The Problem with Education

A recent article noted that roughly 13 percent of today’s high school and college graduates can’t get a job—not because there aren’t any, but because those jobs are being lost to artificial intelligence. For young people with no work history, this is a hard truth. Finding that first paycheck has always been difficult; now it’s even more so, because the very roles that once gave people a foothold are being devoured by machines.

It’s tempting to blame AI for this problem, but the truth is that education itself bears much of the responsibility.

The Broken Assumptions of Education

The public education system—whether in the United States or elsewhere—rests on a fragile foundation: the assumption that memorizing facts equals learning. Tests and quizzes reinforce this assumption, and standardized exams like the SAT enshrine it. But memorization has never been the heart of human learning.

We all know this intuitively. Babies don’t study textbooks before walking. They watch, they attempt, they fail, they try again, and eventually they succeed. At no point does a toddler sit down for a multiple-choice quiz on locomotion. The “test” is simply whether or not they can walk.

And yet, schools persist in equating fact recall with education. This leads to two predictable outcomes. First, it wastes enormous time. People memorize long enough to pass the test and then forget what they studied. Second, it sets up direct competition between humans and machines. AI models can memorize better than we can, faster, cheaper, and with far fewer mistakes. They can also perform repetitive tasks tirelessly. These just happen to be the entry-level jobs that once absorbed young graduates—and now, increasingly, they’re gone.

The Human Advantage

So what’s left? Where do humans still shine?

In the simple but profound act of teaching and learning by doing.

Real teaching is not the delivery of facts. It’s the process of showing someone how to do something, giving them pointers, and letting them practice until they improve. Progress comes through repetition, through doing it badly until you do it better, and eventually mastering it.

Take archery as an example. You could sit in a classroom for an entire semester and learn the vocabulary: bowstring, nock an arrow, draw length. You could pass a test on those terms. But ten minutes on a range with a bow in hand—fumbling with your first arrow, listening to an instructor say, “Try holding it this way”—teaches you more than months of lecture. And the irony is that even in that brief practice session, you’ll naturally memorize the terms anyway. You’ll remember “nock the arrow” because you’ll be doing it, not because you underlined a definition in a textbook.

That is the efficiency of experiential learning. You don’t need to force memorization—it happens as a byproduct of practice.

Why AI Can’t Replace This

AI can give you information about archery or carpentry or design. It can describe the steps, generate instructions, even create practice drills. What it can’t do—at least not yet—is watch you in real time, see where you stumble, and provide the exact feedback that propels you forward.

That’s the human edge. It’s not in fact retention or busywork. It’s in the ability to notice, to adapt, to offer encouragement, and to guide someone through mistakes until they improve.

The Irony of Online Courses

Given that truth, you’d think online education would innovate. Unfortunately, most platforms have doubled down on the broken model. They’ve digitized the very things that don’t work: worksheets, quizzes, exams, and grading systems. The underlying assumption hasn’t changed—if students can regurgitate facts, they must have learned.

The result? An experience that feels eerily similar to school: a lot of material, but little transformation. Students may finish with a folder full of PDFs, but rarely with the skills they actually wanted.

A Different Approach

When I built TeachKit, I wanted to go in the opposite direction. The question I asked was: how do people actually learn? The answer was clear. They learn by seeing, doing, and repeating—by practicing until they improve.

So TeachKit is structured around that principle. Every lesson includes action items. Every download is a guide or a template, not a worksheet or quiz. Videos aren’t there to passively inform; they’re there to show, so students can try.

The goal is transformation. Not more material. Not more “stuff” to pad the course. What students pay for—what they hope for—is to come out changed. To be able to do something they couldn’t before. The faster and more effectively you can help someone make that leap, the more valuable your course becomes.

That’s what TeachKit is built for.

Toward Real Learning

My hope is that courses built on TeachKit will allow students to learn in weeks what might otherwise take months or even a year in a traditional classroom. I know it’s possible because I’ve done it myself. When I learned traditional archery, I didn’t need semesters of memorization. I needed practice, feedback, and repetition.

AI may be closing doors in the job market, but it has also thrown light on the failures of education’s obsession with memorization. The future belongs to those who can teach and learn the way humans always have: not by rote, but by practice, by guidance, and by transformation.