering, finance, law and business, many teacher education programs still devoted little systematic attention to artificial intelligence, its pedagogical implications or how assessment itself might need to change. The very people expected to prepare children for an AI-rich future were often receiving little preparation themselves.
For generations, schools could afford to move slowly because society moved slowly.
A curriculum could remain largely unchanged for ten years without causing much harm.
Not anymore.
Artificial intelligence doesn't wait for curriculum committees.
It doesn't pause while ministries draft new frameworks.
It doesn't delay development until teacher education catches up.
It simply becomes more capable.
Every month.
Every week.
Sometimes every day.
That changes everything.
On Friday evening, Lisa attended Ethan's basketball game.
As parents filled the bleachers, conversations drifted from vacations to mortgages, from rising grocery prices to summer plans.
Then someone mentioned AI.
"My company just rolled out a new AI system," one father said. "It probably saves me two hours a day."
A woman sitting beside him nodded.
"Our hospital is using it now."
"So is my law firm," another parent added.
"Same with us," someone else said. "We'd be lost without it."
Lisa listened quietly.
Finally she asked a question.
"So...if every one of us depends on AI to do our jobs better..."
Heads turned.
"...why are our children being taught that using it to become better learners is somehow unacceptable?"
Nobody answered.
Not because the question was difficult.
Because everyone suddenly realized it was.
Driving home that night, Ethan asked from the back seat,
"Mom?"
"Yes?"
"Do you think school is getting me ready for the future?"
She started to answer.
She couldn't.
Because she honestly wasn't sure anymore.
And that was the moment everything changed.
Not because artificial intelligence had replaced teachers.
Not because schools had suddenly become obsolete.
But because, for the first time, a parent had begun to question whether the institution she trusted most with her son's future was preparing him for the world he would actually inherit.
History suggests that revolutions rarely begin when governments change policy.
They begin when ordinary people begin asking extraordinary questions.
Schools will not lose public confidence because artificial intelligence replaces teachers. They will lose it the day enough parents conclude that schools are preparing children for yesterday while quietly delaying their future.
5. If You Consider Yourself Intelligent, Then Act Like One
The coffee was terrible.
It always was.
The folding chairs were arranged in uneven circles, someone had brought supermarket cookies that nobody intended to eat, and twenty-five parents sat waiting for the Home and School meeting to begin. Like most meetings, the real discussion started long before anyone looked at the agenda.
It didn't take long.
"I hear the school board is looking at an AI policy," one father said, stirring powdered milk into his coffee.
"I certainly hope they ban it," replied another.
Several parents nodded.
"Our kids need to learn to think for themselves."
"Exactly."
"They're already on their phones too much."
"Now they'll have a machine doing the thinking for them."
A woman sitting quietly near the back finally spoke.
"So AI is a bad thing?"
The answers came quickly.
"It destroys critical thinking."
"It encourages cheating."
"It makes students lazy."
"It'll ruin education."
Then came the voice that seemed determined to end the discussion before it had really begun.
"I'll tell you exactly what it'll do."
Everyone turned.
"It'll destroy an entire generation."
He folded his arms.
"And any teacher who allows students to use AI shouldn't be teaching."
Several heads nodded.
The verdict had been reached in less than a minute.
The woman smiled politely.
"May I ask a few questions?"
Nobody objected.
"What do you do for a living?"
The first father smiled.
"I'm a structural engineer."
"Do you use AI?"
"Every day."
"What for?"
He laughed.
"What don't I use it for? It reviews structural calculations, checks building code compliance, identifies potential design weaknesses and runs simulations that would've taken our team days to complete."
She nodded.
"Anyone else?"
"I'm a lawyer."
"AI?"
"I'd be crazy not to use it."
"What does it do?"
"It searches decades of case law in minutes. It drafts research. It finds precedents I might otherwise miss."
A woman near the window raised her hand.
"I'm a nurse."
"Our hospital now uses AI to identify patients who may be at risk of deteriorating before it's obvious to the staff."
Someone else joined in.
"I'm an accountant."
"It reviews transactions before I do."
Another voice.
"I own a construction company."
"Our scheduling system predicts material shortages before they happen."
Another.
"I'm an architect."
"It helps evaluate hundreds of design options overnight."
The woman looked around the room.
"So..."
She paused just long enough to become uncomfortable.
"Every one of you uses artificial intelligence to become better at your profession."
Heads nodded.
"But you don't want children using it to become better learners."
Silence.
One father shifted uneasily.
"That's different."
She smiled.
"I've heard that phrase several times tonight."
"It is different."
"Why?"
He opened his mouth.
Then closed it again.
Before he could answer, the man at the back leaned forward.
"Because children need to learn."
"They do."
