You Can't Tell If You're Learning Anymore

Producing good work used to prove you were getting better. AI broke that link, and the signal you were relying on is gone.

Last week I published a piece called Everyone Is Back in School Now, and the conversations it started have been better than the essay itself. CEOs forwarded it to their kids, professors sent it to their deans, and parents sent it to other parents with a note that said something like “this is what I’ve been trying to say.” One reader, Tiago, asked the question I think matters most: when you actually try to rebuild learning inside a company, what does that look like day to day?

I’ll get to that next week. But I made a promise at the end of last week’s piece that I owe you first, because Tiago’s question sits on top of a harder one that nobody can answer for you:

If everything I used to struggle to learn is now available instantly, what’s actually worth learning?

That’s this week. The personal version of the question, not the company version.

The syllabus problem

I’ve been sitting with a syllabus for weeks. Fall semester at the College of Charleston, a room full of students, three months to spend on something that matters. Figuring out what to teach isn’t the hard part. Figuring out what to assign is.

The readings I put on the syllabus, they can summarize in seconds. The essays I ask for, they can generate a decent first pass in about ninety seconds. The case studies I hand out, they can walk through with a chatbot that will happily do the thinking I was hoping they’d do themselves. And I can’t really blame them, because if I were 21 and someone handed me a tool that could do all that, I’d use it too.

Handshake’s Class of 2026 research found that 58 percent of graduating seniors said they’d need stronger AI skills at work, but only 28 percent said their school had actually taught them any. So the message students are getting is: be careful with AI in class, then show up to your first job knowing how to use it well. That’s not a plan.

The problem I’m solving for my students is the same one they’ll spend their whole careers solving, and it’s the same one you’re solving right now whether you’ve named it or not.

Think about GPS

Here’s the version of this everyone has lived.

Remember learning your way around a new city before phones did it for you? You got lost, you paid attention, you built a map in your head or some of us remember paper maps. Now you follow the blue line and you arrive on time every single time. Better outcome, every trip. But turn the phone off and most of us, for sure me, couldn’t find our way back.

That’s the whole thing in one example. The tool made the trip better and made the driver worse, and the tradeoff was invisible because the trips kept working.

AI is doing that to thinking.

I know because of how I learned my own work. Twenty-something years ago I read the important books, asked people further down the road a lot of dumb questions, and took on problems I didn’t understand while the clock was running and the client was watching. It was slow and expensive and some of it was embarrassing, and it built me. If I started today I could get a version of all of it in an afternoon and sound like I understood the most of the whole field by dinner. I’d absorb faster and build myself less, and that’s the trade abundance is quietly offering everyone.

The abundance trap

The most common mistake I see, in students and executives both, is confusing taking things in with actually growing. I catch myself doing it too.

We’re all consuming more than any humans in history. More podcasts, more videos, more articles, more courses, more newsletters (yes, I hear the irony, and I’m writing this anyway). Almost none of it sticks. Ask someone what they read last week and they’ll struggle to tell you. Ask what they changed or action they took because of it and the pause gets longer.

Consumption isn’t capability. Struggle is.

By struggle I don’t mean suffering. I mean the specific discomfort of working on something just past what you can already do. You’re not sure you have it right, you read the same paragraph twice, you keep reaching for your phone. That feeling isn’t a side effect of learning, it’s the instrument you use to know it’s happening, because the edge of your ability is the only place capability actually changes.

Which is the part that should worry all of us. AI’s whole job is to make hard things feel easy, so when it takes the discomfort away it takes the signal away with it. Someone genuinely learning and someone just producing now feel the same on the inside. Both feel productive but only one is changing.

The basics haven’t changed and AI didn’t get rid of them. You learn something for good when you struggle at the edge of what you can do, get honest feedback from someone who knows more, live with the results of your choices, and try to teach it to somebody else. That’s been the whole list for a very long time. A researcher named Anders Ericsson spent a career proving it, and my mother spent hers doing it in a classroom without needing a study to tell her it worked, which is true of most good teachers I’ve known.

What changed is that all four of those used to be forced on you. If you wanted to know something you had to work for it, if you wanted to look good in front of your boss you had to defend your own reasoning, and if you got it wrong you couldn’t hide from what happened next. Teaching wasn’t optional either, because somebody newer needed help and you were the one standing there. None of that is forced anymore, which means you have to choose it on purpose or it just doesn’t happen.

