Did the student successfully complete the assignment—or did AI complete the learning?
Artificial intelligence is rapidly becoming part of how young people learn, research, communicate, create and eventually prepare to enter the workforce. That reality raises questions that go far beyond whether students should or shouldn't be allowed to use AI in the classroom. The more important conversation may be about what happens to learning itself when technology becomes capable of doing an increasing amount of the intellectual work.
That is the idea behind a new Generation AI™ discussion series I'm calling “Would You Let AI Answer This?”
Over the course of this series, I want to explore situations involving education, careers, misinformation, ethics, relationships, decision-making and leadership. These aren't intended to be arguments for or against artificial intelligence. AI is already becoming part of the world young people are inheriting. The challenge now is helping them learn how to use these extraordinary tools without outsourcing the very human capabilities they will need to succeed alongside them.
Critical thinking, curiosity, judgment, empathy, integrity, accountability and leadership don't become less important because artificial intelligence becomes more capable. I believe the opposite may be true. As technology gets better at providing answers, our ability to evaluate those answers, question them and decide what to do with them may become even more valuable.
So let's begin with a situation that is probably already playing out in classrooms around the world.
The Perfect Assignment
Imagine a 16-year-old student who has been assigned a history paper. The student understands the basic topic but isn't quite sure where to begin, so they turn to an AI assistant.
At first, the student asks AI to help develop a thesis. That's useful. Next, they ask it to organize the argument and suggest some areas for research. Then they ask it to improve a few paragraphs. AI identifies weaknesses in the argument and suggests stronger supporting evidence. Eventually, it helps refine the language, improve the structure and polish the final paper.
The result is outstanding. The facts are accurate. The arguments are compelling. The writing is polished and sophisticated. The student submits the assignment and receives an A.
On the surface, everything appears successful.
Then the teacher asks the student to explain one of the central arguments in the paper.
The student struggles.
The teacher asks why one of the historical events discussed in the assignment was significant and how it influenced what happened next. Again, the student has difficulty explaining the reasoning contained in their own paper.
Suddenly, that A becomes much more interesting.
The assignment demonstrated excellent work. But did it demonstrate that learning actually occurred?
When Assistance Becomes Substitution
This is where I believe the debate surrounding AI in education becomes much more complicated than simply asking whether students should be allowed to use it.
AI can be an extraordinary educational tool. A student can use it to brainstorm ideas, identify gaps in an argument, receive feedback, explore opposing perspectives, simplify a difficult concept or discover new questions worth researching. Used this way, AI doesn't necessarily replace learning. It can potentially deepen it.
But there is a line somewhere between AI helping a student think and AI doing the thinking for the student.
That line isn't always obvious.
If AI suggests a stronger thesis and the student evaluates it, changes it and can explain why the revised argument is better, learning may have occurred. If AI generates the thesis, builds the argument, selects the supporting evidence and writes the explanation while the student primarily transfers the result into an assignment, something very different has happened.
Both students may submit equally impressive papers.
Only one may actually understand what the paper says.
And that presents educators with a challenge that may grow significantly as generative AI improves.
The Finished Product May No Longer Tell the Whole Story
For generations, teachers have been able to look at a student's finished work and draw at least some conclusions about that student's understanding. Essays, reports and research papers weren't perfect measurements, but the work itself generally reflected a significant amount of the thinking required to produce it.
Artificial intelligence changes that relationship.
A beautifully written paper can now potentially tell us more about the capabilities of the technology used to create it than about the capabilities of the person whose name appears at the top.
That doesn't mean the essay is dead. It may mean the way we evaluate learning needs to evolve.
Perhaps the finished assignment becomes only one part of the assessment. A teacher might also ask the student to verbally defend the argument, explain why certain sources were selected, identify something AI initially got wrong, describe which AI suggestions they rejected, or apply the central idea from the assignment to a completely different situation.
Those questions measure something AI cannot demonstrate on behalf of the student: whether the student actually understands.
And perhaps that is where education becomes particularly interesting in the AI era.
Human Intelligence + Artificial Intelligence
In Generation AI: Preparing Tomorrow's Leaders Today, I explore what I call the Four Intelligences: Human Intelligence, Artificial Intelligence, Leadership Intelligence and Character Intelligence. This particular scenario sits directly at the intersection of the first two.
Artificial Intelligence gives young people access to capabilities previous generations could scarcely have imagined. It can retrieve information, analyze material, explain concepts, generate ideas and help transform rough thoughts into polished communication.
But Human Intelligence remains responsible for something equally important: curiosity, critical thinking, creativity, context and judgment.
Knowing how to prompt an AI system may become an important skill. Knowing when its answer doesn't make sense may be even more important.
The goal, therefore, shouldn't necessarily be to remove AI from the classroom. Trying to prepare young people for an AI-powered world by preventing them from using AI would create its own contradiction. These tools will almost certainly be part of the universities, organizations and careers many of today's students eventually enter.
Instead, we should be asking a more difficult question: How do we allow students to use AI while still requiring them to think?
That could mean changing assignments. It could mean changing assessments. It may require educators to evaluate the process behind an answer alongside the quality of the answer itself. And it may require students to learn that getting something correct isn't necessarily the same as understanding why it is correct.
The Assignment Isn't the Destination
There is an understandable temptation to frame this conversation around cheating. Certainly, academic integrity matters, and there will be situations where students use AI in ways that clearly violate expectations. But focusing exclusively on cheating risks overlooking the larger issue.
A student can follow every rule established by a school and still become overly dependent on technology if the rules themselves haven't caught up with what AI can do.
The larger objective of education isn't producing perfect assignments.
It's developing people who can think.
We want young people who can encounter unfamiliar information and make sense of it. We want them to question assumptions, recognize misinformation, defend their reasoning, change their minds when evidence demands it and make decisions when there isn't a convenient answer waiting for them.
Those capabilities will eventually follow them out of the classroom.
One day, the assignment may become a business proposal, a financial recommendation, a medical decision, a hiring recommendation, a piece of journalism, a public policy proposal or a leadership decision affecting hundreds of people.
AI may be there to help.
But someone still needs to understand the answer.
Perhaps that is the lesson hidden inside our perfect history paper. The challenge isn't choosing between Human Intelligence or Artificial Intelligence. It's developing a generation capable of bringing the two together.
Because ultimately, the assignment isn't the destination. The student is.
Would You Let AI Answer This?
So here is the first question in the Generation AI™ — Would You Let AI Answer This? series.
A 16-year-old student submits an outstanding AI-assisted history paper and receives an A. The facts are correct, the arguments are compelling and the writing is excellent. But when asked to explain the central argument in their own words, the student struggles.
Did the student successfully complete the assignment—or did AI complete the learning?
And perhaps there's an even more important question for educators, parents, students and leaders:
How would you redesign the assignment so the student could use AI while still having to demonstrate genuine understanding?
I'd love to hear your perspective. These are exactly the kinds of conversations I hope this series creates, particularly as I begin thinking about the questions worth exploring in the next edition of Generation AI.
About the Series
Would You Let AI Answer This? is a Generation AI™ discussion series exploring the decisions young people, parents, educators and leaders will increasingly face as artificial intelligence becomes part of everyday learning, work and decision-making. The series is inspired by Mike Kehoe's Generation AI: Preparing Tomorrow's Leaders Today and its exploration of Human Intelligence, Artificial Intelligence, Leadership Intelligence and Character Intelligence.
#GenerationAI #WouldYouLetAIAnswerThis #AIinEducation #ArtificialIntelligence #CriticalThinking #FutureOfEducation #HumanIntelligence #ResponsibleAI #Education #FutureOfWork