"They need to struggle."
"They do."
"They need to think for themselves."
"They absolutely do."
"So there."
The room was quiet.
She waited.
Then asked gently,
"May I ask another question?"
He shrugged.
"If your daughter was diagnosed with a serious illness tomorrow, would you want her physician using every available technology to improve the accuracy of that diagnosis?"
"Of course."
"If your home developed a structural problem, would you expect your engineer to ignore AI-assisted safety analysis?"
"No."
"If your lawyer could uncover a critical legal precedent in thirty seconds instead of thirty hours..."
"I get your point."
"No," she replied softly.
"I don't think you do."
Again, silence.
She wasn't trying to win an argument.
She was trying to understand one.
"Why is it," she asked, "that we celebrate professionals for using AI to make better decisions, yet condemn students for using AI to become better thinkers?"
The man folded his arms even tighter.
"I don't care what anybody says."
His voice had become noticeably sharper.
"Nothing will convince me that AI belongs in schools."
Nobody challenged him.
Nobody laughed.
Nobody rolled their eyes.
Instead, something much more interesting happened.
The room quietly moved on.
The engineer spoke again.
"I have to admit..."
He looked around at the others.
"I've never actually thought about the contradiction."
The lawyer nodded.
"Neither have I."
The nurse leaned back.
"When we introduced AI at the hospital, there was resistance too."
She smiled.
"Now I honestly wouldn't want to work without it."
The accountant added,
"I came in tonight convinced AI had no place in education."
He shrugged.
"I'm not nearly as certain now."
The man at the back remained silent.
Not because he had been persuaded.
Because he had stopped listening.
The discussion had become about evidence.
He was still arguing from emotion.
Behavioural scientists have spent decades studying moments exactly like this. Daniel Kahneman demonstrated how our intuitive minds frequently reach conclusions before our analytical minds have begun examining the evidence. Jonathan Haidt argues that reasoning often resembles a lawyer defending an existing position rather than a judge searching for the truth. Paul Slovic's research on risk perception shows that once strong emotions become attached to a technology, contradictory evidence often has surprisingly little influence.
Perhaps the most unsettling conclusion from all of this is that intelligence offers surprisingly little protection.
Sometimes intelligence simply makes us better at defending conclusions we reached emotionally.
The woman reached for her coffee.
It was still terrible.
"May I ask one final question?"
The room nodded.
"If you consider yourself an intelligent person..."
She looked around the circle.
"...and I suspect every one of us does..."
A few parents smiled.
"...then what does intelligence actually require?"
Nobody answered.
She continued.
"Is intelligence simply remembering information?"
"Earning degrees?"
"Winning arguments?"
"Being right?"
She shook her head.
"I don't think so."
"I think intelligence is something much more demanding."
"It is recognizing patterns."
"It is changing your mind when the evidence changes."
"It is abandoning comfortable beliefs when they no longer explain the world."
"It is having the humility to admit that perhaps yesterday's answers are no longer sufficient for tomorrow's questions."
She paused.
"So here's my challenge."
She looked directly at the group.
"IF YOU CONSIDER YOURSELF AN INTELLIGENT PERSON, THEN ACT LIKE ONE."
"Don't tell me artificial intelligence is different."
"Prove it."
"Show me why this technological revolution breaks the pattern established by the printing press, electricity, the automobile, aviation, television, computers and the Internet."
"If your argument is that AI consumes too much energy..."
She smiled.
"IF THAT IS YOUR POSITION, THEN BE CONSISTENT."
"If your argument is that AI makes people intellectually dependent..."
Again she smiled.
"IF THAT IS YOUR POSITION, THEN BE CONSISTENT."
"If your argument is that AI will change employment..."
She didn't need to finish.
Everyone already knew how the sentence ended.
The room was silent.
Not because everyone agreed.
Far from it.
The man at the back still stared at the table, his arms folded, his expression unchanged.
He had come with a conclusion.
He was leaving with the same one.
Everyone else, however, was leaving with something far more valuable than certainty.
Questions.
Good questions have a way of lingering.
They follow us home.
They interrupt our certainty.
They quietly dismantle opinions we once thought unshakable.
Perhaps that is the highest form of intelligence.
Not possessing all the answers.
But recognizing the moment when the questions have become better than the answers.
As parents slowly gathered their coats and headed toward the parking lot, no one spoke about the agenda.
They spoke about the conversation.
And for several of them, that conversation would continue long after the coffee had been forgotten.
Perhaps intelligence isn't measured by what we already know. Perhaps it is measured by our willingness to recognize patterns, question ourselves, and adapt before history once again proves us wrong.
6. Pandora's Box Isn't Opening. It Already Did.
By now, some readers have probably reached one of two conclusions.
