Hi, this is Ray.

Let me show you something humbling. In my office, I have a folder of certifications I earned over the past 15 years. Some of them, at the time, felt like a big deal. Hours of study. Real expenditure. Genuine pride when I passed the exams. And yet, looking through the folder today, a substantial portion of those certifications are for technologies that no longer exist, or have been so heavily superseded that the certification itself is essentially decorative. There's a certificate for a programming framework nobody uses anymore. A platform expertise that the platform itself stopped supporting. A specific software workflow that has been replaced by something that does the same thing better in half the time.

I'm not embarrassed by this folder. I'm just kind of fascinated by it. Because the rate at which the things I learned have become obsolete is genuinely faster than anything previous generations of workers experienced. My grandfather learned a trade in his twenties and practiced essentially the same trade for 45 years. My father learned skills in his twenties, got upgraded skills in his forties, and rode those for the rest of his career. I've had to substantially relearn my professional toolkit roughly every 5-7 years for the past two decades, and the cycle is getting shorter, not longer.

This isn't a complaint. It's the reality of learning in a world where technology is changing faster than any individual education can keep up with. And it has fundamental implications for how we should think about learning itself… what we should learn, how often we should expect to relearn, and what specifically holds its value when the surface layer of skills keeps shedding and reshedding.

Today's newsletter is about how the rate of technological change has changed what learning means. The honest, slightly uncomfortable picture, plus the strategy I've landed on after a lot of trial, error, and one very humbling certification folder. Let's get into it.

Most AI newsletters are written for engineers. This one isn't.

The AI Report is read by 400,000+ executives, operators, and business leaders who want to know what's happening in AI — without wading through code, jargon, or hype.

Every weekday, we break down the AI stories that matter to your business: what's being deployed, what's actually working, and what it means for your team.

Free. 5 minutes. Straight to the point.

Join 400,000+ business leaders staying ahead of AI — without the technical overwhelm.

Read it free today

Let me start with the data, because I think most people sense this is happening but haven't seen the numbers. According to a 2024 industry analysis, the past 10 years alone have witnessed the birth of more powerful smart devices, new home assistants, autonomous vehicles, and advancements in AI and machine learning that have impacted almost all industries… and unlike previous industrial revolutions, technology is not evolving in a linear fashion but progressing at an exponential rate, dramatically shortening the shelf life and relevancy of skills in the workplace.

The "exponential, not linear" framing is the key one. Each generation of technology builds on the previous one, which means new capabilities arrive faster as the foundation grows. The smartphone took maybe a decade to become ubiquitous. Generative AI moved from "interesting research demo" to "transforming most knowledge work" in roughly two years. Whatever comes next will likely arrive even faster.

The workforce data backs this up sobering. The Global Labor Market Conference's 2025 report on skills surveyed 14,000 people across 14 countries. According to coverage of the report, the rapid pace of technology change was the largest disruptor of skills, outpacing globalization, climate change and demographic shifts. Workers themselves rated technology change as the biggest threat to the relevance of what they currently know how to do. Not in some abstract future. Right now. In their current jobs. The very thing that took years to learn is at risk of becoming obsolete within years.

A separate 2024 analysis found similarly stark numbers. According to the research, about 39% of workers' skills will become outdated in the next five years, while 59% of workers will require training within that same time period… a trend that spans every industry from healthcare to manufacturing. Almost 40% of what we currently know how to do will be obsolete within five years. That's a stunning figure, and it's not just relevant to tech workers. Healthcare, manufacturing, every sector. The wave reaches everyone.

So the question isn't whether to engage with this reality. The question is how. And here's where the research gets really useful, because not all skills age at the same rate.

Here's the most important insight I've taken from this research. Skills don't all become obsolete at the same speed. There's a categorical difference between two types of skills, and understanding it changes everything about how you should invest your learning time.

A 2023 study published in Industrial Relations made this distinction explicit. According to the researchers, hard skills, such as knowing how to operate a certain machine, do not age well, whereas soft skills, such as leadership ability, preserve their value over time. Hard skills are the specific technical capabilities tied to particular tools, technologies, and procedures. Soft skills are the underlying capacities like communication, collaboration, problem-solving, judgment, leadership, learning ability itself, and the meta-skills that apply across contexts.

