When I was 15, I wanted an electric guitar more than anything in the world.

Not a nice-to-have want. A burning, all-consuming want. The kind where you fall asleep imagining the weight of it in your hands. I'd read somewhere that good things come to those who wait, and I held onto that line like a lifeline.

I got my guitar. Then my distortion pedal. A couple of years later, I opened my own rehearsal and recording studio and ran it for 14 years.

The studio was in a shelter beneath a small green hill in New Belgrade—one of those little patches of nature in the middle of a concrete jungle. I'd wanted to paint the walls peach, but the color came out straight orange. We had no name yet, so I shrugged and called it Studio Orange.

The kids from the block used to hang out on the hill above us. They started calling it Orange Hill. Years later, when I founded my first software development company, I knew exactly what to name it.

I learned something in that studio that no one teaches you in school: The instrument doesn't make the musician. Passion does. Patience does. The willingness to keep playing when nobody's listening.

I'm telling you this because I need to write a different kind of letter than the one I would have written 10 years ago.

Ten Years Ago, I Tried to Help You Get Started

About a decade ago, my dear colleague Milos and I started the IT Serbia Podcast—the first IT podcast in Serbia. We covered everything: software craftsmanship, agile, career transitions, startups, remote work, women in tech, game development, tax reform for freelancers. We wanted to promote the IT industry in Serbia and activate the broader community of tech professionals.

One of the topics we kept coming back to was helping junior developers break into the industry. Writing a solid CV, preparing for interviews, building practical skills. Back then, those were real barriers—and they were solvable.

I wish those were still your biggest problems.

Today, Your Competition Isn't Human

Let me be direct with you: This is the hardest time in the history of software development to land your first job.

Employment for software developers aged 22 to 25 has declined nearly 20 percent from its peak in late 2022. Entry-level tech postings have dropped between 50 percent and 67 percent, depending on who's counting. Computer science graduate unemployment has hit 6.1 percent—nearly double the rate of philosophy majors.

Read that again. Philosophy majors are finding jobs faster than CS graduates.

And here's what makes it different from every downturn before: Companies aren't laying off their existing developers. They're not hiring new ones. The hiring plans announced in 2025 were the lowest since 2010. Researchers call the phenomenon the "no-hire, no-fire" equilibrium—companies deploy AI, achieve productivity gains, and reduce headcount through attrition.

Nobody gets fired. Nobody makes headlines. And nobody opens a junior position.

The Math That Changed Everything

A GitHub Copilot license costs $19 per month. A junior developer in the U.S. costs $75,000 to $95,000 per year. In Serbia, a junior earns around 1,000 euros net per month—roughly 12,000 euros a year.

Either way, the math is brutal. Even at Serbian rates, AI is cheaper by an order of magnitude. And it works 24 hours a day without complaining.

AI coding assistants now complete tasks about 55 percent faster than developers working alone. Forty-one percent of all code written in 2025 was AI-generated. Eighty-four percent of developers use AI tools during development. And 54 percent of engineering leaders surveyed by LeadDev said they plan to hire fewer juniors because of AI copilots.

Salesforce chief executive officer Marc Benioff froze engineering hiring in 2025, citing 30 percent productivity gains from AI. By January 2026, he confirmed the policy held—engineering hiring stayed "mostly flat" while the company expanded its sales team by 2,000 people. Google and Meta are hiring about half the new graduates they hired in 2021.

But here's where it gets nuanced. Klarna chief executive officer Sebastian Siemiatkowski cut 40 percent of the workforce using AI—then admitted, "We went too far." Klarna started rehiring after quality declined. Harvard Business Review found that companies are laying off workers because of AI's potential, not its performance. Only 2 percent of executives made reductions based on real AI results. Thirty-nine percent cut based on what they expected AI to do.

The fear is running ahead of the reality. But the reality is catching up fast.

The Beginning, Not the End

Here's the number that keeps me up at night.

Ninety-two percent of Fortune 500 companies have adopted generative AI. But only 5 percent to 6 percent have achieved scaled, enterprise-wide deployment. The rest are still in pilot mode.

Think about that. The job displacement we're seeing right now—200,000 to 300,000 U.S. positions displaced or foregone in 2025 alone—is happening while most companies are still experimenting. When the remaining 87 percent move from pilot to production, the disruption multiplies.

January 2026 confirmed the acceleration: 108,435 announced job cuts, the highest January since 2009, paired with the lowest January hiring plans since tracking began. Job openings fell to 6.54 million—the fewest since September 2020.

The economy is growing. GDP rose 2 percent to 2.5 percent in 2025. But employment grew at its weakest pace since the pandemic. Output and employment have decoupled. The gap is not a blip. It's a structural shift.

Not an American Problem

I know some of you reading this might think this is a Silicon Valley story. It's not.