The first is that I am hopelessly optimistic about artificial intelligence.
The second is that I have underestimated its dangers.
The irony is that neither is true.
Artificial intelligence is not the greatest invention in human history.
Nor is it the greatest threat.
It is something considerably more difficult to understand.
It is a force.
Like electricity.
Like the internal combustion engine.
Like the Internet.
Like every technology that fundamentally altered civilization, it will improve some things, disrupt others and create entirely new problems that no one—not even its inventors—fully anticipates.
That is precisely why simplistic arguments on either side miss the point.
If you believe AI will solve every problem facing humanity, history suggests you're wrong.
If you believe banning or slowing AI will somehow preserve the world as we know it, history suggests you're equally wrong.
The future has never shown much respect for either wishful thinking or nostalgia.
Let's begin with the critics.
They're right about more than many AI enthusiasts care to admit.
Artificial intelligence will eliminate jobs.
Not all jobs.
But enough to cause real pain.
Entire occupations will shrink while others disappear altogether. Research from organizations such as the IMF, the OECD, McKinsey and the World Economic Forum consistently predicts substantial labour market disruption over the coming decade. The precise numbers differ, but the direction does not. Millions of workers will need to adapt, retrain or redefine what they do.
The critics are also right about misinformation.
Deepfakes will become more convincing.
Fraud will become more sophisticated.
Algorithmic bias remains a serious concern.
Privacy deserves stronger protection.
The concentration of AI capability within a relatively small number of companies raises legitimate questions about competition, democratic accountability and economic power.
None of those issues should be dismissed.
They deserve regulation.
They deserve oversight.
They deserve public debate.
But they do not justify pretending that artificial intelligence can somehow be put back in the box.
Because that box no longer exists.
It is tempting to believe that governments will eventually gather around a table, negotiate sensible limits and collectively agree to slow the pace of AI development until society catches up.
It is an attractive idea.
It is also deeply unrealistic.
History has never worked that way.
When the steam engine appeared, countries did not compete by agreeing to use fewer steam engines.
When electricity transformed industry, nations did not slow electrification until everyone felt comfortable.
When aviation reshaped commerce and defence, governments did not decide that flying had become too disruptive.
Competition made those technologies inevitable.
Artificial intelligence is governed by the same force.
Only faster.
Imagine for a moment that Canada decided to halt advanced AI development for five years.
Would the United States?
Would China?
Would India?
Would Europe?
Would the hundreds of private companies investing hundreds of billions of dollars simply pause and wait?
Of course not.
The same principle applies to business.
Imagine a pharmaceutical company announcing that it will no longer use AI to accelerate drug discovery because it wishes to preserve traditional research methods.
Its competitors would thank them.
Imagine an engineering firm deciding to ignore AI-assisted design while everyone else adopts it.
Imagine a law firm refusing AI-assisted legal research.
Imagine a hospital insisting physicians diagnose patients without intelligent clinical support because "that's how we've always done it."
Those organizations would not be applauded.
They would become cautionary tales.
That is the uncomfortable reality of competition.
It has no reverse gear.
Every CEO understands it.
Every venture capitalist understands it.
Every military strategist understands it.
Increasingly, every government understands it.
The only institution that still seems to believe it has the luxury of time is education.
There is another irony hiding in plain sight.
Many people argue that AI development should slow.
What they often mean—whether they realize it or not—is that everyone else should slow down too.
History offers little encouragement.
The company that hesitates loses market share.
The scientist who delays discovery is overtaken.
The nation that falls behind spends decades trying to catch up.
Innovation has never been driven by consensus.
It has almost always been driven by competition.
Stanford's annual AI Index documents accelerating investment, accelerating model capability and accelerating adoption across industries. McKinsey reports that organizations using AI are moving rapidly from experimentation to integration. Governments across the world are publishing national AI strategies not because they enjoy technological disruption but because they recognize that failing to compete carries risks of its own.
The race is no longer approaching.
It is underway.
Does that mean anything goes?
Absolutely not.
The faster AI develops, the more urgently society needs guardrails.
We need transparency.
We need independent auditing.
We need accountability when AI systems fail.
We need privacy protections that mean something rather than merely appearing in lengthy terms and conditions nobody reads.
We need international cooperation on safety.
We need ethical standards.
We need laws capable of evolving as quickly as the technologies they regulate.
None of those are arguments against artificial intelligence.
They are arguments for governing it intelligently.
There is an enormous difference.
One accepts reality.
The other attempts to negotiate with it.
Throughout this essay I have returned repeatedly to the same challenge.
IF THAT IS YOUR POSITION, THEN BE CONSISTENT.
If you believe AI should be regulated...
IF THAT IS YOUR POSITION, THEN BE CONSISTENT.
