Specialization Is for Insects

"A human being should be able to change a diaper, plan an invasion, butcher a hog, conn a ship, design a building, write a sonnet, balance accounts, build a wall, set a bone, comfort the dying, take orders, give orders, cooperate, act alone, solve equations, analyze a new problem, pitch manure, program a computer, cook a tasty meal, fight efficiently, die gallantly. Specialization is for insects."

Heinlein wrote that in 1973. It's been the rallying cry of generalists ever since. And for fifty years, they've been wrong about what it means.

The problem isn't that Heinlein was wrong. It's that everyone who quotes him misses the point. They think he's telling them to avoid specialization. He's not. He's telling them something much more interesting: that the choice between specialist and generalist is false.

Think about what Heinlein actually listed. Programming a computer. Building a wall. Writing a sonnet. These aren't casual skills you pick up in a weekend. Each one takes years to do well. What kind of person knows how to do all of them? Not a generalist in the weak sense—someone who knows a little about everything. Someone who has mastered multiple specialties.

This distinction matters more now than ever. Because AI is about to make the weak generalist obsolete.

The T-Shaped Illusion

Marketers love talking about being "T-shaped." Broad knowledge across many areas, deep expertise in one. It's a comforting model. It suggests you can have it both ways—be both generalist and specialist.

I spent twenty years believing this. Learn a little about everything in marketing. Go deep on paid acquisition. Perfect. I was T-shaped.

But there's something unsatisfying about the T-shaped model. It assumes the world stays still while you carefully construct your T. It doesn't. And more importantly, it assumes that broad and deep are equally valuable. They're not.

Here's what I learned: The broad part of the T is table stakes. Everyone has it. Or at least, everyone can fake it well enough with Google and now ChatGPT. The deep part is what matters. But even that's not quite right.

The 95/5 Principle

The Pareto principle says 80% of effects come from 20% of causes. Everyone knows this. What fewer people realize is that you can apply it recursively. 80% of the 80% comes from 20% of the 20%. Do the math: 64% comes from 4%.

But why stop there? Apply it once more. You get something like 95% of value from 5% of effort—but only if you pick the right 5%.

This changes everything about how you think about specialization.

You don't need to know everything about everything. You don't even need to know a lot about everything. You need to know just enough about most things to recognize patterns, and then go frighteningly deep on the few things that actually matter.

In marketing, that might mean understanding the basics of copywriting, design, analytics, and brand strategy—but becoming world-class at paid acquisition and marketing automation. Not just good. Not just very good. So good that you see patterns others don't even know exist.

Why AI Changes the Game

Here's the counterintuitive part: AI makes specialization both more and less important.

Less important because AI can handle the routine parts of any specialty. A good AI can write decent copy, create basic designs, analyze data, and even code simple programs. The shallow specialist—someone who's gone deep but not deep enough—is in trouble.

More important because the gaps AI can't fill become more valuable. AI can write copy, but it can't understand why certain words make people feel something specific about a brand. It can analyze data, but it can't sense when the data is lying because of some subtle market dynamic only someone who's been watching that market for years would notice.

This is why the future belongs to what seems like a contradiction: the highly specialized generalist.

The Highly Specialized Generalist

A highly specialized generalist isn't someone who knows a little about everything. It's someone who knows enough about everything to be dangerous, and who has gone so deep in a few areas that they see connections others miss.

Think about it this way. If you're 95% deep in paid acquisition and 5% broad in everything else, you're not really a specialist anymore. That 5% contains multitudes. It's enough to recognize when your SEO person is BSing. Enough to spot when your designer is solving the wrong problem. Enough to know when your developer is overengineering.

But more importantly, that 5% makes your 95% more valuable. Because you can connect your deep expertise to other domains. You see how paid acquisition principles apply to product development. How customer acquisition cost models mirror hiring decisions. How conversion optimization techniques work in sales conversations.

The specialist sees their domain. The generalist sees all domains but none clearly. The highly specialized generalist sees their domain with crystal clarity and everything else with just enough resolution to spot the patterns.

The Insects Are Coming

Heinlein was more prescient than he knew. Specialization really is for insects. And we're building them. They're called AI agents.

Want something highly specialized? Train a model on nothing but conversion rate optimization. Or derivatives trading. Or sonnet writing. These artificial insects will be better specialists than any human at narrow, well-defined tasks.

But insects, even artificial ones, don't see the big picture. They can't make the leap from derivatives trading to marketing strategy. They can't see that the same psychological principle driving their sonnets could revolutionize their landing pages.

Humans can. But only if they've cultivated the right mix of depth and breadth.

How to Become One

So how do you become a highly specialized generalist?

First, forget the T-shape. Think more like a spike with a platform. The platform is your basic competence across domains—just enough to not be dangerous. The spike is your scary expertise in one or two areas.

Second, apply the 95/5 principle ruthlessly. For every twenty things you could learn, pick one to master. For the other nineteen, learn just enough to have good questions, not answers.

Third, choose your 5% based on leverage, not interest. What are the one or two skills that, if you were world-class at them, would make everything else you do more valuable? For me, it was paid acquisition and basic coding. For you, it might be different.

Fourth, use AI as a multiplier, not a crutch. Let it handle the routine parts of your 95% so you can go even deeper. Let it fill in the gaps in your 5% so you can maintain broader awareness without the time investment.

Finally, and this is crucial: connect everything. The value isn't in being 95% good at one thing and 5% good at others. It's in being the only person who sees how they relate.

The Future of Work

The old model was simple. Start broad, gradually specialize, become an expert. Very linear. Very predictable. Very much like an insect developing through its stages.

The new model is different. Start focused but curious. Go deep enough to see patterns others miss. Stay broad enough to recognize when those patterns appear elsewhere. Use AI to handle everything in between.

This isn't just about marketing. It applies to any field where human judgment, creativity, and pattern recognition matter. Which is to say, any field that will still exist in ten years.

The generalists were right that specialization is limiting. The specialists were right that depth matters. They were both wrong about having to choose.

Heinlein knew this fifty years ago. He didn't say humans should be generalists. He said they should be able to do all those things. Building a wall and writing a sonnet. Taking orders and giving them. The paradox isn't in being both broad and deep. It's in thinking you have to pick one.

The future belongs to those who refuse to choose. Who go so deep in a few areas that they achieve escape velocity, while maintaining just enough breadth to see where they're going.

The insects are coming for the specialists. AI is coming for the generalists. The highly specialized generalists? They're building both.

Which means they're writing the future. Not typing it. Not generating it. Writing it, with all the human judgment, creativity, and cross-domain pattern recognition that word implies.

That's not something you can delegate to an insect. Or an AI. At least, not yet.