The New Rules of Relevance
--
There was a time when a career had one shape: join, rise, retire. My father's generation lived this. One organisation, four decades, a gold watch at the end. Loyalty was rewarded with stability, and stability was the point.
That shape is gone. The new one isn't just about changing organisations — it's less linear altogether: lateral moves, second acts in unrelated fields, deliberate pauses, returns that don't retrace the original path. Each shift, planned or not, demands fresh adaptation. For a while, the market found a fix: a single, deliberate stretch of reskilling somewhere mid-career — a course, a return to study, a sabbatical spent catching up — that reset the clock. It worked because the gap didn't move very fast. You closed it once, and it held its value for a decade.
That model made a quiet assumption: that a fixed body of knowledge, absorbed once, would still be the right body of knowledge years later. The assumption held when change was linear. It doesn't hold now. The half-life of relevant knowledge has shrunk from decades to years, in some domains to months, and a single mid-career stretch of reskilling can't inoculate anyone against a world where the tools, the competitive landscape, and the definition of the job itself keep shifting under your feet. Updating yourself is no longer an event. It has become a condition — something closer to fitness than to certification.
Artificial intelligence has sharpened this problem, but it did not create it. What AI has done is make visible a question that was already there: which human capabilities compound, and which ones expire?
The Case for Breadth
David Epstein's book Range makes an argument worth sitting with here. In a world of narrow, well-defined problems, Epstein contends, specialists win — deep, early, focused expertise is the fastest path to mastery. But in a world of wicked, unstructured problems — the kind that don't come with a rulebook, where the challenge itself is still being defined — generalists tend to outperform. They draw analogies across domains, sample widely before committing, and are slower to specialise — a slowness that's the source of their advantage, not a deficiency to be corrected.
This matters for how we think about staying relevant. If the future belonged to narrow, stable problems, the right strategy would be to specialise early and defend that turf. But AI has moved fastest into the narrow, stable, well-defined problems — because those are the ones it can be trained on most easily. What's left, disproportionately, are the messy, cross-domain, judgment-heavy problems where breadth is an asset rather than a distraction.
One place breadth gets built is a startup stint. Early-stage companies rarely have the luxury of clean role boundaries — the org chart hasn't caught up with the ambition yet, so everyone ends up closer to all hands on deck. A finance hire sits in on product calls. An engineer ends up talking to customers. That pressure is breadth being forged in real time, with real stakes, faster than a decade inside a role with tidy edges tends to produce it. The stint doesn't have to be permanent to be worth something; two years inside that pace can build more durable capability than a much longer run somewhere calmer.
What Doesn't Commoditise
A few categories of human capability seem durable, at least for now, and it's worth naming them plainly rather than romantically.
Creative synthesis. The value here was never really about producing something no one has ever seen — people recombine prior influences too, and generative tools can now produce genuinely novel combinations of their own. What doesn't transfer is the judgment behind the choice: knowing why this framing, this piece, this model, out of everything that could have been made — and being willing to answer for having made it. A musician choosing one phrase over another isn't just generating options; they're exercising intention about what the moment means. That intention, and the responsibility that comes with it, is the economic moat.
Judgment under ambiguity. Knowing which problem is worth solving, when the data is incomplete, when to trust the model's output and when to override it. This is different from knowledge. It is earned through repeated exposure to consequence — getting it wrong, seeing the cost, adjusting.
Trust-based relationships. Boards appoint people they trust with ambiguity, not people who scored highest on a test. Clients stay with advisors who have seen them through a bad quarter. This kind of trust is built slowly, through repeated interaction, and it does not transfer easily to a system with no memory of the relationship. It also has a corollary: a generalist doesn't need to personally hold every specialism, provided they can collaborate well with people who do — knowing who to bring in, and having enough trust banked that they say yes.
The ability to keep learning on purpose. This is a disciplined habit of unlearning more than it is raw intelligence — noticing when a mental model has expired and being willing to discard it before the market forces the issue. Staying curious, staying a student, whatever the label — the posture is what matters: treating expertise as something to keep revising, not a plateau you've earned the right to stop climbing.
A working relationship with failure. Every one of the capabilities above is built through trial, and trial means a fair number of misses. People who treat failure as data — something to be examined, priced, and learned from quickly — accumulate judgment faster than people who treat it as a verdict on their competence. This is what resilience actually is at the individual level: the practical capacity to take a hit, metabolise it, and go again, rather than a toughness that pretends not to feel it. Dweck's research on fixed versus growth mindsets names the belief underneath this — ability that feels fixed turns a setback into a verdict; ability that feels developed turns it into information. It's built, not inherited: small, bounded experiments, repeated, plus the habit of starting again before the sting of the last one fades. Organisations that punish every visible failure quietly train their best people to stop experimenting — precisely when the capacity to experiment, exercise judgment and learn from consequence matters most.