A study called "remAIn relevant," conducted in Serbia in mid-2025 by Inicijativa Digitalna Srbija in partnership with UNDP, surveyed companies operating in the Serbian market. The findings are striking.

Ninety-one percent of companies consider AI the key technical skill for the next five years. Almost 82 percent say flexibility, adaptability, and continuous learning are the most important traits in employees. And 88 percent of companies named AI and machine learning specialists as the most sought-after profiles for the years ahead.

But here's the gap that should alarm you: When these same companies evaluate candidates who apply for jobs, they rate their AI knowledge at a mediocre level. Almost 60 percent of candidates for non-technical positions have basic or no AI skills. Even technical candidates score average at best.

The market is moving faster than formal education can follow. Almost every company in the study said that universities, despite their best efforts, aren't producing graduates with the practical skills the market demands.

Indian IT services companies have cut entry-level roles by 20 percent to 25 percent. EU countries saw a 35 percent decline in junior tech positions in 2024. The World Economic Forum's Future of Jobs Report projects that 22 percent of jobs will be disrupted by 2030, with 41 percent of employers planning to reduce their workforce as AI automates tasks.

The wave is global. And it doesn't discriminate by geography.

So What Can You Do?

Don't wait for your university to save you. It can't. By the time a curriculum is updated, the technology has moved on.

Here's what the data says matters now:

Invest in analytical thinking. The World Economic Forum ranks it the number one most sought-after skill globally, with seven in 10 companies considering it essential. Analytical thinking is the one thing AI still can't replicate well—the ability to look at a complex, ambiguous problem and figure out where to even start.

Master AI tools. Not as a nice-to-have. As a survival skill. Get a license for the best AI coding assistants available today. Learn to use Claude Code, GitHub Copilot, Cursor—whatever is current. The goal isn't to compete against AI. It's to become the human who wields AI better than anyone else.

Focus on product thinking. Invest in understanding UX, UI, and how humans interact with software. AI can generate code. It's still bad at understanding what a user needs, what frustrates them, and what delights them. That understanding is your edge.

Build your people skills. The Serbian study found that the skills companies most want—and most lack in candidates—are flexibility, communication, teamwork, and empathy. Not frameworks. Not languages. Empathy.

Develop the ability to evaluate, not merely produce. The junior role is evolving. It's shifting from "person who writes code" to "person who evaluates, directs, and improves AI output." That requires judgment. And judgment comes from understanding systems deeply.

Two Types of People in This Industry

Let me tell you something I've observed in more than 20 years in software development.

There are two types of people who enter this industry.

The first type came for the money. They heard that developer salaries were high, the work wasn't too demanding, and the career prospects were solid. Fair enough. That was all true.

The second type came because they love building things. They'd write code on weekends for no reason. They'd build side projects nobody asked for. They get lost in a problem and forget to eat.

AI is about to sort these two groups efficiently.

If you're in the first group—honestly, that's okay. AI might be doing you a favor. If software development was never your passion, this is your signal to go find what lights you up. You are being liberated, not eliminated.

If you're in the second group? You'll be fine.

Not because the market will be easy. It won't. But because people who love building things don't stop when the tools change. They didn't stop when we moved from assembly to high-level languages. They didn't stop when frameworks replaced hand-rolled solutions. They won't stop now.

The passionate builders will learn AI tools fastest—not because they're told to, but because they can't help themselves. They're curious. They want to see what's possible. And that curiosity is the most valuable trait in a world where the technology shifts every six months.

The Seed Corn Problem

There's one more thing I need to say, and this one is for the companies, the CTOs, and the hiring managers reading this.

If you stop hiring juniors entirely, you eat your seed corn.

In five years, who becomes your mid-level developer? In 10, who reviews the AI-generated code at a senior level? AWS chief executive officer Matt Garman put it bluntly: Stopping all junior hiring is "one of the dumbest things I've ever heard." Because 10 years from now, you'll have no one who's learned anything.

The pipeline is breaking at the bottom. And the consequences won't be visible until it's too late to fix them.

Good Things Come to Those Who Wait

I started this letter with a story about a kid who wanted a guitar.

Here's what I didn't tell you: There were plenty of moments when I almost gave up. When the equipment seemed too expensive, the dream too far away, and everyone else seemed to already have what I wanted.

But I kept playing. On a terrible guitar, with no pedals, recording on a cassette deck. Because I loved it. Not the equipment. Not the studio. The music.

If you love building software—if you'd do it for free, on the worst hardware, even when nobody notices—then keep going. Learn the new tools. Develop the skills the market is asking for. Be patient. Be relentless.

The instruments are changing. The music doesn't have to.

Good things come to those who wait. But also to those who keep building while they wait.

Tihomir Opačić is the founder of Orange Hill, where he helps organizations bridge the gap between AI potential and reality.