Demand equally rigorous oversight wherever powerful technologies place lives, privacy or democracy at risk.
If you believe AI will transform employment...
IF THAT IS YOUR POSITION, THEN BE CONSISTENT.
Stop asking whether change is coming and start asking whether our schools, universities and workplaces are preparing people for it.
If you believe artificial intelligence is one of the defining technologies of this century...
IF THAT IS YOUR POSITION, THEN BE CONSISTENT.
Then stop treating it like an optional topic to be squeezed into an already overcrowded curriculum.
Teach children how to question it.
Teach them how to challenge it.
Teach them when not to trust it.
Teach them to recognize its limitations as well as its extraordinary potential.
Most importantly, teach them to become better thinkers because AI exists—not despite it.
Perhaps that is where this entire debate has gone wrong.
We have spent years arguing about whether artificial intelligence belongs in education.
The rest of the world quietly decided it belongs in medicine.
Engineering.
Science.
Finance.
Agriculture.
Manufacturing.
Government.
Law.
Architecture.
Virtually every profession today's children will eventually enter.
Only education continues behaving as though time itself can be negotiated.
It cannot.
Pandora's Box is open.
No committee will close it.
No government will uninvent it.
No policy will persuade competitors to abandon it.
No amount of nostalgia will return us to a world where artificial intelligence does not exist.
The real choice before us has never been whether AI should shape the future.
That decision has already been made.
The real choice is whether we prepare the next generation to shape AI—or whether we leave them to be shaped by it.
History has always been remarkably unforgiving toward societies that prepared children for a world that no longer existed.
I see little reason to believe this time will be any different.
We can debate artificial intelligence. We can regulate artificial intelligence. We can—and absolutely should—build intelligent guardrails around artificial intelligence. What we cannot do is pretend that it is patiently waiting outside the classroom until education decides to let it in. It is already sitting in the front row. The only unanswered question is whether we intend to notice before another generation graduates into a world it was never prepared to enter.
7. The Morning After
Tomorrow morning, at exactly 8:12 a.m., Sarah Bouchard will unlock the door to Room 214.
She has been teaching for nineteen years.
She loves her students.
She still arrives before anyone else.
Long before the buses pull into the parking lot, she will arrange the desks, write the day's objective on the board and quietly hope that somehow she can prepare twenty-eight young people for a future she herself struggles to describe.
None of this is her fault.
And yet, everything that happens inside that classroom matters.
The first student through the door is Noah.
He is fourteen.
He loves robotics, teaches himself Python at night and spends weekends building AI applications that solve problems nobody assigned him to solve. Last month he created an AI tutor that helped his younger sister finally understand fractions.
He has never shown it to his teacher.
He doesn't think she's allowed to see it.
Emma arrives next.
She is an artist.
She uses AI to explore colour palettes she never would have imagined on her own, then spends hours turning those ideas into paintings that are unmistakably hers. AI didn't replace her creativity.
It expanded it.
She keeps that to herself too.
Then comes Liam.
School has always been difficult for him. Reading takes longer. Organizing his thoughts feels like climbing a mountain in winter. Six months ago he discovered that an AI tutor could explain the same concept five different ways without ever becoming impatient.
For the first time in years, he began raising his hand in class.
He smiles more now.
No one at school knows why.
By 8:30 the room is full.
Sarah closes the door.
"Good morning, everyone."
Twenty-eight voices answer.
"Good morning."
She smiles.
"Before we begin, I want everyone's phones away."
The room fills with the familiar rustling of backpacks.
"And remember..."
She hesitates for only a moment.
"No artificial intelligence today."
Noah quietly closes the application that helped him understand quantum physics the night before.
Emma slips her tablet into her bag.
Liam stares at the blank page in front of him.
Nobody argues.
They've heard it before.
Sarah begins the lesson.
Outside that classroom, the world is moving at extraordinary speed.
Hospitals are diagnosing disease with the assistance of AI.
Scientists are discovering new materials.
Engineers are designing safer bridges.
Researchers are accelerating breakthroughs that once took years.
Businesses are reinventing themselves.
Entire professions are being reshaped.
Inside Room 214, twenty-eight students are being asked to pretend none of it exists.
Sarah doesn't know that UNESCO has warned education systems that artificial intelligence requires far more than simply adding another unit to the curriculum. It requires rethinking what it means to learn in an age where information is instantly available and judgment has become more valuable than memorization.
She hasn't had time to read the report.
She was coaching volleyball after school.
Then writing report cards.
Then answering parent emails.
She doesn't know that the OECD has reached much the same conclusion—that success in tomorrow's world will depend increasingly upon creativity, critical thinking, collaboration and the ability to work effectively alongside intelligent technologies.
