The rise of artificial intelligence promises speed and efficiency in education, yet a handwritten thank-you note from a student reminded one professor of a vital truth: meaningful learning requires friction. In a world racing to automate, intentionally slowing down, reintroducing effort, and demanding visible thinking may be the most effective way to harness AI’s potential.
The Paradox of Speed
AI excels at accelerating tasks – faster feedback, drafts, translations, and completion. However, simply doing things faster doesn’t guarantee better learning. In fact, it can undermine the very processes that build understanding. The real question isn’t how to use AI to speed up education, but how to use it to deliberately restore the productive friction that makes growth meaningful.
Content Mastery: Depth Over Coverage
With AI capable of generating instant answers, students can appear fluent without actually understanding. This highlights the critical need for deeper content mastery. Students must first build a foundational knowledge base – core concepts, key vocabulary, and timelines – to reliably evaluate AI-generated information. Without this background, new ideas have no context, and learning becomes superficial.
The solution? Prioritize fewer topics and deeper engagement through frequent, low-stakes routines like retrieval practice, quizzes, and short writing exercises. These create the friction of recall, making knowledge durable. AI can assist by providing targeted prompts and spaced review, but the cognitive load must remain with the student.
Transcreation, Not Just Translation
AI translation removes barriers, but can flatten meaning. Accurate translations may lack nuance, cultural sensitivity, or emotional resonance. A mechanical translation of a joyful community event flyer, for example, might be technically correct but feel robotic.
The key is transcreation: a slower, human-driven process that prioritizes tone, cultural fit, and genuine connection. AI can assist with first drafts and terminology, but the final voice should always be human-owned. Rushing to automation risks losing trust and flattening communication.
Making Language Visible
Writing tools polish sentences instantly, erasing the friction that helps language learners grasp underlying structures. Sentence diagramming, often dismissed as outdated, forces a slow, deliberate breakdown of grammar. It makes language visible, revealing how sentences function rather than simply accepting polished outputs.
AI can support this by visualizing structures, prompting revisions, and encouraging intentional analysis. The goal isn’t to punish students with old-school drills but to provide a pause that transforms language into something concrete and improvable.
Accountability in the Age of AI-Generated Work
AI can remove the friction of producing work, but also learning itself. Plagiarism and overreliance are real concerns. The answer isn’t just detection and punishment, but reintroducing friction into assessment. Oral defenses, process reflections, and “walk me through your thinking” prompts shift focus from the final product to the learning journey.
Authorship requires accountability for thinking. This means bringing back practices that demand explanation, revision, and justification, rather than simply submitting a polished result.
The Human Steering Wheel
AI is a powerful engine, but it needs a steering wheel. The most effective approach isn’t blindly embracing automation, but selectively recovering practices that foster meaningful learning.
The goal is to keep the friction that builds understanding while using AI to reduce unnecessary barriers. Borrowing from the past doesn’t make us less innovative; it might be the key to staying human while innovating. The design challenge is simple: keep the friction that matters, and let AI handle the rest.
