Writing for NPR, Columbia University freshman Maximilian Milovidov offers a firsthand perspective on what thoughtful AI integration in education can look like. As a student in a course called “Writing AI“, where AI use was not just permitted but required, Milovidov argues that the real question is not whether students use AI, but how they are taught to use it responsibly.
Notably, Milovidov practiced what he preaches: the essay itself was edited for length using a large language model, while human editor Phoebe Lett handled content edits, a real-world demonstration of the balanced human-AI collaboration he advocates for in the classroom.
Below are key highlights from Milovidov’s eassy:
- Cognitive offloading risk is real, but conditional: The concern that AI weakens independent thinking only holds when students treat it as an all-knowing oracle. When used critically, questioning its outputs and documenting reasoning, AI becomes a study partner rather than a crutch.
- AI as an academic equalizer: For students without access to private tutors, AI can generate practice questions, give paragraph feedback, and suggest counterarguments. A 2025 Harvard study found that students using an AI tutor achieved learning gains more than double those in traditional classrooms.
- Moderation matters: One study found that students using moderate AI assistance during lectures outperformed both those relying fully on automation and those using minimal support.
- AI “Shame” drives underground use: Campus climates that ban or stigmatize AI don’t eliminate its use, they simply push it out of sight, removing any opportunity for guided, critical engagement.
When anyone can generate a passable paragraph, what distinguishes us is not whether we can produce text, but whether we can think, judge and revise.
– Maximilian Milovidov, NPR
In conclusion, Milovidov‘s experience suggests that banning AI in education may be both ineffective and counterproductive. Just as past generations could not halt the printing press, today’s students will reach for AI tools regardless of institutional policy. The more urgent challenge is designing learning environments that teach students to interrogate these systems, understanding their biases, limitations, and appropriate use, so they are prepared to work alongside AI rather than be displaced by it.
Acknowledgement: This summary was generated using AI writing tools as part of my AI Explorations process. For more information, please refer to the original article listed below.