No one has ever shown her how to teach those things in an AI-rich classroom.
No one prepared her for this moment.
Sarah is not failing.
She has been failed.
Failed by teacher preparation programs that never imagined this future.
Failed by curriculum that changes at the pace of committee meetings while technology changes at the pace of discovery.
Failed by a system asking extraordinary teachers to solve extraordinary problems with yesterday's tools.
And while she has been asked to stand still...
The world hasn't.
That may be the greatest injustice in this entire debate.
Not toward artificial intelligence.
Toward teachers.
We have placed them on the front line of the greatest technological transformation in modern history and then criticized them for not knowing how to fight a battle nobody trained them for.
That isn't leadership.
It's abandonment.
But there is another injustice unfolding quietly beside it.
One that belongs to the children.
Imagine Noah graduating in four years.
Imagine Emma.
Imagine Liam.
Imagine them walking into workplaces where every engineer, every architect, every physician, every scientist, every lawyer, every entrepreneur and every designer works confidently alongside artificial intelligence.
Imagine their employer asking a simple question.
"So...tell me about your experience working with AI."
And imagine Noah answering,
"I wasn't really allowed to."
Emma says,
"We were told it might interfere with our learning."
Liam lowers his eyes.
"I learned how to use it at home."
What should that employer conclude?
More importantly...
What should those young adults conclude?
That school prepared them?
Or that they succeeded despite it?
History rarely announces the moment an institution begins losing public confidence.
It happens one family at a time.
One conversation at a time.
One parent driving home after another school meeting asking a question that suddenly seems impossible to ignore.
"If the world outside school rewards my child for mastering artificial intelligence, why does the world inside school punish them for learning to use it?"
That question will spread.
Not because politicians demand it.
Not because technology companies market it.
Because parents love their children more than they love institutions.
And when enough parents begin believing that the education their children receive no longer resembles the world they will inherit, loyalty to the system will give way to loyalty to their sons and daughters.
History suggests that is the moment institutions either reinvent themselves...
...or become museums.
This article has never been about defending artificial intelligence.
It has been about defending children.
It has never been about replacing teachers.
It has been about giving teachers the tools, the preparation and the trust they deserve.
It has never been about technology.
It has always been about responsibility.
One day, Noah, Emma and Liam will look back and ask what the adults in their lives did when the world began changing faster than education.
I hope they discover that we argued.
I hope they discover that we questioned.
I hope they discover that we demanded thoughtful guardrails, ethical standards and responsible leadership.
But more than anything...
I hope they discover that, when the evidence became impossible to ignore, we found the courage to change.
Because if they discover instead that we spent our time protecting yesterday while sacrificing tomorrow...
they will not judge artificial intelligence.
They will judge us.
And history will not ask whether we were afraid of change. It will ask whether we loved our children enough to change with it.
Part 2
How We Used ChatGPT to Develop an Article That Remained Our Own
The article did not begin when ChatGPT generated its first paragraph. It began when I decided what I wanted to say.
That distinction is important.
I did not ask artificial intelligence to choose a topic, form an opinion, and produce something I could attach my name to without further thought. I entered the process with a position based on my own experience in education, my knowledge of technological change, and my concern for the future facing children and grandchildren. ChatGPT helped me explore, test, organize, and express that position, but it did not originate the purpose behind it.
The process unfolded through a series of deliberate steps.
1. I established the intellectual purpose
I began by explaining the central issue I wanted to address: the growing backlash against artificial intelligence. I identified several dimensions of that backlash, including environmental concerns, energy consumption, dependency, employment disruption, and public anxiety.
I also made my position clear. I did not believe that every criticism of AI was wrong, but I believed that much of the opposition was being driven by emotion, inconsistency, and resistance to change rather than by a balanced examination of evidence.
I then identified the larger issue that mattered most to me: education was responding far too slowly to a technological transformation that would profoundly affect the lives and employment prospects of young people.
Before ChatGPT wrote anything, I had therefore established:
the topic;
the central argument;
the intended audience;
the educational implications;
the desired emotional impact;
and the broader purpose of the article.
The AI did not determine what I believed. I told it what I believed and asked it to help me investigate and articulate that position.
2. I asked the AI to help design the inquiry
Rather than immediately requesting a complete article, I first asked ChatGPT to propose a series of prompts that would allow the subject to be examined in depth.
This was an important decision. A weak use of AI would have been:
“Write me a blog about the backlash against artificial intelligence.”
That might have produced a coherent article, but it would probably have been generic, predictable, and intellectually shallow.
