The Half-Life Problem: What Happens When Your Expertise Expires Before Your Career Does

Here is the uncomfortable math that most career advice still pretends doesn’t exist.

Forty years ago, a professional skill had a useful half-life of roughly ten years. Learn something at 25; expect it to remain competitively relevant until your mid-thirties, maybe beyond. That was the architecture of expert careers — learn deeply, master your domain, and let seniority and reputation compound over time. It worked for decades.

Today, according to Kian Katanforoosh, a Stanford University lecturer and CEO of Workera, the half-life of a general professional skill has shrunk to roughly four years. For AI and digital technology skills, he estimates it is closer to two years, as he described to CXOTalk host Michael Krigsman — and that number keeps shrinking. This data was reported by Forbes in April 2024.

Two years. Half-life of two years for the skills at the center of the modern knowledge economy.

If you are forty-five and expecting another twenty years of professional life, you are looking at ten complete cycles of refresh for your most technically current skills. But most professionals are still running a career architecture designed for one.

The numbers that should be keeping more people awake

#

LinkedIn’s Blake Lawit, Chief Global Affairs and Legal Officer, confirmed to TechCrunch in April 2026 that the skills needed for the average job have already changed 25% over recent years — and LinkedIn expects that figure to reach 70% by 2030. His precise framing deserves to be repeated: “So, even if you’re not changing jobs, your job’s changing on you.”

The World Economic Forum’s Future of Jobs Report 2025 puts the structural shift in starker terms: 39% of workers’ core skills are expected to change by 2030, with the WEF projecting 170 million new jobs created globally but 92 million displaced, yielding a net gain that masks enormous underlying churn. Sixty-three percent of employers in the WEF survey cited the skills gap as the largest single obstacle to business transformation — not capital, not technology availability, not regulation. Skills.

IBM’s Institute for Business Value surveyed executives and found they estimate that 40% of their workforces will need reskilling due to AI and automation over the next three years. The Forbes report on this data translated that percentage into a number: 1.4 billion of the 3.4 billion people in the global workforce.

Harvard Business School professor Raffaella Sadun, writing in Harvard Business Review’s September–October 2023 issue alongside colleagues from Boston Consulting Group, puts the half-life of some job skills at five years, and notes that in certain tech roles the figure halves again to roughly 2.5 years. The article arrived before this year’s labor-market data hardened the picture further.

Here is what makes this a qualitatively different problem from any previous period of career disruption: the pace of obsolescence now exceeds the pace at which most individuals can organically reskill while doing their actual jobs. The career math no longer works the way we were told it would.

The architecture problem nobody is talking about

#

When most professionals hear “skills are expiring faster,” they reach for the obvious fix: learn more, learn faster, stay current. Add a certification. Take a LinkedIn Learning course. Dedicate Friday afternoons to professional development.

That response is not wrong. But it is insufficient in a way that matters enormously.

The deeper problem is architectural.

Most professionals — including most senior, highly accomplished ones — have built their career value on a single-mode architecture. Their worth, in their own minds and in the market’s eyes, derives primarily from technical depth in a specific domain. They are the Python developer, the M&A lawyer, the quantitative analyst, the data science manager who runs XGBoost pipelines. Their identity, their compensation, their professional reputation — all of it rests on a foundation of technical expertise.

When that technical expertise carried a ten-year half-life, single-mode architecture worked. When it carries a two-to-four-year half-life, single-mode architecture is fragile. Not challenging. Fragile. There is an important difference.

A fragile career architecture does not bend under pressure — it breaks. And it tends to break silently, over several years, until the professional wakes up to discover that their core value proposition has quietly slipped and that the market has been compensating them on the basis of reputation lag rather than current capability.

The conventional advice — learn faster, stay current — treats this as a learning problem. It is actually a portfolio problem.

Perishable and durable: a different way to see professional value

#

Think about career value as a portfolio with two distinct asset classes: perishable skills and durable assets.

Perishable skills are the technical and methodological capabilities that give you near-term relevance — specific tools, platforms, languages, frameworks, regulatory knowledge, current best practices. They are essential for doing the actual work of your profession. But they depreciate. Some depreciate slowly; some, in today’s market, depreciate at terrifying speed. A detailed knowledge of the GPT-4 API architecture is already less relevant than knowledge of the agents and multi-model orchestration frameworks that replaced it. A deep mastery of Universal Analytics is moot. A specialist in pandemic-era remote-work legal compliance found their niche collapsing as regulations normalized. Perishable skills are not bad investments. They are just finite ones.

Durable assets are the four dimensions of professional value that do not depreciate with market cycles — and which, given intentional investment, tend to appreciate over time.

