As AI becomes part of everyday learning, educators face a growing challenge: how to use technology without weakening critical thinking
“Would you like me to help with the other problems on the list?”
That was the sentence a physics teacher recently found at the end of a pupil’s homework assignment. The solution itself was elegant and correct. Unfortunately, it was not produced by the child. It was generated by artificial intelligence and copied so carelessly that the pupil left in the chatbot’s question.
A video on this went viral because it was funny in the uneasy way bad news can be funny. Today’s schoolchildren, it seems, are not only forgetting how to think, but some are forgetting how to cheat properly.
This might have remained another sad school anecdote if President Vladimir Putin had not instructed the State Council at around the same time to prepare proposals for changing federal education standards and incorporating AI into them. So we are no longer discussing a toy, a novelty, or a passing panic – we’re discussing the future of Russian education.
At first glance, ordinary citizens may think this only concerns teachers and administrators. But the consequences won’t remain inside the classroom. They will shape the way children read, write, argue, remember, and think.
The statistics already tell the story. By 2025, the share of student work written with the help of AI had risen from 17.8% to 24%. Nearly a quarter of presentations, essays, coursework, and even dissertations are now being produced with AI assistance. Among school pupils, the scale is greater still; 29% of Russian pupils admit they use AI tools to do homework, while 23% use them out of boredom, as a substitute for real conversation.
Teachers don’t need surveys, because we see the problem every day. I once had a pupil who wrote excellent essays at home but consistently failed creative assignments in class. His homework passed anti-plagiarism checks perfectly, and I couldn’t accuse him without proof. The Russian language Unified State Exam settled the matter, as when he was deprived of his digital ghostwriter, he suffered a complete fiasco. His supposed literary ability belonged to a neural network.
If we don’t stop and think seriously about this uncontrolled integration of AI into education, the future looks bleak.
The risks identified by teachers and experts are real. In surveys, 36% of respondents say they fear reduced mental effort will damage children’s development while another 31% worry about the decline of face-to-face interaction. A further 27% fear a collapse in motivation and the rise of catastrophic laziness.
This is the central danger because AI doesn’t merely help a child avoid effort, it can imitate effort. It can produce the appearance of thought and even personality. A bad essay written by a child is still a human document as it contains errors, awkwardness, effort, fear, ambition, and sometimes buried inside, a living voice. A polished AI essay might contain none of this.
IT pioneer Natalya Kaspersky has said we risk raising “a generation of complete idiots.” You might not like the harshness of her words, but there’s a grain of truth in them. If a child today cannot even thoughtfully rewrite an answer produced by a machine, what will happen in two or three years? Will our children still write and formulate thoughts of their own, or will they outsource these basic human acts to an algorithm?
Still, pretending we can simply ban AI from the classroom would be childish, and burying your head in the sand never works. Nor does fanatically opposing innovation. Those who try to keep technology out of school entirely will lose, and the only serious answer is to teach children how to use AI intelligently, without surrendering their own minds to it.
AI can already help teachers. It can prepare tests and presentations and reduce the routine workload that consumes so much of a teacher’s time. It can analyze written work and identify recurring errors across many texts. While this doesn’t replace the teacher’s judgment, it does support it, and used properly, AI can become a useful tool rather than a cheat sheet.
The problem is that neither teachers nor pupils have yet been properly taught how to use this tool.
I remember how my parents, who were programmers, feared that calculators would destroy mathematical thinking. In the end, calculators destroyed nothing in those who had first learned to count, and they freed the brain instead for more complex work – but only after the basic skill was formed.
This is the principle we must apply to AI. Children must first learn to think, read, write, argue, doubt, calculate, and express themselves, and only then should the neural network become an additional layer of intellect.
What can a teacher or parent give a child that no neural network can provide? The answer is simple and old-fashioned: Living human contact and real conversation. The spark of shared thought and the discipline of disagreement. Emotion that hasn’t been simulated.
Even the most advanced AI can’t feel anything. But it can imitate feelings, and a lonely child might accept this imitation if nothing better is available.
Are teachers to blame? Only partly, but as long as Russia remains desperately short of teachers, neural networks will remain the most accessible ‘tutor’ for many families. For parents, they are convenient; for pupils, they are obedient. For administrators, they create the illusion of improved results.
That’s why the issue can’t be resolved by scolding children or laughing at them for copying chatbot replies. If we don’t restore the authority of the teacher, reduce bureaucracy, fill staff shortages, and teach both adults and children how to work with AI honestly, then the machine will take the place of thought.
And our children will continue typing ‘essay on the topic’ into a search bar, while the neural network politely finishes their education with the same fatal question:
‘Would you like me to help with anything else?’
This article was first published by the online newspaper Gazeta.ru and was translated and edited by the RT team