Instead, the initial inquiry was broken into distinct areas:
the causes of the AI backlash;
historical resistance to transformative technologies;
the economic implications of AI;
agentic AI and its emerging capabilities;
the failure of education to adapt;
the psychology of technological resistance;
contradictions in the arguments against AI;
the strongest case in favour of adoption;
the strongest case against it;
and the eventual construction of the complete article.
The AI helped create an intellectual roadmap. I then reviewed that roadmap and decided the order in which the ideas should be developed.
This is comparable to asking a research assistant to help organize a complex investigation. The assistant may propose a structure, but the author decides whether that structure serves the intended argument.
3. I worked on one component at a time
I did not ask ChatGPT to generate the entire article in one attempt.
We developed it section by section.
This allowed me to examine every part independently and to decide whether it contributed meaningfully to the developing argument. It also prevented the AI from controlling the overall direction simply because it had generated a polished first draft.
For each section, I considered questions such as:
Was the argument accurate?
Was it consistent with what I believed?
Was it persuasive?
Was it too cautious?
Was it too abstract?
Did it sound like me?
Would it provoke the reaction I wanted?
Was it repeating ideas already covered?
Did it connect naturally to the next section?
The article emerged through a sequence of decisions rather than through a single act of generation.
4. I rejected language that weakened the argument
The first drafts frequently contained ideas I accepted but language I did not.
For example, ChatGPT initially suggested that sarcasm should be directed toward ideas rather than toward readers. I disagreed. My purpose was not simply to analyze the backlash intellectually. I wanted readers to recognize themselves in the patterns being described and to feel sufficiently uncomfortable to reconsider their positions.
I did not accept the AI’s rhetorical recommendation merely because it sounded reasonable. I challenged it.
I explained that attacking an idea can allow readers to remain emotionally detached. I wanted the argument to become personal without simply insulting people. That led to a more precise formulation:
IF THAT IS YOUR POSITION, THEN BE CONSISTENT.
That refrain became one of the central devices in the article.
It did not come from passively accepting the AI’s first suggestion. It emerged from my dissatisfaction with the suggestion, my explanation of what was missing, and the AI’s subsequent revision.
This is a crucial part of responsible AI use. Students should understand that rejecting an AI response is not evidence that the tool has failed. It may be evidence that the user is thinking.
5. I repeatedly revised the tone
The early drafts were often too wordy, too conciliatory, or too fragmented.
I asked for greater concision.
Then I asked for more bite.
When the writing became too choppy, I requested longer, more connected sentences and a more natural flow. When it became too analytical, I asked for descriptive passages and narrative. When it became too clinical, I asked for dialogue, recognizable people, and situations that readers might experience in their own homes, schools, and workplaces.
These were not cosmetic changes.
Tone affects meaning.
A paragraph that is technically accurate may still fail if it does not make the reader care. I therefore treated style as part of the intellectual work rather than as decoration added afterward.
At different points, I directed ChatGPT to make the writing:
more concise;
more sarcastic;
less conciliatory;
more flowing;
more descriptive;
more uncomfortable;
more lyrical;
and more emotionally resonant.
The final voice developed through repeated negotiation between what the AI produced and what I recognized as appropriate for the article.
6. I used my own judgment to reshape the argument
The order of the original prompts was changed as the article developed.
I initially planned to move from the historical discussion into the economic consequences of AI. I later decided that agentic AI had to come first because readers needed to understand how rapidly the technology itself was changing before considering its broader consequences.
That was an editorial decision based on the needs of the argument.
I recognized that readers who still understood AI primarily as a chatbot would not fully appreciate the urgency of the economic or educational sections. The article therefore needed to explain, through everyday examples, the shift from systems that answer questions to systems that can pursue objectives, coordinate tasks, and complete increasingly complex work.
The AI did not independently make that structural judgment. I changed the sequence because I saw that the existing plan would not provide sufficient context.
7. I insisted that abstract claims be made real
Many initial responses discussed AI in broad categories such as productivity, labour markets, healthcare, and education. Those concepts were valid, but they remained distant.
I asked ChatGPT to put people inside the argument.
The result included scenes involving:
a physician beginning an overcrowded day with patient information already reviewed;
a teacher sitting at a kitchen table late at night with marking still unfinished;
a small business owner spending Sunday afternoon reconciling invoices;
a parent listening to a child explain why AI was prohibited at school while adults used it professionally;
and a Home and School meeting where parents gradually confronted the contradictions in their own positions.
These narratives were not included simply to make the article entertaining. They translated abstract claims into human consequences.
The discussion of agentic AI became a discussion about time returned to doctors, teachers, families, and business owners.
The discussion of educational policy became a conversation at a kitchen table.
The discussion of cognitive bias became a debate among parents who initially agreed with one another but gradually recognized that their reasoning was inconsistent.
The research remained important, but the narratives gave the research a human face.
