The Old Model
For most of the industrial age, specialization was a remarkably successful strategy. Institutions were designed around it. The engineer engineered. The accountant accounted. The scientist researched. Entire professional identities were built around mastering a relatively narrow domain and spending decades refining that expertise.
That model made sense in a world where systems moved more slowly and where the boundaries between disciplines were easier to maintain. A person could remain highly effective while understanding only a small slice of the larger machine because the machine itself was less interconnected.
I'm not sure that's true anymore.
Where Modern Work Happens
Modern problems rarely stay confined to a single domain for very long. Technology bleeds into psychology. Data becomes organizational behavior. Artificial intelligence becomes philosophy, economics, infrastructure, labor politics, and ethics all at once. More and more meaningful work exists somewhere in the messy overlap between systems that were never really designed to cooperate cleanly.
For a long time, I thought my own career reflected a lack of focus. I studied physics, worked in restaurants, moved into technical consulting and client support, studied Data Analytics, and eventually settled into implementation consulting. On paper, it never looked especially linear. It looked like someone moving sideways through unrelated worlds.
But over time, I started noticing that the actual work I gravitated toward was strangely consistent. Again and again, I found myself sitting between technical systems and human systems, trying to make them cooperate.
Part of my work involved understanding reporting logic, integrations, workflows, and data structures. Another part involved translating those realities to executives, recruiters, operations teams, sales leaders, and end users who were often operating from entirely different assumptions. The technical team might understand the architecture perfectly while missing how people actually performed their jobs day to day. Leadership might understand the strategic objective while remaining disconnected from the operational friction underneath it. Clients often understood their pain intimately without understanding the constraints of the platforms they had purchased.
The Real Organizational Problem
What struck me after enough years doing this was that most organizations do not actually suffer from a shortage of intelligent people. They suffer from fragmentation.
Every department slowly develops its own internal worldview. Sales teams optimize around urgency and promises. Engineering teams optimize around scalability and constraints. Operations teams optimize around survival and process stability. Leadership operates several abstraction layers above the operational reality of the people using the systems every day. Everyone involved may be competent and well-intentioned, but the organization still drifts toward incoherence because fewer and fewer people are capable of translating between those perspectives.
The result is that many modern problems that appear technological on the surface are actually systems problems underneath. Sometimes they are communication problems disguised as software problems. Sometimes they are incentive problems disguised as workflow problems. Sometimes they are human problems hidden inside reporting structures and implementation timelines.
And increasingly, the most valuable people inside organizations are not necessarily the deepest specialists in a single discipline. They are the people capable of moving between worlds without losing the thread of the larger system.
When Information Gets Cheap
AI is accelerating this rather than reversing it. As large language models make raw information abundant, the scarce resource shifts. Not just possessing information matters anymore, but understanding which information matters, how systems interact, where constraints actually exist, and what unintended consequences emerge when one part of a system changes another. The translators become more valuable precisely as the information they're translating becomes cheaper.
This is one reason I think some conversations about AI replacing knowledge workers miss something important. Work rarely consists of isolated tasks neatly separated from human systems. Most real-world work involves ambiguity, conflicting incentives, incomplete information, shifting priorities, institutional politics, emotional dynamics, and invisible dependencies that never appear on process diagrams.
A project plan may look perfectly logical until it collides with how human beings actually behave. A reporting structure may look efficient until you realize nobody involved shares the same operational vocabulary. A software platform may technically support a workflow while remaining practically unusable for the people expected to live inside it every day.
The Hidden Failure Mode
These are not purely technical failures. They are failures of translation.
Ironically, I think many people who feel professionally "scattered" may already be adapting to this shift before institutions fully recognize it. They learned a little technology, a little psychology, a little communication, a little business strategy, a little systems thinking. Because none of those pieces fit neatly into a traditional professional identity, they assume they lack specialization.
But maybe synthesis itself is becoming a specialization.
A Different Kind of Specialization
Not shallow generalism. Not collecting disconnected trivia. Something more difficult than that. The ability to build conceptual bridges between domains that increasingly depend on one another while speaking entirely different languages.
The strange thing is that modern civilization may depend on these translators more than ever precisely because our systems have become too complicated for isolated expertise alone. No single person fully understands the systems we rely on now. Not global supply chains, not financial systems, not healthcare infrastructure, not machine learning ecosystems, not large organizations. Everything has become interconnected enough that translation itself starts to function like infrastructure.
I also suspect this helps explain why so many intelligent people feel exhausted right now. Modern life requires constant movement between incompatible modes of thought. We bounce between analytics dashboards and emotional conversations, between technical constraints and political narratives, between algorithmic systems and deeply irrational human behavior. Many people are functioning as translators all day long without ever formally recognizing that labor for what it is.
Translation as Infrastructure
And most organizations still struggle to measure or reward it because translation work is diffuse. It often looks less impressive on paper than narrow technical output, even when it is the invisible layer preventing systems from collapsing into dysfunction.
The future will still belong to experts. Civilization will always need people who dive deeply into specific domains. But the highest-leverage individuals over the next few decades will be the people capable of translating between worlds: between humans and machines, between technical possibility and organizational reality, between data and meaning, between systems and the people trapped inside them.
For years, I assumed my broad interests reflected a lack of direction. They didn't. They were building a capacity that linear careers rarely develop — the ability to connect things that were never supposed to fit together.
The People Between Worlds
That turns out to be exactly what this moment requires.