Generalist vs Specialist in the Age of AI: Rethinking Talent for Sustainable Business Performance

The generalist vs specialist in the age of AI debate is going to become increasingly important for businesses. We were always told to specialise. Choose a field and focus everything on it. If you want to do marketing, fine. That is where it starts. And then you find yourself in the job market as an intern. And if you’re lucky enough not to simply be making tea, you will find yourself doing the menial work no one else wants to do and that is hard to get catastrophically wrong.

My experience of that was being thrown at “the Facebook” at an advertising agency where I interned in my final year of university. That was in 2009.

We had no idea what we were doing. This was a time of MySpace and half a dozen other social networks – anyone remember Hi5? – when social media strategies were being assembled with the methodological rigour of a toddler building a suspension bridge out of Weet-Bix. But I had, by complete coincidence, visited a friend at Harvard in 2005 and seen this strange thing called “The Facebook” in the wild. This apparently qualified me as the agency’s resident expert. And so began a career in social media.

Over time I got bored of the content side and became more interested in the strategy side, which led me into the broader digital ecosystem, which eventually took me out of purely digital and into what marketers used to call “through-the-line.” We mostly just call it marketing now because life is too short.

What it meant along the way was that I loved operating as a generalist. I had no interest in becoming the world’s most decorated social media expert, the high priest of SEO, a Python wizard, or a WordPress master who could identify a rogue plugin by scent alone. I loved understanding enough about how all the parts worked that I could lean into an area when needed, push a problem forward, or know when a genuine specialist had to be pulled in.

And that has probably been the most consistent thread of my working life. My value has rarely been in having the deepest technical answer in the room. It has been in understanding enough of the moving parts to see how they connect, spotting where the blockage actually is, reducing the fog around an issue, and either getting the work moving myself or bringing in the right person before everyone wastes three weeks solving the wrong problem. I have never particularly wanted to own a narrow trench. I prefer walking the whole battlefield.

The problem is that we have tended to treat specialisation as the mature, serious, adult version of capability, and generalism as something slightly suspicious – as though breadth means you failed to commit. I think that hierarchy is about to age very badly.

To be clear, I am not talking about the shallow generalist who flits from topic to topic, leaves half-built decks in their wake and proudly describes themselves as “strategic” whenever anyone asks what they actually do. I mean the high-agency generalist: the person who can understand enough across disciplines to make sense of an ambiguous problem, frame the right question, pull the relevant pieces together and create movement where previously there was only fog.

There is a reason these people become so useful in volatile environments. The strongest generalists often do three things remarkably well: they notice shifts earlier because they are looking across more of the system; they design responses by translating between disciplines rather than getting trapped inside one functional lens; and they help implementation move because they are less paralysed by narrow role boundaries. That pattern maps closely to the capabilities resilient organisations need most: the ability to anticipate, design and implement change.

The research does not prove that “generalists” as a broad category automatically outperform specialists. It does, however, support a more precise claim: people who combine breadth with agency, role-breadth confidence and boundary-spanning behaviour are more likely to surface issues, build workable responses and carry them into action. [Parker, Williams & Turner, 2006; Parker, Bindl & Strauss, 2010; Fuller & Marler, 2009; Ancona & Caldwell, 1992; Troy, Hirunyawipada & Paswan, 2008]

Why the Generalist vs Specialist Debate Matters in the Age of AI

Where am I going with this?

Well, as time has marched on and the world has evolved, we have been issued more and more capability. Most recently, and most impactfully, AI. I wrote recently about the opportunity AI gives us to get our humanity back, but I think it gives us the opportunity to try something else too: to rethink the kind of people we most value inside organisations.

Because we still cling to an extraordinary amount of received wisdom. “This is what we’ve always been told.” “Don’t fix what isn’t broken.” “Stay in your lane.” These expressions may have been more useful in a world where “The Facebook” was one of the only major social networks, dial-up modems shrieked like trapped demons every time you tried to get online, and Solitaire was the most advanced digital entertainment available on your parents’ office computer.

But that is not the world we are operating in now.

Specialists have always been vital, and they will remain vital. When the problem is known, the lane is clear, and the cost of getting it wrong is high, deep expertise matters enormously. You want a tax specialist doing complex tax work. You want an engineer designing the bridge. You do not want a cheerful generalist with a premium ChatGPT account standing next to a load-bearing structure saying, “I’ve asked a few follow-up questions and I’m feeling good about this.”