You can watch a two-hour lecture and feel smarter, or you can spend those same two hours making something bad, having it torn apart, and fixing it. The wrong idea running around right now is that more access means more learning. More access means more exposure, and exposure without effort is how you end up knowing about everything and understanding almost nothing.

There’s a second thing happening underneath all of it. When the machine can produce competent-looking work on any subject in seconds, competent stops meaning anything. I wrote last week about taste, the ability to tell fluent from true and competent from excellent, and I’ve been thinking about it every day since. In a world flooded with output that all looks fine, taste is becoming the thing that separates people who can be trusted with a decision from people who can only produce something that resembles one.

The people in the middle

Most of this conversation drifts toward students and people starting out, which makes sense. But the hardest questions I get come from people ten, twenty, thirty years into careers they worked hard to build.

Learning when everyone expects you to be a beginner is one thing. Learning in public after years of being the expert is something else. Asking for help can feel like weakness, trying a new tool can make you worse before it makes you better, and giving up a familiar way of working can feel like giving up part of who you are. I’ve felt all three of those in the last six months.

Selma Bensalah left a comment on last week’s piece that named what I’d been circling. She wrote that passing things forward means “accepting that one day we will no longer be at the center.” The hardest work for experienced people isn’t picking up the new thing. It’s putting down the identity built around the old one, and no training program helps with that.

Companies make this worse without meaning to. They announce a new tool, offer a few courses, tell everyone to experiment, and then change nothing about workload, deadlines, or what gets rewarded. That’s not a learning plan. That’s more homework on top of a full day.

Experience still matters enormously. It gives you context, pattern recognition, and the kind of judgment that only comes from watching decisions play out over years. But experience can also trick you into reading a new world with old categories. Sometimes the work is learning, sometimes it’s unlearning, and mostly it’s knowing which parts of what you know still hold.

The future doesn’t ask us to throw away experience. It asks us to keep experience moving.

What’s actually worth learning

Here’s the promise I made last week, and here’s my honest attempt at keeping it.

When intelligence is everywhere, the things worth learning are the things it can’t hand you. Six keep showing up, in my classroom prep, in my client work, and in conversations with people genuinely wrestling with this instead of pretending they have it figured out.

Judgment. The machine gives you an answer. Judgment tells you whether the answer is any good, whether you asked the right question, and whether the thing should be done at all. As output gets cheap, this gets rare, because speed without discernment is just faster mistakes. You build judgment by deciding things and living with what happens, not by reading about other people’s decisions.

Curiosity. Sounds soft until you watch someone without it. Abundance without curiosity is just noise you scroll past. Curiosity turns access into pursuit, and it’s the difference between asking a machine to summarize something and asking it a question you actually needed answered. It’s also the one thing on this list nobody can install in you.

Seeing the whole picture. Almost every failure I’ve watched in five years was a systems failure dressed up as a technology problem. Somebody fixed one piece and broke three others they couldn’t see. AI makes this more dangerous because it makes fixing one piece effortless, so you can improve a single task at machine speed while quietly wrecking the thing that task was serving. Anyone who has fixed one leak and caused another knows exactly what I mean.

Knowing what you’re handing over. Every time you hand something off, you get speed and you give something up. Sometimes that trade is obviously worth it, and sometimes it quietly costs you a capability you were going to need. The skill isn’t knowing the tool. It’s noticing the trade while you’re making it, and being honest with yourself about which things you’re delegating because they free you up for better work and which ones you’re delegating because they were hard. The GPS didn’t make me a worse navigator by accident. I let it, one trip at a time, because each individual trip was easier that way.

Putting your own thinking into words. You don’t really know what you think until you try to say it. The struggle to find the right words is the thinking, and when the machine finds them for you, you skip the part where the understanding actually forms. This isn’t about being a good writer. It’s a nurse explaining a diagnosis to a frightened family, a parent answering a hard question from a kid, anyone standing up in a room and making a case. The fastest way to find out whether you understand something is to try explaining it to a person who doesn’t, without help.

The human subjects. I know how this sounds coming from a guy who spends most of his time in meetings with leaders. History shows us that every new technology arrives inside an old human story, philosophy teaches us to question the assumptions buried in the excitement, and literature lets us live inside a life that isn’t ours. AI can hand you an answer. These are how you become someone able to judge it.

Look at what those six have in common. Not one of them is something you can download, and all of them get built the same way, through trying and failing and being corrected over time.

