By Khaled Hassan, M.Ed., CELTA
The arrival of generative AI in education has sparked two opposing reactions. One camp views AI as an existential threat to learning, while the other celebrates it as the greatest educational innovation since the internet. I believe both perspectives miss a more fundamental question.
The real question is not whether students should use AI.
The real question is whether classrooms can preserve the kind of dialogue that has always been at the heart of meaningful learning.
The Classical Classroom
For centuries, education has relied on dialogue. From Socrates questioning his students in ancient Athens to modern classrooms built around discussion, debate, questioning, and collaborative problem-solving, dialogue has never been merely conversation. It has been the engine of thinking.
Educational theorists from Lev Vygotsky to contemporary learning scientists have argued that knowledge is socially constructed. Students do not simply receive information; they develop understanding by explaining ideas, defending arguments, questioning assumptions, and negotiating meaning with teachers and peers. Dialogic teaching views classroom talk not as an interruption to learning but as learning itself. (ScienceDirect)
A good teacher therefore does far more than deliver information. The teacher creates productive intellectual friction. A thoughtful question, a carefully timed challenge, or a disagreement between classmates often produces deeper learning than a perfect explanation ever could.
The New Participant in the Conversation
Today, every student carries an intelligent conversational partner in their pocket.
When students encounter difficulty, many no longer ask:
“What do I think?”
Instead, they ask:
“ChatGPT, what do you think?”
This seemingly small shift has profound implications.
Traditionally, classroom dialogue followed a familiar pattern:
Teacher ↔ Student
Student ↔ Student
Now a third participant has entered every conversation:
Teacher ↔ Student ↔ AI
Student ↔ AI
Teacher ↔ AI
This changes not only who speaks, but how learning unfolds.
The Hidden Risk
The greatest danger of AI in education is not plagiarism.
It is cognitive outsourcing.
If students increasingly delegate interpretation, summarisation, argument construction, and problem solving to AI, they may also reduce the very mental effort through which learning occurs.
Learning has always required productive struggle.
We become better writers by revising.
Better thinkers by defending ideas.
Better readers by wrestling with difficult texts.
When AI removes every obstacle, it may also remove many of the cognitive experiences that build expertise.
This does not mean AI should be banned. It means educators must carefully distinguish between using AI to support thinking and using AI to replace thinking.
Dialogue Must Be Redesigned, Not Abandoned
Rather than eliminating classroom dialogue, AI offers an opportunity to redesign it.
Imagine students first asking AI to generate three competing explanations for a controversial issue.
Instead of accepting the first response, students compare the arguments, identify assumptions, challenge weak reasoning, and defend their own conclusions in discussion with classmates.
In this model, AI becomes the starting point for dialogue rather than its replacement.
Recent work in dialogic pedagogy suggests that AI-mediated learning can be most effective when activities deliberately promote questioning, reflection, and collaborative discussion instead of passive acceptance of AI-generated answers. Researchers have proposed classroom routines that introduce “epistemic friction”—activities that require students to question, revise, and justify ideas before reaching conclusions. (Tandfonline)
The teacher’s role therefore becomes even more important.
Not as the primary source of information.
But as the architect of learning.
The Teacher of the Future
As AI becomes increasingly capable, excellent teaching will depend less on delivering information and more on designing learning experiences that cultivate curiosity, intellectual autonomy, critical thinking, and self-regulation.
Teachers will need to know when AI should explain, when students should struggle independently, and when dialogue among learners is more valuable than an immediate answer.
Perhaps the defining question for education over the next decade is not:
“How do we teach students to use AI?”
Instead, it may be:
“How do we preserve the uniquely human forms of dialogue through which students learn to think?”
If we answer that question well, AI may become one of the greatest educational tools ever developed.
If we answer it poorly, we risk creating classrooms where answers become easier to obtain, but thinking becomes increasingly rare.
The future of education will not be determined by artificial intelligence.
It will be determined by how wisely teachers redesign pedagogy to ensure that dialogue, reflection, and deep thinking remain at the centre of learning.