But specialists are most valuable when the terrain is stable enough for a clearly defined lane to exist. Generalists become disproportionately valuable when the lanes are blurring, the problem is half-formed, and someone needs to work out what is actually going on before a specialist can even be useful. In stable environments, specialisation wins. In transitional environments, translation becomes power.

Joi Ito and Jeff Howe make this point beautifully in Whiplash. Much of our education and business training has been built to reward focus and execution: solve the known problem, master the defined discipline, follow the prescribed path. That made sense in slower-moving environments. But it has also meant we have chronically undervalued the people who explore, connect, translate and operate comfortably between domains. They argue for using a compass rather than a map – not because direction is irrelevant, but because in fast-changing terrain, the route often only reveals itself as you move.

That is the generalist’s natural habitat.

A strong generalist does not always begin with a complete map. They begin with a compass: a clear enough sense of the problem, the outcome that matters, and the next useful move. AI now helps sketch the map as they walk.

That is the shift.

There is a huge amount of organisational work that previously stalled because someone lacked the exact technical knowledge to move the next piece forward. A strategy problem got stuck waiting for the analyst. A marketing team could see the insight but not wrangle the data. A leader knew a dashboard would help but had no idea how to get one built. A researcher had a strong question but not the time to synthesise the mountain of source material.

In many of those spaces, AI narrows the gap dramatically. It allows a capable generalist to research, test, prototype, question, compare, structure and get far closer to a useful answer before calling in the expert.

That does not mean expertise disappears. It means the runway before specialist intervention is needed gets longer.

McKinsey’s 2025 State of AI research found that the organisations beginning to extract meaningful value from generative AI are not simply handing out shiny tools and waiting for miracles. They are redesigning workflows around the technology. In fact, among the organisational attributes McKinsey tested, workflow redesign had the strongest relationship with self-reported bottom-line impact from generative AI. The same research found that 78% of respondents’ organisations use AI in at least one business function, while 71% report regular use of generative AI in at least one function. The work is already shifting; the question is whether our ideas about talent shift with it.

This is why I believe high-agency generalists are going to become some of the most valuable individuals in organisations. Not because they replace specialists. Not because depth suddenly becomes irrelevant. But because their ability to anticipate change, integrate across boundaries, and keep work moving becomes more powerful when AI expands how far they can travel before deep technical expertise is required.

AI does not turn generalists into specialists. It increases the amount of valuable work a capable generalist can carry before specialist expertise is needed. [Brynjolfsson, Li & Raymond, 2023; Dell’Acqua et al., 2023; McKinsey, The State of AI: How Organizations Are Rewiring to Capture Value, 2025]

They are led by EQ as much as IQ. They understand enough of the system they operate in that they can shift between roles without needing six weeks, a formal induction programme and a commemorative lanyard. They can be working in marketing one day and alongside finance the next – not because they are specialists in either, but because they can understand the problem, clarify the destination, identify the constraints and use AI to help resolve much of the blank space in the middle.

Why Businesses Need Adaptable Capability for Sustainable Performance

For businesses, the generalist vs specialist in the age of AI question points toward a far more fluid model of capability.

Start-ups work like this all the time, largely out of necessity. Nobody in a five-person business gets to say, “Sorry, I only do brand strategy.” You write the deck. You call the client. You help shape the offer. You unblock the invoice. You might also hold the bin bag open while someone else drags the broken office chair out of the meeting room because entrepreneurship is, above all else, glamorous. It is messy. The mistakes are many. But the learning is fast, the context is shared, and the growth can be hyper.

None of this should become an excuse for businesses to dump three jobs into one vague “versatile” role and call it innovation. Generalists are not organisational duct tape. Their value comes from judgement, synthesis and movement across boundaries – not from being the person everyone quietly overloads because they are competent and unlikely to complain.

The real opportunity is to preserve some of that start-up fluidity as businesses grow. Large organisations naturally become more structured. They need to. But too often structure hardens into narrowness. Roles shrink. Permission chains lengthen. People become less connected to the whole and more responsible for defending their little square of the org chart as if marauding horsemen might arrive at any moment.

That becomes a resilience problem.

