There is a piece of advice that gets handed to ambitious people with uncomfortable regularity.

Pick one thing. Go deep. Stop spreading yourself thin.

It sounds like wisdom. It has the ring of discipline. And for a specific kind of world — a stable one, with fixed rules and predictable rewards — it might even be correct.

But we do not live in that world. And we haven't for a while.

The problem with the advice.

Here is what nobody tells you about specialisation: the deeper you go in one direction, the harder it becomes to see anything coming from another.

Karim Lakhani, a professor at Harvard Business School, spent years studying how difficult problems actually get solved. He analysed hundreds of challenges posted on InnoCentive — a platform where organisations publish their hardest unsolved problems and invite anyone in the world to attempt a solution. What he found was counterintuitive: the further a problem was from a solver's own area of expertise, the more likely they were to solve it.

A chemist solved a molecular biology problem. A physicist unlocked a polymer engineering challenge. The outsider — the person who had no business being in the room — walked in, looked at the problem through a completely different lens, and saw what years of specialist immersion had made invisible to everyone else.

Lakhani concluded: "The further a challenge was from a solver's specialty, the greater the likelihood of success. That is very counterintuitive."

It is counterintuitive. But once you've lived it, it's also completely obvious.

What cross-pollination actually looks like.

Reed Hastings did not build Netflix because he understood the entertainment industry. He built it because he understood mathematics, computer science, software engineering, logistics, and the psychology of consumer frustration — and he saw how those things could be assembled into something the entertainment industry had never imagined for itself.

His background spanned mathematics at Bowdoin College, a master's in computer science at Stanford, the Peace Corps in Swaziland, and founding a software testing company before Netflix ever existed. None of those things look like a straight line toward building the world's most influential streaming platform. Every single one of them was necessary to build it.

This is what genuine cross-pollination looks like in practice. Not a dilettante who tries many things and commits to none. Someone who goes deep in multiple directions, holds the tension between them, and eventually sees a connection that nobody standing inside any single domain could have found.

Harvard Business School research on R&D teams has found that generalist researchers — those with knowledge across multiple domains — produce innovations with significantly higher impact when they succeed, precisely because breakthrough ideas rarely emerge from within single disciplines. The specialist optimises within the existing map. The person with range redraws it.

The thing nobody warned me about multiple interests.

Here is the part that took me longest to understand.

The shame of having too many interests is not really about focus. It's about legibility. A person with one clear skill is easy to categorise. Easy to hire. Easy to explain at a dinner party. A person who has spent serious time in multiple worlds — who has genuine depth in more than one domain — is harder to place. They don't fit the available boxes. And in a world that rewards quick categorisation, that can feel like a liability.

For a long time, I felt it as one.

I started my career in IT and enterprise systems — the kind of work that keeps infrastructure running at 3am and teaches you, quietly but indelibly, how complex systems behave under pressure. Then photography and cinematography. Then camera technology at RED Digital Cinema, then virtual production at Disguise, then production pipelines at Amazon and Netflix, and now generative AI at scale. Each time I moved, someone somewhere raised an eyebrow. Where exactly are you going? What are you specialising in?

What I was actually doing — though I couldn't have articulated it clearly at the time — was building a compound asset. Each domain made the previous ones more useful. Understanding how creative people think made me a better technologist. Understanding technology at depth made me more genuinely useful to creative people. Understanding production systems made the AI work more grounded. Understanding AI made the production systems more imaginative.

None of these interests competed with each other. They were in conversation — constantly, productively, in ways that continue to surprise me.

On the fear of being half of everything.

The standard objection to all of this is reasonable: doesn't spreading across multiple domains make you mediocre in each?

Sometimes. For a while. Breadth without depth is just distraction.

But the research — and the lived experience of people who have navigated this well — suggests something more nuanced. Studies comparing career-focused and broader education across twelve countries found that while specialists were more likely to be hired immediately and started with higher incomes, their growth rates were slower and less adaptable over time. When shocks hit their industry, they were more likely to go backward or end up out of work entirely.

The specialist wins early. The person with range wins over time.

And in an era where AI is compressing the timeline on early-career specialist work — the analysis, the research, the entry-level tasks that used to take years to master — the value of being the person who can see across domains, who can connect things that don't usually sit in the same room, is increasing faster than almost any other professional asset.

Naval Ravikant puts it simply: "Specific knowledge is found by pursuing your genuine curiosity and passion rather than whatever is hot right now." The path that looked scattered was actually the pursuit of genuine curiosity. And genuine curiosity, compounded over enough time and across enough domains, becomes something that cannot be replicated by someone who stayed in one lane.

What I'd tell anyone who's been told to focus.

The question is not whether to go deep or go broad. It's whether your interests are compounding or just coexisting.

Compounding interests talk to each other. They make each other sharper. They create a vantage point that no single domain could provide.

Coexisting interests are just hobbies that compete for your time.

The difference is not in how many interests you have. It's in whether you actively look for the connections — whether you bring what you know from one world into the problems of another, whether you let the friction between different ways of thinking produce something new.

That is the work. Not the acquisition of interests, but the integration of them.

I have been lucky enough to spend my career at an intersection that almost nobody else was standing at — IT and enterprise systems, photography, cinematography, camera technology, virtual production, generative AI. Each move looked sideways. Collectively, they built something that I could not have designed deliberately and would not trade for the straightest possible line.

The advice to pick one thing is well-meaning. It comes from a world that valued predictability over possibility.

We live in a different world now.

Pick the things that won't leave you alone. Go deep enough in each to have real standing. Then pay close attention to what they say to each other.

That conversation is where the best work comes from.