This is the central observation. Hard skills are what's getting eaten by the pace of technological change. The specific framework you learned, the particular software workflow you mastered, the exact platform you became expert in… those have short half-lives. They're going to obsolete on you, sometimes within just a few years. The expertise was real, the certification was real, the time investment was real. And then the world moved.

Soft skills don't work that way. The ability to communicate clearly, to lead a team, to understand a problem deeply before solving it, to negotiate, to manage your own emotions under pressure, to learn new things efficiently… these don't go obsolete when the technology changes. The technologies that demanded those skills 30 years ago demand them today, and will demand them in 30 more years. As the Industrial Relations study emphasized, the labor market returns to soft skills hold up over time in ways that hard skills increasingly do not, especially in technology-intensive sectors.

The implication is strategically important. If you spend most of your learning time on hard skills, you're effectively running on a treadmill… relearning every few years as your previous expertise expires. If you spend a meaningful portion of your learning time on soft skills, you're making investments that compound over decades. Both types of learning are necessary, but the ratio matters more than most people realize.

The phrase "lifelong learning" gets thrown around so much it's become almost meaningless. Let me try to give it some teeth. In the current pace-of-change environment, lifelong learning isn't a noble ideal you pursue when you have time. It's a structural requirement of remaining functionally useful in your career.

A particularly interesting paper looking at older workers and skill obsolescence found a fascinating dynamic. According to the researchers, when workers experience skill obsolescence more or less continuously in their job, this may indicate a healthy ongoing engagement with technological change, and workers who experience skill obsolescence and respond by participating in training have better employment outcomes than those who don't perceive obsolescence at all. The key insight: noticing that your skills are becoming obsolete is GOOD, because it triggers the response that keeps you employable. The dangerous state is being unaware that obsolescence is happening, because you don't update.

This connects to a related finding that I think is genuinely important. The same paper found that when workers with long job tenures decrease their training participation, this is an early indicator of future job loss. The first sign that someone is on track to lose their job, statistically, is that they've stopped learning. Not necessarily because they're bad workers. Because they've stopped paying the maintenance cost of staying current, and the gap between their skills and the requirements of their job grows quietly until it becomes too large to ignore.

The implication: continuous learning is the maintenance cost of professional relevance in a fast-changing environment. You don't pay it once. You pay it forever. Stopping is, in a real sense, opting out.

Okay, here's how I actually think about this now, after a lot of years and a lot of obsolete certificates. Steal what's useful.

Layer 1: Soft Skills (The Compounding Layer)

Spend a meaningful portion of your learning time on capabilities that don't go obsolete. Communication. Writing clearly. Reading deeply. Negotiation. Leadership. Emotional intelligence. Judgment under uncertainty. Collaboration. Self-management. Learning how to learn (which is, meta-amusingly, what this entire newsletter has been about). These aren't fluffy. They're the skills that make you valuable across any technological era.

The investment here is patient. You don't see dramatic short-term returns. But over decades, these compound enormously, and they apply to whatever new world you find yourself in. The ability to write clearly was valuable when I started my career, is valuable now, and will be valuable when most of the tools I currently use are museum pieces. Bet on the compounding layer.

Layer 2: Domain Fundamentals (The Slow-Decaying Layer)

Within whatever field you work in, there are concepts and principles that change slowly even as the surface technology changes fast. The fundamentals of marketing apply across every channel… the channels change, the principles don't. The mathematics underlying machine learning don't change when the specific frameworks evolve. The principles of good design transfer across the medium. Identify these in your field and invest in them deeply.

These take more effort to develop than just learning the current tools, because they're less immediately applicable. But once you have them, you can absorb new specific implementations much faster, because you understand what they're trying to do at a deep level. The fundamentals aren't outdated by the new tools. They're the lens through which new tools become quickly comprehensible.

Layer 3: Current Hard Skills (The Treadmill Layer)

Yes, you still have to keep up with the current technical reality of your field. The specific tools, frameworks, platforms, software, and systems that are actually used in your work today. These ARE the treadmill. You're going to keep paying this cost forever, because the surface-level technology will keep shifting.

