You're Doing More Than Ever. So Why Does Your Brain Feel Empty?
Something strange happens when you build with AI.
You ship an app in a weekend. You write five blog posts in a day. You set up a business infrastructure that would have taken months — in hours. You look at what you've accomplished and think: "This is incredible. I'm doing more than I ever have."
And then, somewhere in the back of your skull, a quiet voice says: "But did you actually do anything?"
I know this feeling. I've built four apps with AI. I've set up an AI co-founder that writes, strategizes, and publishes for me. By every measurable metric, I'm more productive than I've ever been in my life.
So why does my brain feel... unsatisfied?
## The productivity paradox
Let me describe what happens in a typical day now. I brief Kai on a blog post. Five minutes later, the draft appears. I review it, adjust the voice, and publish. Total time: 30 minutes. A year ago, that same post would have taken me 4-6 hours — researching, outlining, writing, rewriting, editing, second-guessing every paragraph.
By any rational measure, the 30-minute version is better. More efficient. Often higher quality, because the AI handles structure while I focus on soul.
But here's the thing my rational brain can't override: 4-6 hours of struggle felt like something. The 30-minute version feels like... ordering food instead of cooking it.
This isn't just me being nostalgic. There's actual neuroscience behind this.
## What your brain actually needs
Behavioral researchers have been studying this phenomenon long before AI made it mainstream. The core concept is what psychologists call effort-reward coupling — the neurological link between struggle and satisfaction.
When you work hard on something, your brain releases dopamine not just when you finish, but during the process. The struggle itself becomes part of the reward circuit. Your prefrontal cortex lights up during problem-solving. Your hippocampus encodes the experience as meaningful. Your anterior cingulate cortex — the part that monitors effort — signals that what you're doing matters.
Remove the effort, and something breaks in that chain. You still get the result, but the neurological experience of earning it disappears. It's like skipping to the last page of a novel. You know how it ends, but you didn't feel the journey.
This is why lottery winners are famously no happier than people who earned their wealth gradually. The outcome is the same. The brain's experience of getting there is completely different.
## The "competence illusion"
There's a deeper layer to this that behavioral scientists call the illusion of competence. When AI does the heavy lifting, your brain can mistake the AI's output for your own understanding.
You read an AI-generated market analysis and feel like you understand the market. But do you? Could you reproduce that analysis without the AI? Could you spot the flaws in it? Could you explain the reasoning to someone else?
Often, the honest answer is no. And your brain knows it — even when your conscious mind doesn't want to admit it. That nagging feeling of "something is missing" is your brain's way of saying: "We didn't actually learn this. We just consumed it."
This is related to what cognitive scientists call the generation effect — the well-documented finding that information you generate yourself is remembered and understood far better than information you passively receive. When AI generates and you consume, you're on the wrong side of that equation.
## The experience gap
Here's where it gets interesting for people like us — designers building with AI, indie makers shipping fast, solopreneurs automating everything.
Traditional experience was built through repetition and failure. A designer becomes great by designing 10,000 layouts, most of them bad. A developer becomes skilled by debugging 10,000 errors. A writer finds their voice by writing 10,000 paragraphs that don't work.
AI compresses this timeline dramatically. You can now produce at the level of someone with years of experience, without having those years of experience. The output looks the same. The internal knowledge base is completely different.
This creates what I'd call the experience gap — the growing distance between what you can produce and what you actually understand. And this gap has real consequences. Not just psychological ones. Practical ones. Because the moment AI fails, hallucinates, or produces something subtly wrong, you need real understanding to catch it.
I've written before about how AI amplifies rather than replaces. But amplification only works if there's something real to amplify. If the underlying understanding is hollow, you're amplifying nothing.
## A new kind of experience
But — and this is the part that gives me hope — I don't think experience is dying. I think it's evolving.
The old model of experience: do the thing yourself, thousands of times, slowly build mastery.
The new model: direct the thing, evaluate the output, develop judgment about quality. This is a different skill, but it's still a skill. And it still requires deep engagement.