They are:

1. Contextual judgment

#

The capacity to read a complex situation, synthesize incomplete information, recognize patterns across contexts, and make decisions under ambiguity. This is not a skill in the conventional sense — it is an accumulated cognitive asset that deepens with varied experience. A professional who has navigated ten organizational change programs has judgment that no certification can replicate and no AI tool can currently substitute. The market prices this more generously as AI compresses the value of mere information processing: what remains distinctively human is the judgment that sits above the data.

LinkedIn’s “Skills on the Rise” ranking for 2025 — released by Forbes in March 2025 — lists Analytical Thinking, Adaptability, and Innovative Thinking in its top five. These are not skills you can acquire in a weekend; they are cultivated over years of increasingly complex work.

2. Relationship capital

#

Trust networks, reputational endorsements, and deep professional relationships are among the most durable career assets a professional can hold. They compound. They transfer across domains. They generate opportunity, information, and support in ways that are nearly impossible for newcomers to replicate quickly.

Relationship capital depreciates only when it is neglected — and it can be rebuilt, though slowly. In practice, most professionals underinvest in it during their years of perceived technical peak (when they feel they don’t need it) and then find themselves scrambling when the technical peak passes. The irony is that the same people who invest obsessively in technical credentials often invest almost nothing in building a trust network that would outlast any single credential.

3. Learning velocity

#

The capacity to acquire new capability quickly, efficiently, and without requiring extended periods of disruption. This is itself a meta-skill — and it is one of the few professional capabilities that can improve with intentional practice rather than decaying with time.

Professionals with high learning velocity can integrate new perishable skills relatively cheaply. When the next platform shift arrives, they do not need to start from zero; they translate existing conceptual models into the new context, identify the delta, and close it efficiently. Low learning velocity turns every skills refresh into a crisis. High learning velocity turns it into a cost of doing business.

This distinction matters enormously for career longevity. The professionals who are genuinely resilient in the current environment are not necessarily the ones who know the most about today’s AI tools; they are the ones who can acquire knowledge about tomorrow’s AI tools faster than their peers.

4. Career narrative

#

The ability to articulate your professional value compellingly, translate experience across domain boundaries, and make your relevance visible in any market context. This is a constructed asset — it does not arise naturally from doing good work — and most professionals dramatically underinvest in it.

Herminia Ibarra, in her updated edition of Working Identity (Harvard Business Review Press, 2023), makes the counterintuitive case that career reinvention does not begin with introspection — it begins with action and experimentation. But action without narrative is invisible. Professionals who can construct and deploy a compelling career story that bridges what they have done and what they can do next hold a durability advantage that compounds over time. Their value is visible and legible; their pivot costs are lower; their network activation is faster.

Why this is a portfolio problem, not a learning problem

#

The conventional response to skills obsolescence is to run faster on the treadmill of technical learning. This is necessary but not sufficient — and for one very important structural reason.

When a professional’s career architecture relies primarily on perishable skills, every cycle of obsolescence depletes the portfolio significantly. The response — acquiring new perishable skills — replenishes that one asset class. But it does nothing for the others. And it consumes the finite cognitive bandwidth and time that might otherwise be invested in building judgment, relationships, learning velocity, and narrative.

A professional who spends 80% of their development time on technical skills maintenance is making a portfolio decision, even if they don’t experience it that way. They are concentrating investment in an asset class with a two-to-four-year half-life. The portfolio is not diversified. When the technical skills layer turns over — as it now reliably does — the rest of the portfolio is underdeveloped, and the transition is more painful than it needed to be.

The portfolio reframe matters because it changes the decision logic. The question is no longer just “am I learning fast enough?” It is: “what proportion of my development investment is going into durable assets versus perishable ones, and does that proportion match my career horizon?”

A 28-year-old early in their career should heavily weight perishable skill building — technical depth creates differentiation and compensation leverage at that stage. A 42-year-old who has built some technical depth should be actively rotating investment toward the durable layers. A 55-year-old running primarily on deep but aging technical expertise, with thin relationship capital and a weak learning velocity, is in a structurally precarious position regardless of how impressive their résumé reads.

Auditing your own architecture

#

This is a practical exercise. Answer these five questions honestly.

Question 1: Where does your professional income primarily come from? Is it from specific technical knowledge that someone with the right training could acquire in 12–24 months? Or is it from judgment, trust, relationships, and pattern recognition that took years to accumulate and cannot be rapidly replicated?

Question 2: How fast is your primary technical layer depreciating? If you have not updated your core technical skills in three or more years, are you confident the market still values your version of that expertise at its previous price? Have you tested this in a live market conversation recently?