The trick here is to be honest about this layer. Don't fall in love with current tools. Don't let your identity become "the person who knows X." X will be replaced. Plan for the replacement. Stay alert for what's coming. Build the habit of comfortable beginner-mode… being willing to be a novice again at something new every few years. As one analysis of skill obsolescence noted, economic theory implies that rapid technological change will lead to obsolescence of a worker's skills unless the worker makes investments in human capital to keep up with and adapt to the new technology. This investment is recurring, not one-time.

Layer 4: Adjacent Possibilities (The Optionality Layer)

Spend a small amount of your learning time exploring fields adjacent to yours. Not deeply… just enough to know what's there. The reason: when major technological shifts hit, opportunities often emerge at the intersections of fields. If you only know your own field deeply, you miss those opportunities. If you have at least surface familiarity with adjacent fields, you can spot intersections that aren't obvious to people who only know one or the other.

This is also where serendipitous re-skilling happens. Many of the most successful career pivots in the past decade have been people who built deep expertise in one domain and then transferred it to a new domain that suddenly needed it. The transfer requires having at least some prior exposure to the new domain. Without it, you can't even see the opportunity.

In terms of what to actually do, here's the structure I've adopted:

A learning hour, most days. 30-60 minutes of deliberate learning, spread across the four layers above. Some days it's reading a book on leadership (Layer 1). Some days it's deepening my understanding of a domain fundamental (Layer 2). Some days it's getting up to speed on a new tool (Layer 3). Some days it's reading something from outside my field (Layer 4). The mix matters more than any single session.

Quarterly skill audits. Every three months, I look at what I'm doing professionally and ask: which of my skills feel like they're aging? What's emerging that I'm not yet competent in? Where do I notice myself falling behind in conversations or work? The audit is uncomfortable. It's also the early-warning system that prevents the catastrophic obsolescence the research keeps warning about.

Aggressive curiosity about new things. When something new emerges that's adjacent to my work, I make a deliberate effort to engage with it before I've decided whether I "need" to. Half the time it's a fad and I move on. The other half, I'm 6-12 months ahead of where I'd be if I'd waited until it was obviously important. The early access to genuinely important new things is worth the time spent on the duds.

Comfort with being a beginner again. This one is psychological more than tactical, but it might be the most important. The willingness to be visibly novice at something new (to ask the dumb questions, to be slow at the basics, to not know what you're doing) is the single most predictive trait of people who stay relevant over decades. The people who can't tolerate being beginners eventually stop learning new things and slowly slide into obsolescence. Don't be them. Embrace the beginner state.

Here's what I want you to take from all this. The world we're learning in has fundamentally changed from the one our parents and grandparents learned in. The half-life of specific technical knowledge is shorter than it has ever been, and it's continuing to shrink. The certifications, the courses, the years of expertise built on specific tools and platforms… those investments have shorter and shorter payoffs.

This sounds discouraging. It actually isn't. Because the structural change in the learning landscape favors a particular kind of learner… the one who has built strong foundations, deep meta-skills, and the habit of continuous engagement with new material. That learner thrives in this environment. The tools change; their ability to absorb new tools doesn't. The platforms shift; their underlying judgment stays sharp. The technology obsoletes; their capacity to learn the next technology has been compounding the whole time.

You can be that learner. Most people aren't, because they didn't get the memo that the rules have changed. They're still investing as if their first big set of skills will carry them for 40 years. They're going to be in for a hard landing somewhere along the way, when the gap between their skills and the world's requirements gets too large to bridge.

You don't have to be them. The strategy is clear. Soft skills compound. Fundamentals decay slowly. Current tools are a recurring expense. Adjacent fields are optionality. Constant low-grade learning, treated as routine maintenance rather than crisis response, keeps you on the right side of the curve.

Even Gandalf had to learn new things over his thousands of years. The ones who stop adapting become Sauruman. The ones who keep learning, keep mattering. Don't go full Saruman.

Keep learning (because you don't really have a choice anymore),

Ray