8. I integrated evidence into the story
I did not want a narrative section followed by a separate academic explanation. That would have interrupted the emotional momentum and made the research feel attached rather than integrated.
I therefore asked that evidence from UNESCO, the OECD, psychology, behavioural science, and AI research appear within the developing scenes.
The research was used to confirm what characters were experiencing:
UNESCO’s warnings appeared as a parent investigated whether education systems were prepared for AI.
OECD findings appeared as she considered the growing importance of judgment, creativity, collaboration, and human-AI interaction.
Research by Daniel Kahneman, Paul Slovic, and Jonathan Haidt was introduced during the Home and School debate to explain why emotional beliefs often survive contradictory evidence.
Forecasts concerning AI adoption, competition, and labour disruption were incorporated into the conclusion rather than isolated in a technical appendix.
This approach demonstrated that research and narrative do not have to compete. Evidence can deepen a story, while a story can make evidence understandable.
9. I tested the argument against opposing views
The article was not built on the assumption that every criticism of artificial intelligence was foolish.
I wanted both sides of the argument represented.
The conclusion acknowledged that AI could:
eliminate or transform jobs;
intensify misinformation;
increase fraud;
reinforce bias;
threaten privacy;
concentrate economic power;
and create serious governance challenges.
These concerns were not added as token concessions. They were treated as legitimate reasons for regulation, oversight, transparency, and enforceable guardrails.
At the same time, I wanted the article to confront an unavoidable reality: competition among corporations and nations makes a complete halt to AI development extraordinarily unlikely.
That produced a position more nuanced than either blind enthusiasm or outright rejection:
AI requires governance, but governance is not the same as pretending the technology can be uninvented.
The final conclusion was stronger because the opposing case had been acknowledged rather than caricatured.
10. I used the AI as an intellectual opponent
At several points, ChatGPT offered caution that I believed weakened the article.
Sometimes I accepted its reasoning. At other times I rejected it.
This created a form of intellectual friction.
I had to explain:
why a claim was too soft;
why a rhetorical strategy would allow readers to remain detached;
why a scene felt unrealistic;
why a character needed greater depth;
why a paragraph lacked flow;
why research needed to be placed inside the narrative;
and why the conclusion needed to be brutally honest without becoming apocalyptic.
Each criticism required me to articulate what I wanted more precisely.
The AI’s imperfect responses therefore became useful. They forced me to clarify my standards and sharpen my thinking.
This is one of the most educationally valuable uses of AI. Students do not learn merely by receiving strong answers. They also learn by diagnosing weak ones.
11. I made the article increasingly personal
Near the end of the process, I identified something missing from the article.
It was being written by two grandfathers who had spent their careers in education and who cared deeply about the world their grandchildren would inherit.
That perspective changed the emotional centre of the work.
The article was no longer simply an argument about artificial intelligence. It became an expression of intergenerational responsibility.
I asked that the opening explain why two grandfathers were writing the piece and that the closing return to the same concern: whether adults cared enough about children to prepare them for the world they would actually enter.
This was not an idea supplied by the AI.
It came from knowing who we were, why the issue mattered to us, and what we ultimately wanted readers to remember.
The AI then helped express it.
12. I remained responsible for every accepted sentence
ChatGPT generated many sentences, scenes, transitions, and possible arguments.
I did not accept all of them.
Some were rejected completely. Others were rewritten several times. Some concepts survived while the language changed. Some structures were moved. Some sections were abandoned or combined. Certain phrases became central only after repeated revisions.
The completed article therefore reflects a chain of human decisions:
what to ask;
what to keep;
what to reject;
what to challenge;
what to emphasize;
what evidence to require;
what tone to adopt;
what stories to tell;
and what conclusion to defend.
The AI contributed language and possibilities.
I exercised authorship through intention, selection, organization, revision, judgment, and accountability.
A fair description would therefore be:
This was an AI-assisted article directed, shaped, and critically edited by its human authors.
That description acknowledges the contribution of ChatGPT without surrendering the role of the people who determined what the article was ultimately meant to say.
Why This Is a Model for Student Use
This process bears little resemblance to asking AI to write an assignment and submitting the result unchanged.
A student following the process used here would have to:
begin with a position or genuine question;
explain the purpose and audience;
divide the inquiry into manageable parts;
evaluate each AI response;
challenge weak reasoning;
request evidence;
compare opposing arguments;
detect repetition and inconsistency;
make structural decisions;
revise tone and style;
integrate personal knowledge or experience;
and assume responsibility for the final work.
Those are not shortcuts around thinking.
They are acts of thinking.
The educational question should therefore not be whether a student used artificial intelligence. That question tells us almost nothing.
The better questions are:
What decisions did the student make?