How you build them

Three practices. Worth choosing on purpose, because nothing forces them anymore.

Try before you ask. Before you type the question, take five minutes and write down what you already think. Not to prove you don’t need help, but so the help lands somewhere that’s already been working. Ask after trying and you learn from the answer. Ask first and you learn that the machine is fast.

Teach what you’re learning. The second you have to explain something to another person, you find out whether you actually understood it. Write it down for a colleague. Say it out loud in a meeting. Put it somewhere people can push back. Understanding shows up when you deliver, and delivering means stepping onto the stage instead of rehearsing in the wings.

Work in front of people. Practicing alone with a chatbot builds one thing. Performing in front of colleagues, clients, or a classroom builds something else entirely.

AI can rehearse for you, but it still can’t step on the stage for you.

That line was commented on a lot last week, and I think it’s because everybody knows the difference between practicing and performing, even if we stopped talking about it.

None of this asks you to use AI less. It asks you to keep the work that builds you in your own hands, especially when the machine could do it faster.

Then take it beyond yourself. Ask someone in a different role or a different generation the hardest version of the question: where is AI making you more capable, and where is it just making you look more capable? That one belongs at a dinner as much as the boardroom, and it’s what this whole series is built around.

Stay a beginner

Back to the syllabus.

I’m going to assign things I couldn’t have imagined five years ago. Attempts before answers, leveraging your own reasoning written down before you touch a tool. Problems solved with the AI turned off and students teaching each other and defending what they think they know out loud.

I’m doing it because my education worked precisely because I couldn’t skip any of it. What I owe my students has to earn the same thing, in a world where every shortcut is free and sitting right there on their phone.

And here’s the part that isn’t about students at all. We are all our own dean now. We decide what we assign ourselves, what we try before asking, and what we keep doing the hard way even when the machine would be faster and probably better on the first pass.

I want to be careful with that line, though, because being your own dean is a great deal easier if you have time, a job that rewards it, and somebody in your life who models it. A lot of people have none of those, and telling them to self-direct is just one more way to feel behind. If you do have those things, some of what you owe is helping build them for someone who doesn’t.

The stuff that matters most is the stuff nobody writes down. Whether the silence in a meeting means agreement or confusion or quiet resistance. Whether the problem someone described is the problem they actually have. That doesn’t come from reading more. It comes from being in the room, getting it wrong, and learning to hear what nobody said. AI can summarize the meeting. It can’t tell you what the silence meant.

Capability is a practice, not a collection. It includes knowing what you can still do when the tool isn’t there, and the most valuable thing you can do for the rest of your life is stay a beginner on purpose.

Last week was about why everyone is back in school. This week is about the homework, and I don’t mean the kind someone assigns you. I mean the kind you assign yourself.

Tiago, I haven’t forgotten your question. Next week we go inside the organization and talk about what companies owe people when AI takes away the experiences that used to build them.

Let’s move together from Now to Next!

P.S. Same as last week, if anyone wants to talk about any of this, email me at jason@nowtonext.ai. I read every one, and the replies from last week are genuinely shaping where piece three goes. Much love.

Next in the series: Where the Hidden Curriculum Goes Now. What organizations owe their people when AI takes away the experiences that used to build them.

---

About Jason

Jason Averbook is the co-founder of Now to Next, an adjunct professor of business, and a globally recognized thought leader, advisor, and keynote speaker working at the intersection of AI, human potential, and the future of work. He spent the last few years as Senior Partner and Global Leader of Digital HR Strategy at Mercer, helping the world’s largest organizations reimagine how work gets done, not by implementing technology but by transforming the mindsets, skillsets, and cultures that have to come first.

Over the last two decades, Jason has advised hundreds of Fortune 1000 companies and their leaders, founded Knowledge Infusion and Leapgen, authored two books on the evolution of HR and workforce technology, and become a world renowned keynote speaker who has delivered hundreds of talks on the future of work. His work challenges leaders to stop treating digital transformation as an IT project and start treating it as a human strategy.

Through his Substack, Now to Next, Jason shares honest, provocative, and practical insights on what’s actually changing in the workplace, from generative AI to skills-based organizations to emotional fluency in leadership. His mission is simple: to help people and organizations move from noise to clarity, from fear to possibility, and from now to next.

You can reach him at jason@nowtonext.ai or connect on LinkedIn.