The Theorist and The Builder

I was reading Walter Isaacson's Steve Jobs biography a few weeks ago, and something kept nagging at me. Everyone knows how Jobs saw the future—the iPod, the iPhone, the vision of technology as something beautiful and intuitive. But here's the thing: Jobs didn't build any of it. Not really.

In the beginning, it was Steve Wozniak who actually built the hardware. Wozniak had something working, and Jobs marketed it, sold it, made people believe in it. Jobs himself said he couldn't have done it without Woz. He needed someone who could turn his ideas into actual, functioning products.

As time went on, Jobs kept doing what he did best—having ideas, pushing engineers, and just knowing when something was right. He didn't code. He didn't solder circuits. He saw what needed to exist and made sure it got built.

The Divide

There's this tension in tech that's always bothered me. Hardcore engineers look at people like Jobs and criticize the way he worked. "He didn't even code anything," they say. And on the flip side, you have product people and strategists who think engineers just make things too complicated, who get lost in the details when they should be thinking about the bigger picture.

I think both sides are wrong. Or rather, they're missing the point entirely.

The Physics Parallel

This is where my mind went: think about the Manhattan Project. You had theoretical physicists—people doing the math, coming up with theories, understanding abstract concepts, debating ideas. Were they building actual experiments? No. That's literally why they're called theoretical physicists.

But you also had the experimental physicists and engineers who built the actual tests, who made the theories real and tangible.

Neither could exist without the other.

The theorists needed the builders to prove their ideas worked. The builders needed the theorists to know what to build and why. This isn't unique to physics—it's everywhere. The visionary and the executor. The architect and the construction worker. Steve Jobs and Steve Wozniak.

That gap between theory and execution? It's always been there. And until recently, you needed both types of people to build anything meaningful.

Something Changed

Here's what I've been thinking about: AI is changing this dynamic in a fundamental way.

For the first time, you can be both. You can be the theorist and the builder, all by yourself.

The AI we have access to today knows most technical concepts. It knows general ideas that have been explored before. It won't understand things that haven't happened yet—it's not magic—but it can help you think through everything that's already been figured out. And more importantly, it can help you actually build the technical thing, even if you're not technical yourself.

The gap is shrinking.

It still exists—don't get me wrong. Large, successful teams still need people who respect each other's expertise. But for someone starting from scratch? Someone in their garage, their home office, working on an idea at night after their day job?

That person doesn't need to wait to find their Wozniak anymore.

Why This Matters for Hardware

Software has had a ridiculously low barrier to entry for years now. You can learn to code online for free, spin up a website in an afternoon, deploy an app without owning a server. It's been democratized.

Hardware? Not so much. Hardware has always been expensive, complicated, and unforgiving. You buy the wrong components, you've wasted money. You don't know what you need, so you over-buy or under-buy. The iteration cycle is slow and costly.

But I think that's changing fast.

With AI, you can figure out exactly what parts you need for your idea before spending a dime. You can test your theories, validate your approach, and research components without the traditional waste. The barrier to entry in hardware is dropping in the same way it did for software over the past decade.

My Own Test

I'm testing this theory myself right now. A few weeks ago, I went from building zero hardware projects to building a device you can actually have a conversation with. It connects to LLM APIs, does internet searches, plays audio through a speaker—the whole thing. A normal, functional conversational AI device.

It took about two and a half to three weeks. A lot of hours, nights and weekends, day after day. Not going to sugarcoat it—it was hard work.

But here's the thing: I didn't have to buy much hardware I didn't end up using. I was able to research ahead of time and figure out exactly what I needed. As new ideas came up, I could add things without starting over. The waste was minimal. The learning curve, while steep, wasn't insurmountable.

A few years ago? This would have taken me months, maybe years. I would have needed a team, or at minimum, a technical co-founder. Now I can do it alone, in my spare time, in a few weeks.

What Comes Next

I'm excited about what this means for the future.

If people can build hardware projects as easily as they've been building quick software projects, we're going to see an explosion of innovation. Not from big companies with massive R&D budgets. From individuals. From garages. From anywhere and everywhere in the world.

We're going to have an abundance of better consumer products. Professional-quality products built by people working from their homes, not just manufactured there, but designed and prototyped there too.

The future isn't about choosing whether you're a theorist or a builder, a Jobs or a Wozniak. It's about being able to do both, and seeing what happens when that becomes possible for millions of people who previously couldn't.

I don't know exactly what that world looks like, but I'm excited to find out.

And I'm definitely going to keep building things in my home office to see how far this goes.