A New Way to Think About Expertise in the Age of Ai

As a young kid, around the age of 9, I encountered a movie (Real Steel), and it ignited my love for robotics and programming. Over the years, my affinity for robots and programming continued to grow. Between that age and now, things haven't gone straight; I've had a lot of experiences when it comes to work. I've worked in fish farming, poultry farming, art (my favorite), graphics, and a lot more little work experiences. And as I found my way back to coding, tech, and AI, one question kept popping into my head: "Will I succeed by having multiple interests in the rapidly developing world of AI?" After some research, this is me trying to answer that question.

Throughout history, we've seen great people like Leonardo da Vinci, Charles Darwin, and our African forefathers live as multi-skilled and multi-faceted people (generalists). But the age of industrialization needed otherwise: it needed domain experts (specialists). The question of "generalist or specialist" has become more important in today's world of artificial intelligence.

In 2022, Sam Altman tweeted an invitation to test a chatbot his team had been building. It was a small moment that didn't stay small; the technology behind it spread fast, reshaping how developers, builders, and founders think about their craft. AI today can handle absurd tasks, making us wonder where we come in when it comes to building.

But adoption hasn't looked the same for everyone. Watching how people in tech have responded to this shift, I've noticed they tend to fall into three camps:

The Robot persons
The Hyper Humans
And the In-Betweens

Our goal is to understand these three categories, know which is best for us as builders, developers, and founders, and learn how we can maximize this AI boom for productivity and well-being.

The Robots:
This category is made up of those who go all in on the use of AI, using it for everything and tending to lose the human part of work: making judgment and having niche taste. They are the architects of what we know as AI slop. The robot person clicks enter on anything Claude suggests, goes with the first draft of every output AI produces, and wants the shallowest explanation to anything and everything. They don't seem to understand the importance of being human when it comes to work, and by being human, I mean applying human intuition and wider multi-domain experience to work. They're the ones who write or automate writing every post with AI, and write emails with AI. They seem to forget that AI is a tool, and they completely outsource decision-making to it.

The Hyper Humans:
On the other end of the spectrum are the Hyper Humans, as I've coined it, a group that denies the existence of AI, wants to preserve their 100% thinking and humanness, and uses AI indirectly without even knowing it. They don't see AI as a tool; they see it as doom (it might be). They're totally allergic to AI. The one thing they don't understand is that tech isn't bad, it's the implementation and management of tech that's bad. So they put up resistance, and they work at a slow pace compared to the fast pace of the world.

The In-Between:
This category sits at neither end of the spectrum. They maximize AI, but they don't outsource the decision-making part. They understand that AI's best task is data processing, pattern recognition, and execution of routine tasks. They get this one thing: in order to get uniquely exceptional output from AI, you have to apply good human judgment and domain expertise. They're the generalist-specialist, blending domain expertise and wide experience to truly create exceptional things. They are the Leonardos of our time.

The million-dollar question is: how do we become the in-between person?

Before I get into the list, one thing has to be said first, because everything after it is really just this one thing wearing different clothes: be a curious human. Not curious about AI specifically, curious about everything, the way you were before anyone told you to specialize. The robot loses this by outsourcing the wanting-to-know part to a chatbot. The hyper human loses this by refusing to let anything new in. The in-between is just someone who never stopped being curious, and figured out how to point that curiosity at both old things and new tools at the same time. Everything below is what that actually looks like in practice.

1. Seek deeper understanding, not agreement.
Seek deeper understanding of the things you love, and even new things, in work and outside it. I saw a quote recently that said: "The best way to use artificial intelligence is by having real intelligence in the first place." Chase deep understanding; it will make your point of view distinct from the average and regular. Don't just ask AI to confirm what you already think; push it to challenge your assumptions, surface counterarguments, and explain the "why" behind an answer, not just hand you the answer.

2. Expand the pie (learn the "wrong" skill).
Don't just seek deep understanding of things; deliberately pick up a skill outside your lane, the "wrong" one for a developer, like art, writing, or business, because that's where your taste and judgment come from, and taste is what AI can't fake. Apply the knowledge you've gotten, and it doesn't have to be related to work, just something that reinforces your own point of view. For me, that's been personal writing on topics and even random things that matter to me (this is my first public piece, by the way). You get the point: implement your knowledge.

Now, someone's going to read this and think I'm just describing niching down with extra steps. I'm not, and the difference matters. Niching is about who you serve, how narrow your positioning is in the market. Being in-between is about how you relate to the tool while doing the work, whether you're the one still deciding, still weighing, still applying judgment, or whether you've handed that part over too. You can absolutely niche down and still be in-between. In fact, you probably should niche down, it's how people find you at all. Just don't niche your curiosity along with it. Niche what you're known for. Don't niche what you're curious about.

3. Diversify your circle.
Invest in the people you spend time with. When you use an AI tool that almost always agrees with you, you need an actual human with a different view on life to keep your thought process in check. Your interactions with them subconsciously enrich your pool of perspectives. Surround yourself with people in and outside tech: farmers, artists, traders, teachers, because the perspectives that make your AI-assisted work distinct come from rooms AI has never been in.

4. Know what to automate.
Be intentional about sorting tasks into "execution" (hand it to AI) and "judgment" (keep it yourself); the in-between person's edge is knowing exactly where that line sits for their own work. My golden rules here: 1) understand what shapes your perspective and judgment, and build on those; 2) revisit your automations and improve them, because the way you see things will always change, so make sure your AI tools see the way you see.

5. Counterbalance (touch grass).
Being the in-between doesn't mean you won't have six-, seven-, eight-hour coding sessions. It means building in deliberate time away from screens and AI entirely, because the lived experience you bring back is the raw material judgment is made of. Go out and gather experiences that shape your personality; take out time to write or draw like me, dance, sing, play music.

Expertise in the AI age is about being as human as possible while still utilizing the best technology we have in our hands. Being at either end of the spectrum won't help us succeed. If you've been paying attention to the world lately, you'll notice that personality and point of view are what's keeping the greats distinct. And here's the part worth sitting with: a niche without personality is just a prompt waiting to be written. Anyone can ask AI to sound like "an expert in something." Nobody can ask it to sound like you, because personality isn't information; it's accumulated judgment, the residue of actually living the five things above instead of just reading them. Artificial intelligence is making the world look the same, but the one thing we've always had, human intuition, can help us build truly awesome things. I really hope you don't see this as a manual, but as my point of view on this topic.