Not Behind Enough to Panic. Too Afraid to Learn.

On December 26th, 2025 — the day after Christmas, when most of us were horizontal on a couch — Andrej Karpathy posted something that stopped me cold.

Karpathy is, by most reasonable accounts, one of the best programmers alive. He co-founded OpenAI. He led AI at Tesla. He built neural networks that became foundational to how the field works. If you were to make a list of people who should feel confident about where they stand in the age of AI, Karpathy would be first or second on that list.

Here's what he wrote:

"I've never felt this much behind as a programmer. The profession is being dramatically refactored... Clearly some powerful alien tool was handed around except it comes with no manual and everyone has to figure out how to hold it and operate it, while the resulting magnitude 9 earthquake is rocking the profession. Roll up your sleeves to not fall behind."

The post got 16.7 million views. And it resonated for a reason that has nothing to do with programming.

If Andrej Karpathy feels behind, what does that mean for the rest of us?

The American Management Association recently surveyed over 1,100 professionals across North America. 58% said they feel behind in keeping up with AI. Among individual contributors, it's 63%.

My honest reaction? The other 37-42% must be brain dead. (I'm mostly kidding. Mostly.)

Behind is the only rational response to what's happening right now. The tools are changing monthly. The capabilities are changing weekly. The expectations are changing daily. If you feel like you can't keep up, congratulations — you're paying attention.

But this is where I think a lot of people are making a critical mistake. And I almost made it myself.

There's a very good chance you're feeling the same way — what I’ll call the wrong kind of behind — and that distinction matters more than almost anything else about how the next twelve months go for you.

There Are Two Fears in the Room

1. Fear of Being Behind

This one is real. Structural. Universal. And — this is key — useful.

The fear of being behind is what Karpathy is describing. It's not existential dread. It's the recognition that there's a gap between where you are and where the tools could take you. That gap is real. It's big. Karpathy's word for it was "skill issue." In other words: the gap is closable. The behind-ness isn't a verdict. It's an invitation.

(If I'm being honest, sometimes the feeling is strongest right after I discover something new AI can do that I didn't know about. The more you learn, the more you realize how much you're missing. It's a feature, not a bug.)

This fear has a direction. It points somewhere. When you feel behind, the right move is to start. Pick one thing. Learn it. Build it. Run the experiment. The feeling of behind — held correctly — is fuel.

2. Fear of Replacement

This one is different. And this is the one that's actually causing the paralysis I see in organizations everywhere.

Fear of replacement doesn't point at a gap you can close — it points at a threat you can't neutralize. It sounds like: What if it's already better than me? What if I learn all of this and still get displaced? What if leaning in only accelerates my own obsolescence?

But this fear leads somewhere very different from behind-fear. It leads to paralysis. To fatalism. To a quiet disengagement that looks, from the outside, like calm — but is actually surrender.

And here's the part that really matters: this fear is often self-fulfilling. The person who believes they're going to be replaced stops experimenting. Stops learning. Stops being curious. And then, eventually, they do fall behind — not because AI replaced them, but because they replaced themselves. They opted out.

A good friend called me this week — a competent professional, a leader in his field, someone others regularly turn to for advice. He wanted what he called a "personal one-on-one consultation" about AI. He didn't want to post his questions to a group. He was too self-conscious.

And I realized: the feeling of being behind wasn't just uncomfortable for him. It felt like it said something about him. Like the gap was evidence of a personal failing, not a normal response to an unprecedented rate of change.

I found myself citing Karpathy to him. Because that's the reframe. If one of the most accomplished technologists alive feels behind, then behind isn't a character flaw. It's the weather. The question is just what you do about it.

And here's where the neuroscience stops being abstract and starts being genuinely alarming.

Why Threat Kills Curiosity

Here's what the research actually shows: your emotional starting point before an AI interaction determines how much you learn from it. Curiosity, it turns out, is not a personality trait. It's a neurological state with preconditions — and threat kills it before it starts.

Chantel Prat is a cognitive neuroscientist at the University of Washington who studies how individual brains differ in the way they learn. We recently interviewed her on Beyond the Prompt, where she walked us through what researchers Matthias Gruber and Charan Ranganath call the PACE model — the neurological sequence that has to fire for learning to actually stick.

The P is Prediction — you have to know that you don't know. You have to register the gap. (This is why experts sometimes learn less from a new domain: they assume they already know how it works.)

The A is Appraisal — your brain asks a safety question before it lets you get curious. Am I safe here? When curiosity fires, the brain actually turns down the volume on signals associated with risk. But it won't do that until it's assessed the threat level. If the threat is too high, curiosity never gets switched on.

The C is Curiosity — if prediction and appraisal are both satisfied, dopamine fires. And this is not just a feel-good chemical. Dopamine is a neuroplasticity catalyst. It's what allows the learning to stick. Without curiosity, you can be in the room, you can go through the motions, but the learning doesn't land.

The E is Exploration — which curiosity then intrinsically drives.

Here's what Chantel said that I can't stop thinking about:

"If you think you already know the answer, you will feel zero curiosity. And if you feel zero curiosity, your brain is not set up to learn. And that's just if you think you already know the answer — that doesn't have anything to do with if you feel insecure. If you think it could be better than you, if you think it could take your job, then it's not just not knowing the answer, but feeling some kind of threat. That would cause you to actively sort of create these defense mechanisms and move away from this new experience."