Question 3: What is your learning velocity benchmark? When was the last time you acquired a meaningfully new capability from scratch? How long did it take? Was it faster or slower than the time before?

Question 4: How much of your current opportunity flow comes from your relationship network versus from open-market visibility? If most of it comes from people who have known you for more than five years, your network may be a strength — or it may be a comfort zone that has allowed you to avoid building new relationship capital.

Question 5: Can you articulate your professional value to someone who knows nothing about your current industry in under two minutes, in a way that translates across two or three different domains? If not, your career narrative needs work — and that work has direct return-on-investment implications.

The goal of this audit is not to generate anxiety. It is to surface the portfolio allocation you have made by default so you can make a deliberate choice about whether that allocation still serves your career horizon.

What this means for the decisions in front of you

#

The Half-Life Problem is not an abstract threat. It is already reshaping compensation, mobility, and career trajectories for a large share of professionals in their thirties, forties, and fifties.

The professionals who are navigating it well share a common structural pattern: they treat perishable skill development as a continuous operating cost, not a periodic crisis, and they invest systematically in at least two of the four durable assets. They are not necessarily the fastest learners or the most technically current people in their organizations. They are the ones who can walk into a new context and establish relevance quickly — because their portfolio works in multiple market conditions.

For the organizations asking these questions at scale: the HBR reskilling analysis from Tamayo, Sadun and colleagues argues convincingly that reskilling programs fail when they are treated as standalone HR functions rather than connected to corporate strategy and individual employee growth. IBM’s data shows 40% of the workforce needs reskilling over the next three years. The strategic question is not whether to invest in reskilling — it is whether to wait until the need is acute or build the infrastructure now.

For individuals: the most practical immediate step is not a new certification. It is an honest audit of where your career value actually lives — which layer of the portfolio — and whether the proportion of your development investment matches a deliberate thesis about your career’s next phase.

The question that outlasts any single answer

#

Here is where I want to leave you, and it is not comfortable.

Most career development conversations are about what to learn next. That is a perishable skills question, and it has its place. But the more important, less frequently asked question is: what are you building that will still be valuable when what you know today is worth half as much?

The professionals who navigate the next decade well will not be the ones who learned the most AI tools. They will be the ones who built careers that were not fragile in the first place — careers anchored in judgment no model can replicate, relationships no algorithm can substitute, learning velocity no credential can confer, and a narrative clear enough to survive any market.

The half-life of expertise is shrinking. The durable assets of a career are not.

The question is which one you are investing in.

References

#

• Perez, Sarah. (April 15, 2026). “LinkedIn data shows AI isn’t to blame for hiring decline… yet.” TechCrunch. https://techcrunch.com/2026/04/15/linkedin-data-shows-ai-isnt-to-blame-for-hiring-decline-yet/ (Accessed April 25, 2026)
• McKendrick, Joe. (April 30, 2024). “AI Puts The Squeeze On The Shrinking Half-life Of Skills.” Forbes. https://www.forbes.com/sites/joemckendrick/2024/04/30/ai-puts-the-squeeze-on-the-shrinking-half-life-of-skills/ (Accessed April 24, 2026)
• Tamayo, Jorge; Doumi, Leila; Goel, Sagar; Kovács-Ondrejkovic, Orsolya; Sadun, Raffaella. (September–October 2023). “Reskilling in the Age of AI.” Harvard Business Review. https://hbr.org/2023/09/reskilling-in-the-age-of-ai (Accessed April 24, 2026)
• Castrillon, Caroline. (March 19, 2025). “LinkedIn Reveals The Most In-Demand Skills On The Rise For 2025.” Forbes. https://www.forbes.com/sites/carolinecastrillon/2025/03/19/linkedin-reveals-the-most-in-demand-skills-on-the-rise-for-2025/ (Accessed April 24, 2026)
• World Economic Forum. (January 2025). The Future of Jobs Report 2025. https://www.weforum.org/publications/the-future-of-jobs-report-2025/ (Accessed April 24, 2026)
• Ibarra, Herminia. (October 2023). Working Identity: Unconventional Strategies for Reinventing Your Career (Updated Edition). Harvard Business Review Press. https://herminiaibarra.com/working-identity-book/ (Accessed April 24, 2026)
• CXOTalk. (n.d.). “Demystifying AI: From Deep Learning to Workforce Transformation.” Interview with Kian Katanforoosh, Stanford University and Workera. https://www.cxotalk.com/episode/demystifying-ai-from-deep-learning-to-workforce-transformation (Accessed April 24, 2026)