From scarce code to abundant builders

| In Tech
| 8 minute read |by Constantin Gonzalez
Interior of a grand steampunk-gothic cathedral with ornate wooden architecture, glowing brass gears and mechanical details on the walls, illuminated by golden lights and lanterns. A massive arched doorway in the center glows with bright golden light and is framed with ornamental designs. Stained glass windows with colorful patterns are visible at the top, and a bright blue sky shines through the open doors, creating a dramatic contrast with the warm interior lighting. The floor features an intricate geometric stone pattern.

According to media pundits, software stocks are “crashing”, and there’s a “SaaSpocalypse” going on. Recently, Noah Smith wrote about “The Fall of the Nerds,” painting a picture of software engineers as the new master weavers: skilled artisans about to be displaced by AI-powered looms. Meanwhile, I’m seeing more and more layoff announcements in my LinkedIn feed.

My heart and hopes go out to everyone who is affected by layoffs. I wish you all the best, and I believe I see good reasons for how this may turn out for the better, after some time.

Meanwhile, the anxiety is real. But I think the narrative is wrong.

What we’re seeing now isn’t the death of software, or the end of tech careers. It’s more like a structural shift from scarcity to abundance. A pattern that has repeated throughout economic history. And once you see the pattern, you can also see something much more useful: new options about what you can do next.

Why software companies exist

Here’s a question nobody seems to be asking: Why do software companies exist in the first place?

The answer is deceptively simple. For decades, the ability to write software was a scarce resource. Building anything useful required specialized coding expertise that most companies simply didn’t have. So the market did what markets do with scarcity: it centralized that rare capability. Software companies emerged as aggregation points: “coding centers” that concentrated scarce programming talent on behalf of the rest of the industry.

This made perfect sense. If you couldn’t build it yourself, you bought it from someone who could.

But here’s the structural tension that was always there, hiding in plain sight: software companies’ domain expertise is inherently horizontal. They have to serve hundreds or thousands of customers, which means their products must work for the general case. Meanwhile, each customer’s actual needs are deeply vertical—specific to their industry, their processes, their competitive edge.

That gap between horizontal product and vertical need was tolerable when the only alternative was building nothing at all. But the alternatives have changed.

The abundance shift

AI coding tools have fundamentally changed this equation. Not by replacing programmers, which is a narrative that misses the point entirely. What they’ve done is: they made the ability to build software dramatically more accessible. The cost of building has dropped. The expertise required to get started has dropped even further.

The scarce resource that justified centralizing coding capability is now becoming abundant.

This doesn’t mean every company will suddenly build everything from scratch. But it does mean that the horizontal-vs-vertical gap, the one that was always there, now becomes the deciding factor. When building is hard and expensive, you tolerate an 80% solution from a vendor. When building becomes accessible, you start asking: “Why am I paying for features I don’t need, while missing the ones that actually matter for my business?”

As App Economy Insights recently noted, even the pricing model is under pressure. Seat-based SaaS pricing assumed that value scales with headcount. In a world where AI agents can execute work autonomously, that assumption breaks down.

Build what makes you, you

One of my favourite customers from my time as a Principal Solutions Architect at AWS offers a great example here.

This German Mittelstand company, an automotive OEM supplier, had built their own Manufacturing Execution System (MES) in-house. For years, I watched them face intense pressure from management to cut development costs and replace it with a standardized, off-the-shelf solution. The argument sounded reasonable: why maintain expensive custom software when you can buy a proven product?

I always argued against it. Their core competency as a German industrial company is precisely the ability to orchestrate complex just-in-time manufacturing processes within the intricate supply chain that defines the automotive industry. Every optimization point, every process refinement they’d developed over decades was encoded in that system. By handing control to a standard software provider, they wouldn’t just be outsourcing code. They’d be surrendering one of their core competitive advantages.

Fortunately, they resisted the pressure. But it was a close call.