Taken together, these five — creative synthesis, judgment under ambiguity, trust, purposeful learning, a working relationship with failure — are a reasonable answer to which capabilities compound rather than expire. But naming them only pushes the real question back a step.
Relevant for What?
Compound toward what, exactly? There's a question underneath all of this that's easy to skip past: relevant for what end? Employability is the default answer, because it's the most measurable one — but it's rarely the whole answer. Some people are optimising for income and the security it buys. Others are optimising for autonomy: the freedom to choose the work, not just perform it well. Others still are optimising for something quieter: time with family, the mental space to think without an inbox pulling at it, work that feels like a continuation of who they are rather than a role they perform.
These aren't competing definitions of success so much as different points on the same person's timeline, and they demand different strategies. Someone optimising for pure employability might reasonably chase every emerging skill the market rewards. Someone optimising for peace of mind might choose depth over breadth in one domain and let the rest go. Neither is wrong. What's costly is not choosing — staying vaguely "relevant" in the abstract, chasing every signal the market sends, without asking which version of the life you actually want.
Morgan Housel's The Psychology of Money makes a related point about wealth that applies just as well to relevance. Housel argues that one of the most common financial failures isn't earning too little — it's never defining "enough," so the goalposts keep moving no matter how much is accumulated. He also makes the case that the real dividend money pays isn't what it buys, but the control it hands you over your own time. Swap "money" for "relevance" and the same trap appears: without a defined "enough," there's always one more course, one more certification, one more tool to learn, because the goalpost was never fixed in the first place. The real payoff of relevance, like the real payoff of money, is optionality — the ability to choose the next chapter rather than have it chosen for you.
Abraham Maslow's hierarchy gives that same "enough" problem a psychological name. Relevance, pursued the way most people pursue it, operates at the esteem level — status, recognition, keeping pace with peers. Esteem needs are supposed to be satisfiable, met and then left behind on the way toward self-actualisation; without Housel's "enough," they never resolve, and a person can cycle at the esteem layer indefinitely, mistaking a treadmill for a rung. There's a regression risk in the same dynamic: the pace of change can pull even accomplished people back toward safety-seeking — hoarding credentials, treating every new tool as a threat rather than something to be curious about. It looks like growth. It's safety-seeking wearing growth's costume.
There's a harder version of the safety question that actually keeps people up at night: if my skills lose relevance, will I have built enough financial security to survive the gap? That's the argument's own logic catching up with itself — the five capabilities compound, but compounding takes time, and someone needs runway while it works. Housel's own answer travels well here: savings that exist not to be spent but to buy room for error, the ability to survive a bad stretch long enough to let judgment and relationships do their slower work. Accepting "enough," done honestly, means defining a financial runway as explicitly as a professional one — treating it as what makes acceptance affordable, not something competing with it.
Bill George's idea of a "True North" is useful here, even outside the leadership context he built it for. George argues that authentic leaders operate from an internal compass, formed by their own values and life story, rather than by chasing external markers of success — and that leaders who lose touch with that compass tend to drift, however impressive their credentials look on paper. The same test applies to the relevance question. A True North doesn't tell you which skill to learn next. It tells you which of the answers above — income, autonomy, family, peace of mind — is actually yours, as opposed to one inherited from a peer group, a parent, or a LinkedIn feed. Skills built on a borrowed compass tend to get abandoned the moment the market stops rewarding them. Skills built on your own tend to survive the reward changing.
George's work draws on an older idea from Warren Bennis, one of the first serious scholars of leadership: that becoming a leader and becoming yourself are, in the end, the same project. Bennis's central claim was that people spend enormous energy imitating a style of leadership — or a style of relevance — that belongs to someone else, and that the ones who eventually lead well are the ones who stop doing that and get comfortable operating as who they actually are. That claim matters more, not less, in an AI-saturated world: a model can imitate a style instantly and at scale — tone, phrasing, the surface pattern of someone's judgment — but it cannot manufacture a person's own accumulated experience, expressed without disguise. Authenticity, in that sense, isn't a soft add-on to relevance. It may be the least automatable input into it.
This is where clarity does more work than any skill on the list above. Once you know what you're actually optimising for, the question of which capabilities to build stops being abstract. It becomes a much narrower, more answerable question: what does this version of a well-lived life require of me next?
That clarity is also what makes a non-linear career defensible rather than merely explainable. A sabbatical, a lateral move that looks like a step down on paper, a pause to raise children or care for a parent, a return to study in your fifties — these used to be gaps to justify in an interview. Increasingly they are strategy, provided they follow from a clear answer to "relevant for what" rather than from drift. A break taken to recover breadth, or to test a domain before committing to it, fits Epstein's argument for generalists: it's sampling, not wasted time. A break taken because the treadmill became unbearable, with no other reason examined, is a different thing — and it tends to show. The difference isn't the break itself. It's whether the person can say, afterward, what it was for.
