The specialist had the answer. The generalist knew which question to ask.

For most of the last century, the corporate ladder rewarded a fairly simple trade. Go deep in one area, spend three or four years at university sharpening a specific set of skills, and the depth itself became the asset. You became the supply chain person, the M&A analyst, the compliance expert. That specialism was your moat, and for a long time it was a genuinely good one, worth every year it took to build.

That trade is starting to look different, and the reason isn't that depth stopped mattering. It's that AI can now hand almost anyone a working version of that same depth, on demand, for a specific task, without the years that used to be the price of admission.

01Why specialisation is losing some of its shine

The shift is easiest to see in technical work, coding, data analysis, drafting, the kinds of tasks that used to require someone who had spent years building the specific domain knowledge to do them well. AI tools can now produce competent first drafts of all three for someone who doesn't have that background at all. That doesn't mean deep expertise has become worthless, a person who deeply understands a codebase or a dataset still catches things a generalist with an AI tool will miss. But the floor has moved. Getting a reasonable result out of a technical task no longer strictly requires the years of specialised training it once did, and that changes the calculation for what an organisation actually needs to hire for.

Industry research on this shift has converged on broadly the same read. Analysis from PwC and Deloitte points to a growing "hourglass" shape in knowledge work, strong junior and senior tiers, with the specialised mid-tier roles, the analysts, the coordinators, the narrow domain managers, increasingly compressed as AI agents take over more of what used to be their core deliverables. A separate industry survey found that 56 percent of workers already use AI in their jobs, and 58 percent expect their own skill set to change substantially within five years. That's not a forecast about some other industry. It's most people's own expectation about their own job.

02The powerhouse effect of moving fluidly

Here is the part of this shift that matters most for how a company actually gets built. When someone can move fluidly between projects, functions, and departments, rather than being permanently anchored to one narrow lane, they stop being a single input into a single process and start becoming something closer to a connective layer across the whole organisation. They see the handoff between sales and delivery that nobody else notices because nobody else sits in both rooms. They spot the redundant step that exists only because two departments never actually compared notes. That kind of value doesn't show up on a job description built around a single specialism, because it was never really about depth in one place. It was always about the connections between places.

This is also where departmental silos genuinely start to break down, not through a reorg or a new piece of collaboration software, but because the people inside the organisation are no longer structurally confined to one box. Cross-disciplinary problem solving stops being a workshop exercise that happens twice a year and starts being how the actual daily work gets done, because the person doing it has genuine standing in more than one domain rather than borrowed knowledge from a meeting they sat in once.

"A duck can walk, swim, and fly. It won't outswim a shark or outfly an eagle, but it survives and adapts because it was never trapped in a single element."

03Why this doesn't mean depth stops mattering

It would be a mistake to read this as an argument that specialism is now worthless, and the evidence doesn't actually support that reading either. Economist David Autor has made the point that AI amplifies a sound foundation, but if there's nothing underneath, it becomes a liability rather than a shortcut. A generalist who can move across five domains but has never gone deep enough in any of them to develop real judgement is not actually more valuable than a specialist. They're just more scattered.

The winning shape looks less like pure breadth and more like a T, or sometimes an M, depth in one or two areas a person has genuinely earned through hard-won experience, paired with a wide horizontal ability to connect that depth to everything happening around it. The anchor is what makes the range trustworthy. Without it, adaptability is just guessing with more confidence.

A useful reframing of this debate comes from academics who've pushed back on "generalist versus specialist" as the right axis entirely. The more important divide, in this view, is between people trained mainly to recall and reproduce information, and people who can direct, verify, and govern intelligent systems well. AI has made the first category far less scarce than it used to be. It has made the second category more valuable than it has ever been.

04The org chart was built for a world that's ending

There's a structural assumption sitting underneath almost every organisation that's rarely questioned out loud. Companies have historically hired around what their core business happens to be. A financial services firm hires business and finance graduates. A law firm hires law graduates. A hospital hires clinicians. The logic seemed obvious enough that nobody needed to defend it, the work was domain-specific, so the hiring had to be domain-specific too, and the department structure that followed was really just that logic made permanent in an org chart.

That assumption gets shakier every year foundation models get better, and it may not survive the trajectory AI is currently on at all. As the field moves toward something resembling artificial general intelligence, and eventually further still toward systems that exceed human capability across most domains, the specific domain knowledge that used to justify hiring a finance grad for a finance job, or a law grad for a legal one, becomes something the organisation can increasingly draw on directly through the model itself. The domain knowledge doesn't disappear as a requirement. It just stops being the scarce thing a human has to personally hold in their head to be useful in that seat.

We're already seeing the early version of this inside organisations that have started paying attention. More staff want to contribute across functions and across departments, not because they've suddenly become less loyal to their own team, but because the old excuse for staying in a lane, that's not my department, I don't have the background for that, is a much weaker excuse than it used to be when a capable AI tool can help bridge the gap. The people asking to work across boundaries aren't overreaching. They're responding correctly to where the value has actually moved.

What this means practically is that new hiring increasingly needs to select for a different DNA entirely. Less for a credential that proves someone spent four years mastering one narrow field, and more for innovative thinking, the ability to simulate and apply information across contexts it wasn't originally trained on, and a genuinely bold, daring willingness to take on a problem outside their formal training and work it through anyway. That combination is much harder to screen for on a resume than a degree title, which is exactly why most hiring processes haven't caught up to it yet.

05How Trusted Hub is building for this shift

This is precisely the shape of worker Trusted ActionVerse is designed to create space for. A no-code, agentic workflow platform lowers the technical bar enough that someone doesn't need years of engineering training to build something that works, which means the people best placed to design a workflow are increasingly the people who actually understand the problem end to end, not just the people who happen to sit in the technical department.

Inside our own team, the staff building agents on TAV aren't exclusively engineers. They're the people closest to the actual friction, in operations, in finance, in hiring, who understood the problem well enough to know what a good solution needed to do, and had a tool capable enough to let them build it without waiting for a specialist to get to it. That's the powerhouse effect in practice. Give someone the ability to move fluidly and the tools to act on what they see, and the boundary between departments starts to matter a great deal less than what they can actually solve.