Read that slowly. The fear of replacement doesn't just make learning uncomfortable. It mechanically blocks it, at a neurological level. The brain, trying to protect you from the threat, actively suppresses the dopamine pathway that would otherwise allow you to learn your way through the situation.

Replacement-fear is literally self-fulfilling. It prevents the very learning that would address the underlying insecurity.

Even wilder? Chantel told us: "Given the same number of experiences, the same skills, the same background — if you feel curious instead of threatened, the benefit you'll get from each interaction with AI will be higher."

Same inputs. Different emotional starting point. Measurably different outcomes.

Why Organizations Keep Getting This Wrong

I know, I know — individual psychology is one thing. But here's the organizational implication, and it's the one I think leaders are almost universally missing.

After we recorded that episode, I sat down with Henrik — my co-host — and he said it plainly: "If organizations really knew that speaking about AI in a way that is threatening keeps their people from being curious, I think they would speak about it differently."

Therefore: every all-hands presentation, every change management communication, every enthusiastic demo from your most AI-forward leader — each of those is an appraisal event. Your people's brains are running a threat-assessment in real time. If the framing triggers replacement-fear rather than curiosity, the dopamine doesn't fire. And without the dopamine, the learning doesn't happen. You can mandate AI adoption all you want. You cannot mandate neuroplasticity.

Serious question: how does your organization talk about AI right now? Is the frame "here's what's possible" or "here's what's coming for you"? Because those are not the same frame. One passes the appraisal check. One fails it. The downstream learning outcomes are categorically different.

And here's the uncomfortable version of this question — the one worth sitting with if you're reading this newsletter: you're probably already enthusiastic about AI. Which means you're probably the most well-intentioned threat-trigger in the room. The more impressive your demo, the more fluent your prompts, the more it demonstrates what's possible — the higher the appraisal threat you may be generating in someone who isn't yet fluent.

We actually asked ourselves that, sitting there post-podcast: Is our enthusiasm scaring people or getting them curious? It's an uncomfortable question. But it's the right one.

What to Do When You Feel Behind

Feeling behind is not the problem. Staying behind is. Here's how to convert that fear into forward motion:

1. Name which fear you're actually carrying

This sounds simple but it's not. "I'm behind" and "I'm going to be replaced" feel almost identical at 2am. But they lead to completely different behaviors.

If you feel paralyzed — frozen, avoidant, quietly hoping it all plateaus — it's probably the replacement fear. That one needs to be challenged: "Is the gap really closed by not engaging? Or does disengaging guarantee the outcome I'm afraid of?"

If you feel restless — like you should be doing something but don't know what — it's probably the behind fear. That one needs to be channeled. The Karpathy reframe is useful here. If the world's best programmer calls it a skill issue, then it's a skill issue. Skills are closable.

2. Lower your appraisal threshold

You can't skip straight to the good part. The sequence is the sequence.

The PACE model says curiosity needs safety first. So manufacture safety. Don't start with the AI tool that will touch your most critical workflow. Start with something genuinely low-stakes — the one that helps you draft an email, plan a trip, brainstorm a title. Get the wins. The wins lower the perceived threat. Lower threat means the appraisal step passes. The appraisal step passing means curiosity fires. Curiosity firing means dopamine. Dopamine means learning actually lands.

Give yourself permission to be bad at this. Karpathy — the guy who helped build the AI revolution — described the experience as holding an alien tool with no manual. The standard isn't mastery. The standard is movement.

3. Have a real conversation with AI about your actual work

Not a demo. Not a tutorial. Not "write me a haiku about leadership." Open ChatGPT or Claude right now and paste this:

"I'm a [your role] at a [your type of company]. Here are the three things I spend most of my time on each week: [list them]. Before giving me any suggestions, interview me — ask me 3-5 questions about my work so you understand what would actually help."

That's it. That single prompt will teach you more about what AI can do for you than a dozen webinar recordings. (I mean it. Stop reading and try it. I'll wait.)

4. Find one other person who's also figuring it out

Not an expert. Not a guru. Just someone else who's behind and willing to say so out loud. The single fastest way to move from threat to curiosity is to realize you're not the only one fumbling around. Karpathy's post got 16.7 million views because millions of people felt the same thing and didn't know anyone else did. That's not a coincidence.

5. Set a small, recurring appointment with yourself

Fifteen minutes. Twice a week. Dedicated to trying something new with AI. You don't need a course. You don't need a certification. You need reps. The behind-ness doesn't shrink by thinking about it. It shrinks by doing something about it — even something embarrassingly small. (If you want a structured version of this, I built a free 15-day AI bootcamp that gives you one small challenge per day. It's designed for exactly this moment.)

The Only Question That Actually Matters

Andrej Karpathy is behind. Fifty-eight percent of the professional workforce says they're behind. I feel behind regularly — and I spend every single day immersed in this stuff. The more I learn, the more I realize how much I'm missing. I've decided to treat that as a feature.

The question isn't whether you're behind.

It's what you do when you realize you are.

One kind of response gets curious. The other gets defensive. One treats the gap as a problem to close. The other treats it as a threat to avoid. One generates dopamine and neuroplasticity and measurably better outcomes from every single AI interaction. The other builds defense mechanisms that move you further from the thing you actually need to learn.

Roll up your sleeves. Not because it's not scary. Because scared is not the same as stuck — unless you let it be.

Related: You're Not Tired of AI. You're Scared of It.
Related: It's a Skill, Not a Pill
Related: The Imagination Ceiling

Join over 33,147 creators & leaders who read Methods of the Masters each week.