There's a generational version of this worth noticing. Younger professionals are often described as restless — moving roles every year or two, chasing experiences over titles, uninterested in the linear climb their parents assumed was the whole point. Read uncharitably, that looks like a lack of commitment. Read against everything argued here, it looks more like an instinct that arrived before the theory did: sampling widely, treating a role as a chapter rather than an identity. It won't serve everyone equally — sampling still has to convert into depth somewhere, or it's just motion — but the instinct sits closer to Epstein's generalists than to a lack of discipline.
Pivots raise a second worry, just as real: will I ever reach the top of the game again? The honest answer is that a pivot usually means leaving the old ladder behind — arriving in a new domain rarely means keeping the seniority earned in the last one. But "top of the game" quietly assumes the game stays the same one, and it usually doesn't. The fairer question isn't whether you'll stay first at what you used to do; it's whether the new summit is one you're choosing on purpose, not settling into by default. That's the True North question again, and answering it doesn't make the climb easier. Resilience isn't there to make a pivot comfortable — it's there because it won't be, and an honest description of that serves someone better than a reassuring one.
When Do You Pivot?
There's a question underneath everything so far that has been assuming its own answer: when do you actually pivot? The honest answer is before circumstances remove the choice. A layoff doesn't end a career — plenty of good pivots start there — but it does compress the timeline: the deliberation that would otherwise happen on your own schedule now has to happen on someone else's, and the range of moves available has already narrowed by the time you're making them. Building situational awareness early is what keeps the choice yours for as long as possible.
Two families of signal are worth tracking, and they're usually both present before either gets loud enough to notice on its own. Push signals live inside the work: the learning curve has visibly flattened, you're solving this year's problems with last year's playbook because it's comfortable, the work still rewards you but has stopped teaching you anything. Pull signals live outside it: the market is asking for something you don't have and can't fake, the same "maybe it's time" conversation has come up with more than one person you trust, there's a pull toward something specific rather than a vague discomfort with where you are. Push alone can curdle into complaint. Pull alone can be restlessness mistaken for a plan. Both together is usually the real signal.
Situational awareness isn't a one-time audit; it's the shrinking half-life from earlier, turned personal and made continuous. If the market's half-life keeps shrinking, the gap between when something shifts and when you notice needs to shrink with it. That's a habit, not an event — staying in occasional, deliberate contact with adjacent domains, not becoming an expert in them, just not a stranger to them, so a shift registers as a signal instead of a shock.
Networking is the mechanism underneath much of this, and it's a trait a surprising number of capable people never deliberately build. It isn't collecting contacts. It's staying visible enough that when a pull signal from elsewhere arrives, someone already knows what you're capable of before you have to prove it from scratch — shortening the gap between a shift happening and someone noticing you.
Self-awareness is the harder half, and it needs two views held together, not one. A flattened learning curve doesn't feel like stagnation from the inside; it feels like competence, so introspection alone only takes you so far. It needs pairing with an outside view — someone trusted enough to say what you can't yet see about yourself. Clarity about timing, like clarity about purpose, comes from holding both at once.
The muscle for staying continually prepared is the same one named earlier as learning on purpose, aimed specifically at this question rather than left general: a standing habit of asking, on some regular cadence, whether you would still choose this — not only when something has already gone wrong.
One more distinction belongs here before any of this reads as a call to reinvent yourself entirely: a pivot is rarely a jump from one specialisation to a completely unrelated one. Far more often it is a move into adjacency — the plant controller who becomes a CFO is still in finance, just operating at a different altitude; the compliance officer who moves into risk strategy is still reading the same kind of signal, just earlier and more broadly. Even the story of Dr. Govindappa Venkataswamy, known widely as Dr. V, is really an adjacency move dressed as a leap. Venkataswamy retired from government medical service in his late fifties and founded Aravind Eye Hospital, modelling its operating efficiency on chains like Sears and McDonald's — but he stayed in ophthalmology his entire career. What he borrowed from outside it was an operating model, not a new specialisation.
Musk's jump from software into aerospace and automotive is the example that gets told because it's dramatic, not because it's typical — and it worked partly because he could fund the leap himself and hire the specific expertise he lacked. For most people, the more reliable path is adjacency: close enough that existing judgment transfers, far enough that it counts as genuine movement.
The destination doesn't have to be another employer, either. A pivot can just as easily mean stepping out of employment altogether and building something of your own. Musk, Dr. V, and Bikhchandani weren't only people who moved into new territory; they built the vehicle themselves rather than joining someone else's. Self-employment doesn't exempt anyone from the arguments made here — if anything it sharpens them. The financial runway matters more, not less. The credibility narrative — the specific, translated case for why your experience counts here — has to be built from nothing rather than adapted from a résumé. And the pull toward something specific matters more than ever, because no institution is supplying the structure while you work it out.