Today, AI coding tools are vindicating that decision. Instead of being locked into a vendor’s horizontal product, they can now accelerate their custom MES development, expand its capabilities, and grow their development team’s responsibilities from maintaining code to building intelligent tools and services that amplify the entire company’s capabilities. Their software developers are becoming something more valuable: domain-expert builders who understand both the business and the technology.

We’ve seen this movie before

If the “scarcity becomes abundance” pattern sounds familiar, it should. Dan Pink described almost exactly this dynamic in his 2004 book A Whole New Mind. He identified three forces threatening traditional knowledge workers: Abundance, Automation, and Asia (outsourcing). His argument: when analytical, “left-brain” skills become commoditized, the value shifts to what remains scarce—creativity, empathy, systems thinking, meaning.

I wrote about this on my blog back in 2010, applying Pink’s framework to technology careers. Sixteen years later, the same pattern has simply moved up a level. Back then, Abundance, Automation, and outsourcing threatened individual knowledge workers. Today, those same forces are reshaping the companies that employed them.

The lesson is consistent: when something scarce becomes abundant, don’t mourn the scarcity. Look for what remains scarce—and invest there.

The solar grid, not the ice age

So are software companies doomed? I don’t think so. But the sensationalist framing doesn’t help.

Here’s a better analogy than “apocalypse”: think of centralized power generation meeting distributed solar.

For decades, if you wanted electricity, you bought it from a utility. They had the power plants, the expertise, the infrastructure. Then solar panels got cheap enough for homes and businesses to generate their own power. Just last summer, I installed a balcony solar plant for our apartment that’s been humming along nicely since then. Did utilities disappear? No. But their role fundamentally transformed: from sole provider to grid balancer, integrator, and backstop for what distributed generation can’t cover.

The parallel to software is striking. And if you’ve been in the cloud computing world, you’ll recognize another pendulum swing: cloud computing centralized IT infrastructure into hyperscale data centers, and now hybrid cloud and cloud sovereignty are redistributing some of that back to the edge. Centralization and decentralization aren’t opposites—they’re phases in a recurring cycle. The balance point just shifted.

Software companies that survive this shift will be the ones that stop selling scarce code and start selling what genuinely remains scarce: deep domain expertise, complex system integration, outcomes for organizations that can’t yet build for themselves, and platforms that make distributed builders more effective.

What this means for you

If you’re a business leader, here’s the strategic takeaway: reconsider your build-vs-buy defaults. For anything close to your core value chain—the processes and capabilities that define your competitive edge—building may now be the smarter investment. Not because AI makes building trivial, but because it makes building accessible enough that you no longer have to sacrifice your vertical needs for a horizontal product.

If you’re a tech professional feeling anxious right now, I want to offer some perspective. As Andrew Ng pointed out this week, most current layoffs are pandemic hiring corrections and general cost-cutting, not AI replacing jobs wholesale. But the demand is shifting, and the skills that complement AI are the ones worth investing in.

The good news is: those skills aren’t exotic. They’re the Expert Generalist skills I wrote about last year, and they echo what Werner Vogels described as the “Renaissance Developer”: systems thinking, cross-domain pattern recognition, the ability to bridge business and technology, and the curiosity to keep learning by doing. These are precisely the capabilities that AI amplifies rather than replaces. And they rhyme well with Dan Pink’s creativity advice checklist: Design, Story, Symphony, Empathy, Play, Meaning.

Start building. In every sense.

My challenge to you: invest in learning this week. That might mean building your own tool instead of buying one. For example, I recently built my own video background generator instead of paying €60 for Apple Motion, and the experience taught me more than the tool itself was worth. Check out Simon Willison’s collection of 178+ AI-built tools for inspiration.

Because “building” goes beyond code and tools. Build cross-domain expertise. Build your ability to see systems. Build bridges between the business problems you understand and the AI tools that can help solve them.

The age of scarce code is ending. The age of abundant builders is just beginning. And that’s not a crisis—it’s an opportunity.

For an individual perspective on “Build vs. Buy,” and some more background (pun intended) on the background generator mentioned above, check out this companion video on the Constant Thinking YouTube channel:

YouTube video thumbnail
Play Video

References