The Real AI Revolution in Education? Time.

Education has always been shaped by time. School days are divided into periods. Curriculum maps are plotted week by week. Teachers race the clock to fit instruction, grading, and planning into finite hours. Students, too, operate on a schedule not of their own making, often waiting days for feedback or spending hours in class without personalized attention. Time, in education, is both a structuring force and a constraint.

AI could change that dynamic.

We often talk about artificial intelligence in education as a tool for personalization or efficiency. But beneath those familiar buzzwords lies a deeper, more radical possibility: AI can reshape how time functions in schools. And in doing so, it could either humanize or mechanize learning — depending on how we use it.

Time: The Silent Architect of School

For over a century, instructional time has been one of the central currencies of schooling. As David Berliner wrote in 1990, the amount and quality of time spent on learning tasks strongly correlates with academic achievement. Yet instructional time is frequently squandered — not out of neglect, but because human teaching is hard.

Teachers must plan lessons, differentiate instruction, monitor behavior, assess work, and respond in real time to student needs. These are cognitively demanding tasks, made more difficult by rigid schedules and large class sizes. TNTP's The Opportunity Myth (2018) found that students spent only 17% of classroom time on grade-level assignments. Much of the rest was lost to logistical churn, slow pacing, or content mismatched to student needs.

In professional development, I sometimes ask teachers to estimate how many minutes per class period are lost to transitions, disruptions, or logistical slowdowns — five minutes, ten, sometimes more. Then I have them calculate how many days of instruction that adds up to over a year. The results are always eye-opening. What seems like a few moments here and there can easily become weeks of lost learning time over the course of a school year.

AI as Time-Maker

Enter AI. Large language models and intelligent tutoring systems don’t just personalize content. They compress time. An AI tutor provides immediate feedback. An AI co-planner generates differentiated lesson ideas in seconds. An automated grader can return scores and suggestions before a student even stands up from their desk.

Pearson's report Intelligence Unleashed (2016) argued that AI’s most transformative potential is not just in tailoring instruction, but in radically increasing the speed and responsiveness of learning systems. UNESCO's Artificial Intelligence in Education (Holmes et al., 2019) echoed this, emphasizing AI's ability to "save teachers time" and extend students' active learning hours.

One pilot program using Khanmigo, Khan Academy's GPT-based tutor, found that students completed more practice problems per session and received more targeted guidance — not because they were working harder, but because they weren’t waiting for feedback. This is time made productive. But is it time well spent?

The Risks of Time Compression

Here lies the paradox. Learning is not merely an input-output process. Reflection, struggle, boredom, and dialogue are all part of the human texture of learning — and they take time. Compressing every step risks mistaking speed for depth.

Psychologist Gavriel Salomon warned in 1990 of "cognitive offloading" — when tools that make tasks easier also reduce deep engagement. The same risk applies here: if AI handles all the planning and responding, when do students (or teachers) wrestle with complexity? When do they reflect, revise, or question?

The push to "accelerate learning" post-COVID, while well-intentioned, has already shown signs of reducing learning to a treadmill. AI, used uncritically, could supercharge that treadmill. It could also reinforce inequities, as students in under-resourced schools are given AI tools in place of human contact, while wealthier peers use AI to augment already rich learning environments.

What Will We Do With the Time We Save?

The deeper revolution lies not in AI giving us more time, but in how we choose to use that time.

We could use it to double down on test prep, automate lesson delivery, and further deskill teaching. Or we could use it to make space for the things only humans can do: mentoring, listening, adapting with care, slowing down for meaning.

In the end, AI may not replace teachers entirely — but it might in some contexts, and it will almost certainly transform the role in ways we haven’t fully reckoned with. But it might redefine what teachers do with their time. If we’re intentional, that could be the most humanizing change of all.