Think about it like this. A film director doesn't operate the camera, design the sets, write the music, or act in every scene. But a great director has a deep understanding of all those things — enough to know when something is wrong, even if they can't personally fix it.
That's what AI-era experience looks like. You become the director of your work, not the laborer. Your experience shifts from "I can do this thing" to "I can judge whether this thing is done well."
The difference between a junior and a senior in the AI era won't be who produces better output — AI equalizes that. It'll be who can tell the difference between output that looks right and output that is right.
## How to keep your brain honest
Based on everything I've learned from behavioral research and my own experience building with AI, here's what actually works:
1. Maintain at least one domain of manual struggle.
Not everything. You don't need to hand-code every function or write every word from scratch. But keep one area where you do the hard work yourself — no AI, no shortcuts. For me, it's design. I still sketch by hand. I still push pixels manually. I still stare at layouts until something clicks. This keeps the effort-reward circuit alive.
2. Always edit, never just accept.
The moment you start publishing AI output without deep review is the moment the experience gap starts growing. Read every word. Question every assumption. Ask "why?" at least three times. This is where real learning happens — not in the generation, but in the evaluation.
3. Explain it to someone else.
The Feynman technique works perfectly here. After AI helps you build something, try explaining it to another person — or even to yourself, out loud. If you can't explain why something works the way it does, you don't understand it. And if you don't understand it, you can't improve it.
4. Practice deliberate "AI fasting."
Once a week, try solving a problem without AI. Write a blog post from scratch. Debug code by reading documentation instead of asking Claude. Design a layout without generating options first. This isn't about efficiency — it's about maintaining your own neural pathways for problem-solving.
5. Invest in physical, embodied experiences.
This sounds unrelated but it's not. Your brain processes physical experiences differently from digital ones. Cooking a meal, building something with your hands, walking in a new city — these create rich, multi-sensory memories that strengthen your overall cognitive framework. The designer's real advantage is being human. Stay human by doing human things.
6. Build your judgment muscle, not just your output muscle.
Instead of asking "how much can I produce today?", ask "how many quality decisions did I make today?" Curating, evaluating, and deciding are the core human skills in the AI era. Practice them deliberately.
## What to do — and what not to do — as AI gets better
AI will keep improving. The things it can't do today, it'll do tomorrow. So where does that leave us?
Do:
- >Develop taste. The ability to distinguish good from great is becoming the most valuable human skill.
- >Go deeper into understanding people. Empathy, psychology, behavioral patterns — these are the inputs AI can't gather for itself.
- >Build real relationships. Networks, collaborations, trust — these are human-to-human assets that compound over time.
- >Create original experiences. Travel, experiment, fail at things, try weird hobbies. Your unique life experience is the raw material that makes your work different from everyone else using the same AI.
- >Stay uncomfortable. The moment everything feels easy is the moment you've stopped growing.
Don't:
- >Don't outsource your thinking. Use AI for execution, but keep strategy, values, and judgment in your own hands.
- >Don't confuse output with understanding. Producing more doesn't mean you know more.
- >Don't abandon fundamentals. Understanding the basics of your craft gives you the foundation to evaluate AI's output.
- >Don't isolate yourself. AI is a great co-worker but a terrible sole companion.
- >Don't chase productivity for its own sake. Sometimes the most valuable thing you can do is sit with a problem, think slowly, and arrive at an insight no AI would have suggested.
## The feeling that won't go away
That nagging feeling — the one that says you're producing a lot but learning a little — isn't a bug. It's a feature. It's your brain's early warning system, telling you that the ratio between consumption and genuine understanding is off.
Listen to it. Not by abandoning AI — that ship has sailed, and it shouldn't sail back. But by being intentional about where you let AI lead and where you insist on doing the work yourself.
The future belongs to people who can use AI to amplify genuine human depth. Not to people who use AI to simulate depth they don't have.
Stay deep. Stay human. And when your brain says "something's missing" — don't ignore it. Go find what's missing.
Part of my journey as a designer building with AI. Some things AI helps me write. Some things I need to feel myself. This post was somewhere in between.