As organisations face faster change, denser systems and more ambiguous problems, their capability cannot only live inside increasingly narrow specialist silos. They also need people who can see across the whole, translate between disciplines, identify second-order effects and turn fragments of expertise into coordinated action. Robert Anderson and William Adams make a related point in Scaling Leadership: the complexity of the environment can outgrow the complexity of our internal operating systems.

This is not only a talent argument. It is a resilience argument. Organisations built around rigid roles and narrow capability pathways will struggle to adapt. Organisations that cultivate people who can move between problems, work across boundaries and extend their own capability with AI will have far more room to respond.

There may also be a human upside to valuing breadth properly.

The HBR Guide to Beating Burnout makes the point that monotonous work, routine overqualification, low autonomy and a lack of novelty can all contribute to disengagement and burnout. It also explores job crafting: the idea that people can redesign elements of their work to create more meaning, variety and ownership.

That resonates with me. I have always felt more energised when I have multiple things to bounce between, when I can make connections across seemingly unrelated areas, when I am not trapped inside a narrow lane pretending that specialisation alone is the highest form of professional virtue. For some people, depth in one area is thrilling. Good. We need them. For others, the energy comes from synthesis. We need them too.

For years, I treated my own tendency to move between ideas, projects and domains as something to discipline out of myself. Increasingly, I suspect it was never a flaw. It was simply a poor fit with a model of work that prized narrow consistency more than connective intelligence.

The Future of Work Needs Human Judgement, Cross-Functional Skills and AI Leverage

AI does not erase that distinction. It amplifies it.

But there is an important guardrail here. AI is an incredible work partner. It is a terrible colleague.

As a work partner, it can help you accelerate what is in your head. It can help you make connections faster, structure ideas, interrogate thinking, generate first passes, spot gaps, surface questions and turn raw thought into something you can react to. I know that from using it daily. It is extraordinary.

But as a colleague? Awful. No empathy. No recognition. No sensing the temperature in a room. No noticing that the person who said “fine” is, in fact, very much not fine. No ability to build trust slowly, to carry relational history, to understand that the stated problem is sometimes just a socially acceptable decoy for the real one. It cannot create the human conditions under which complex work actually succeeds.

Which means the generalist of the AI era is not simply “person plus tool.” The valuable generalist is human judgement plus systems understanding plus relational intelligence plus AI leverage.

That combination is powerful. It also suggests that we may need to rethink how we identify talent. In Whiplash, Ito and Howe write about skill diversity and the danger of assuming traditional credentials are the best predictor of usefulness. They argue for matching talent to task more intelligently rather than simply matching CVs to predefined roles.

The person who can write, analyse, sense tension in a room, ask sharp questions, use AI well, move comfortably between departments and learn rapidly may matter more in many contexts than someone whose résumé looks beautifully narrow and reassuringly conventional. The future will still reward expertise. But it will also reward uncommon combinations of capability.

This shift is already beginning to appear in how stronger organisations think about talent. McKinsey’s research on CFO priorities found that finance leaders view capability building and advanced technologies as two of the most valuable ways to strengthen organisational resilience. It also found that respondents from top-performing organisations were more likely to see talent rotation – including movement across roles and into business functions – as an effective way to build the capabilities they need.

So no, I am not arguing for the death of the specialist. That would be idiotic, and also a wonderful way to become deeply unpopular with lawyers, surgeons, engineers, accountants and the kind of software developer who can make your website load before civilisation collapses.

I am arguing that we have over-indexed on specialisation as the only respectable form of professional value.

The generalist vs specialist in the age of AI debate creates space to rebalance that. Specialists will remain indispensable. But high-agency generalists – particularly those who can anticipate emerging shifts, design across boundaries and help implement movement – will be able to travel further, solve more, translate faster and create greater value before the specialist even needs to enter the room.

Let the foundations of the work shift and move. Call in specialists where the work demands depth, precision and consequence. But let generalists run further. Let them roam until they run out of runway. Give them the tools, the context and the trust to operate across the organisation rather than inside a shrinking square of it.

I did not set out in 2009 to build a career as a generalist. I was handed “the Facebook” because I had seen it once and the adults in the room were no less confused than I was. But the path that followed – learning enough, crossing boundaries, moving from task to system, calling specialists when needed – may have been better preparation for this moment than a much neater career would have been.

Because in a world that refuses to sit still, the most useful people may no longer be the ones who know one thing more deeply than anyone else.

They may be the ones who can make sense of many things quickly enough to help everyone else move.