Why Mentorship Fits Here
Clarity about what you're optimising for doesn't arrive by introspection alone; it usually arrives through conversation with someone who has already lived a few rounds of the question. It's the same pairing named earlier for self-awareness — the inside view and the outside view held together — now applied to purpose rather than timing. None of the five capabilities above can be taught in a classroom in the way a technical skill can. They are transmitted, not instructed — through proximity to someone who has already navigated ambiguity, made the judgment calls, taken the reputational risk, and can explain the reasoning after the fact in a way a textbook cannot.
An honest admission belongs here too. For most people, what's actually available isn't a senior mentor — it's a peer, someone at a similar stage who'll have the honest conversation. That's not a consolation prize: a trusted peer does the self-awareness work well, sometimes better than a mentor would, because there's no hierarchy in the way and the contact tends to be more frequent. What a peer generally can't do is calibrate a decision at an altitude neither of you has reached yet — they can test whether your reasoning holds together, but they haven't personally made that trade-off. That narrower thing is what an actual mentor adds, rarer precisely because it requires someone who has already stood where you're about to stand. Both are worth having. Neither substitutes for the other.
This is the old apprenticeship model, and it is arguably more relevant now than it has been in decades. A senior executive facing a genuinely novel strategic problem — not a templated one — benefits less from another course and more from time with someone who has stood in a similar spot and can pressure-test their thinking. The mentor's value isn't information transfer. It's calibration: helping someone trust their own judgment faster, and know when not to.
The best mentor relationships are closer to barter than to a transaction. Neither side is paying market price: the mentor, often past the point of chasing the next role, gets a live window into how the ground is shifting, and the quieter satisfaction of relevance passed forward rather than defended alone — closer to what Maslow, late in his life, called self-transcendence. That exchange resists being priced cleanly, which is exactly why it's durable. A market transaction can be undercut by a cheaper competitor. A good barter mostly can't.
It also suggests what should replace the one-time reskilling stretch: not a single credential, but an ongoing structure that pairs current thinking with sustained access to people who have already carried real decisions — judgment transmitted through relationship, not delivered once and left to age. Content is increasingly available everywhere, often generated at low cost. Relationship sustained over months is not, and that scarcity is exactly why it holds its value as everything else gets cheaper.
The Practical Takeaway
None of this argues against learning technical skills, including how to work with AI tools — for a senior executive today, that fluency is close to table stakes, even if it isn't a differentiator on its own. But it argues against treating any single skill, credential, or tool fluency as a permanent hedge. The one-time-reskilling model assumed the world would hold still long enough for a static credential to matter for the rest of a career. It won't hold still again.
The more durable strategy looks like Epstein's generalists: breadth deliberately maintained, judgment built through varied exposure, a willingness to run small experiments and be visibly wrong along the way, and — crucially — relationships with people a few steps further down the road who can help you see around corners you haven't reached yet.
What ties these threads together — Bennis's authenticity, George's True North, Housel's "enough," even the case for a non-linear career — is a form of acceptance that's easy to mistake for resignation but isn't. Accepting your own compass means accepting its limits too: the things you were never going to be good at, the version of success that was always going to belong to someone else.
Accepting "enough" means the chase has a stopping point, and you get to set it rather than have the market set it for you. Accepting failure as data rather than verdict means accepting the outcome without accepting a story about your worth. This is active, not passive: the older idea that effort is best spent on what's actually within your control — your own response, your own compass, your own next move — and wasted on trying to control the pace of change itself, which was never anyone's to control.
But underneath the strategy is a simpler instruction, and perhaps the only one that doesn't expire. Don't chase being relevant. The chase has no finish line — there is always a newer tool, a new capability, a faster mover, and relevance defined purely as keeping pace with them is a race you eventually lose, because everyone does. Stay authentic instead. Know your own True North, and let relevance be the by-product of that — not the goal itself. It won't make the runway shorter, the resilience easier to practise, or the choice to leave a ladder behind painless. What it does is tell you which of that effort is actually worth making. Chasing relevance is exhausting and, in the end, unwinnable. Being unmistakably yourself is not.
How Crossmentors Can Help
This is roughly the shape Crossmentors' CXO Fellowship takes: a nine-month, mentoring-led journey rather than a workshop, pairing accomplished former CXOs with leaders working through a real organisational challenge, inside a small peer cohort. Its five competencies — which the programme calls PIVOT — read like a working definition of the judgment this piece has been arguing for: partnering across boundaries, presence, vision, talent, and enterprise-wide thinking, built through relationship rather than curriculum alone. It won't hand you your own True North. But it puts you, for nine months, in exactly the kind of company that helps you find it faster.
Trust us to get your leaders to be at their best!
Comments