What did the student reject?
How did the student verify claims?
How did the argument change during the process?
What personal knowledge did the student contribute?
Can the student defend the final conclusions?
Can the student explain why the work took its final form?
A student who cannot answer those questions probably did not author the work in any meaningful sense.
A student who can answer them may have engaged in a more demanding process than one who wrote a conventional essay from memory the night before it was due.
What Educators Should Assess
If students are permitted to use AI, assessment must move beyond examining only the final product.
Teachers should also examine the intellectual trail that produced it.
Students could be asked to submit:
their original statement of purpose;
the prompts they developed;
selected AI responses;
examples of output they rejected;
explanations of major revisions;
sources used to verify factual claims;
a reflection on what the AI misunderstood;
and an oral defence of the final argument.
This makes the student’s thinking visible.
It also makes it much harder to outsource the entire task without understanding it.
A student who simply submits generated text will struggle to explain the choices behind it. A student who has genuinely directed the process will be able to discuss the development of the work in detail, including the moments when the AI was unhelpful, inaccurate, repetitive, or stylistically inappropriate.
The process therefore offers educators something valuable: not the disappearance of assessment, but the possibility of assessing judgment more authentically.
The Larger Lesson
The article was not mine because I personally typed every sentence.
Typing is not the highest form of authorship.
It was mine because the concern was mine, the purpose was mine, the argument was mine, the standards were mine, the editorial decisions were mine, and the responsibility for the final position was mine.
ChatGPT expanded what I could produce. It helped generate alternatives, expose weaknesses, create scenes, and reformulate ideas. At times it offered language I would not have produced independently. At other times it produced language I rejected immediately.
That is what collaboration looks like.
Used poorly, artificial intelligence can allow a student to avoid thinking.
Used well, it can force a student to think more carefully, make more decisions, examine more alternatives, and defend those decisions with greater precision.
The difference does not lie in the technology.
It lies in the process.
The purpose of education should not be to prove that students can complete intellectual work without artificial intelligence. It should be to ensure that, with or without artificial intelligence, the thinking remains unmistakably their own.
A New Assessment Model for the Age of Artificial Intelligence
The arrival of large language models has exposed a fundamental weakness in the way schools have traditionally assessed learning.
For generations, the finished product—a report, an essay, a presentation or a research paper—served as the primary evidence that learning had taken place. Teachers naturally focused on the quality of the final submission because, until recently, the path that produced it remained largely invisible. Students either demonstrated what they knew or they did not.
Artificial intelligence has changed that assumption forever.
Today, a student can produce a polished essay in minutes. Some educators have interpreted this as the end of authentic assessment and have responded by attempting to ban AI from classrooms. Others have simply added a line to existing rubrics asking whether students "used AI appropriately."
Neither response addresses the real issue.
The arrival of AI has not simply changed how students write.
It has changed what it means to learn.
For that reason, adding a few new criteria to yesterday's writing rubric is rather like attaching a faster horse to the front of an automobile. The appearance changes, but the underlying model remains exactly the same.
The question before us is much more profound.
What should schools actually be assessing in an age when information is abundant, writing can be assisted by intelligent systems, and the ability to direct those systems has become a valuable intellectual skill in its own right?
The answer, I believe, requires a complete shift in educational thinking.
For more than a century, schools have primarily assessed what students know, how accurately they remember it and how effectively they communicate it. Those skills remain important, but they are no longer sufficient.
Artificial intelligence has fundamentally altered the nature of intellectual work.
Students are no longer working in isolation.
Increasingly, they are learning, investigating, questioning, creating and solving problems in partnership with intelligent systems.
That reality demands a different assessment model.
The Fundamental Shift
For decades we have assessed writing.
Today we have the opportunity to assess something far more important.
WE ARE NO LONGER ASSESSING WRITING.
WE ARE ASSESSING THINKING.
Writing is no longer the scarce skill.
Judgment is.
Typing is no longer the scarce skill.
Reasoning is.
Finding information is no longer the scarce skill.
Determining whether that information is accurate, relevant and meaningful has become the real intellectual challenge.
In other words, the arrival of AI has shifted the educational spotlight away from production and toward judgment.
The central question is no longer,
"Can this student write an essay?"
The more important question has become,
"Can this student think well enough to direct, challenge, refine, improve and ultimately take responsibility for work developed in collaboration with an intelligent system?"
That is a profoundly different form of learning.
More importantly, it is the form of learning our graduates will encounter in medicine, engineering, law, science, business, journalism, architecture and virtually every other knowledge profession.
Schools now face a choice.
We can continue assessing the final product while pretending the thinking process has not changed.
Or we can begin assessing the thinking itself.
The framework that follows is built upon the second assumption.