why mastery still matters in the age of AI

Everyone thinks AI is going to replace us.

Experts predict mass job displacement. We see AI writing copy, coding apps, creating content—and suddenly everyone's wondering if their skills even matter anymore.

We see these CEOs saying most tasks are going to be done by AI. But I think we're reading this the wrong way. It's not so much that AI is replacing jobs—it's changing how we approach our work and creations.

I should know. I was a copywriter when ChatGPT launched. People assumed my job was toast.

But something funny happened instead.

when AI "replaced" my copywriting

When I first started using AI for copywriting, the output was terrible.

Generic, soulless, obviously AI-generated garbage. You could smell it from miles away.

But as the models improved and I learned to work with them, something interesting emerged.

I began teaching AI to think like a copywriter.

I'd feed it frameworks, mental models, examples of what worked and what didn't. The copy started getting good.

Suddenly I could deliver more emails to clients, work with more people, scale my solo freelance operation. The productivity gains were real.

At first I was feeling guilty. “Why are my clients paying me?” They could do this themselves.

In fact, they knew I was using AI in my workflows and were completely fine with it.

But as I thought deeply about it, I realized they couldn't. Or sure they could try, but they would get very different results.

It hit me. My clients weren't paying me to use AI. They were paying me to know what good copy looks like.

Take the average person using AI versus someone who's spent years learning copywriting, reading great copy, understanding what works. When we both use the same AI tool, the outputs are completely different.

The untrained user might go from creating 10% quality copy to 60% with AI. Not bad. Someone who learns prompting and uses templates might hit 80%. But that last 20%—the difference between good and great—requires deep understanding of the craft.

Your ability to recognize quality, guide the AI's thinking, and know what's missing when something doesn't work. That's what matters.

the vibe coding breakthrough

Last year I started what I call "vibe coding"—building apps using AI tools like Bolt.new, Lovable, V0, Cursor AI, Windsurf etc.

My goal was creating internal software for my company and experimenting with my own ideas.

I had basic coding experience from my school days when learning data science, but as I’ve had my early career in marketing, I wouldn’t consider myself technical.

Now, from my experience vibe coding, while I was able to end up creating these working prototypes and micro software tools, it was through alot of pain and persistance. I spent 80% of my time was spent basically scolding and cursing the AI, trying to get it to move a red box or fix some basic functionality.

As such, the biggest realization for me was that the more I vibe coded, the more I wanted to actually learn programming.

Not because AI couldn't code, but because I lacked the vocabulary to communicate with it effectively.

When you don't understand the fundamentals of what you're trying to build, you can't identify root cause problems. You can't explain to the AI what's actually wrong or what you really need. An expert developer who's been through the trenches can immediately diagnose issues, recognized patterns, and communicate them to AI in technical detail, getting results 10x faster than someone just winging it.

The revelation was that understanding the fundamentals would make me exponentially better at using AI.

the new skill requirement

We're not heading toward a world where AI replaces knowledge workers. We're heading toward a world where everyone is augmented by AI.

This means the bar is shifting higher. AI tools will become as standard as email or spreadsheets. The people who survive and thrive will be those who can effectively collaborate with these tools.

But effective collaboration requires understanding. You need the right vocabulary, pattern recognition, and domain expertise to guide AI toward quality outputs.

When you've written copy from scratch before and for long enough time, you can immediately spot when AI is using weak emotional triggers or hitting the wrong market sophistication level. When you've debugged code without assistance, you can tell AI "this is a state management problem, not a styling issue" instead of spending hours on random fixes.

The future of programming is through words, but you can't communicate complex technical concepts without understanding system architecture, data flow, and why certain approaches scale. The future of copywriting is AI-assisted, but you can't recognize great copy without understanding market awareness, emotional triggers, and what makes something "feel" right.

why I still practice

This is why I spend an hour and a half every day practicing coding or writing—without AI doing the work for me, but with AI helping me learn faster.

Naval Ravikant once said, "Learn to build, learn to sell, and you'll be unstoppable." That's exactly why I chose these two fields to focus on. I've spent the past six years learning how to sell through copywriting and marketing. Now I'm learning how to build through coding.

I'm not trying to become the world's best programmer or writer but well enough to be able to get the first versions of my ideas out. And more so to build deep enough skills to be effective in an AI-augmented world. I'm chasing mastery even as I depend heavily on AI tools.

As an entrepreneur, I see myself as a generalist who can do both reasonably well. Eventually, I'll hire specialists to handle the heavy lifting—and I hope they use AI too. But I want to understand the fundamentals well enough to communicate effectively with those specialists and guide the vision.

Because here's what I believe. The people who are actually great at what they do will never be replaced by AI. Their output will just be tremendously accelerated.

AI hasn't eliminated the need for expertise—it's made expertise more valuable than ever. The gap between someone with deep skills using AI and someone just prompting without understanding is massive and growing.

the symbiosis

The most productive people in the AI age won't be those who abandon skill development or those who ignore AI entirely. They'll be the ones who develop deep craft knowledge specifically to enhance their AI collaboration.

Think of building your internal compass. When you deeply understand your domain, you can quickly spot when AI is going off track, guide it back to quality, and push it toward outputs you could never achieve alone.

The future belongs to humans who can think clearly about complex problems and communicate those thoughts effectively to powerful tools.

That's why I'm still learning. That's why mastery still matters.

The machines aren't coming to replace us. They're coming to amplify us.

But only if we're worth amplifying.