AI is changing software engineering faster than anyone expected. In the United States, 6.1% of computer science graduates are unemployed — double the rate of art history majors.
According to Ray Kok, CEO of Mendix, a subset of Siemens, this trend shows that traditional software engineering is being redefined by AI.
“AI has catalysed the extinction of software engineering and coding as we know it today,” Kok says. He isn’t suggesting that software engineers are going away — far from it. Instead, he argues that the skills and methods that define the job are being rewritten at an unprecedented pace.
From code to concepts
In an interview with Kok, he explains that software development has always evolved alongside advances in computing power. “We started with machine-level programming languages like assembly because compute and memory resources were scarce,” he says. “Back then, programming was about giving very specific instructions so you could squeeze the most out of limited hardware.”
Then came the 1970s and the rise of the C programming language. Developers learned that reusable routines and higher levels of abstraction made development far more efficient. Combined with improvements in hardware, that shift laid the foundation for decades of software progress.
Now, Kok says, the next evolution is here. “With AI being able to produce code better than most humans, we should be ready for the next level of abstraction for software engineering.”
That new level, he believes, centres on model-based software engineering — where developers focus on visual modelling, automation, and logic rather than manually writing every line of code. Low-code platforms like Mendix already reflect this approach by allowing teams to design software visually while AI handles the underlying complexity.
“Soon, the mainstream software engineering community will conclude that model-based engineering should be the primary method of building software,” Kok predicts. “We’re already seeing this happen with emerging agentic development tools that rely on visual modelling instead of verbose code.”
The job market reality
For many new graduates, the timing couldn’t be tougher. AI tools are now automating large portions of the development lifecycle — from testing and debugging to code generation and deployment. That has slowed hiring across the industry, particularly for entry-level roles.
“Two factors are at play,” Kok says. “AI is automating many software engineering tasks, reducing the need for developers, and at the same time, the skill sets that matter most are shifting.”
The first factor — automation — is easy to see. Many tech companies have cut back on large engineering teams as AI systems handle routine coding. But the second factor may have a deeper impact.
“We’re seeing subject-matter experts with business know-how and a technical background emerge as the software engineers of the future,” Kok says. “AI and model-based techniques remove the need to understand every detail of compute, which allows a broader range of professionals to take part in software creation.”
This widening participation means that coding itself is no longer the core value of software development — understanding problems and designing effective solutions is. As Kok puts it, “AI changes how we develop software and what software is.”
How developers can stay relevant
If AI can now write better code, what should trained developers focus on? Kok’s answer is straightforward: mindset and adaptation.
“Trained developers must accept that their knowledge work will fundamentally shift,” he says. “They need an AI-first mindset — one they exercise daily.”
That means developers should move beyond the idea that coding is their main contribution. Instead, they need to learn how to use model-based and low-code tools to design, assemble, and optimise applications. At Mendix, Kok says, developers from traditional backgrounds like .NET, Java, and C++ often see dramatic efficiency boosts when they integrate visual modelling and generative AI into their workflows.
The rise of non-functional skills
As AI takes over more feature development, the focus of human developers is moving toward what Kok calls non-functional requirements (NFRs) — qualities like maintainability, scalability, and usability.
“With AI now able to handle your feature development, a lot of your knowledge work will be around composition and scaling,” he says. “There’s an art to this — it requires critical thinking and intuition, which we can’t yet find in large language models.”
These skills may not sound as glamorous as building new features, but they’re becoming vital. They’re what determine whether an AI-generated application actually works in real-world conditions, scales efficiently, and delivers a positive user experience.
Rethinking how companies hire
AI is also reshaping what employers look for. “Having great programming skills effectively means nothing in today’s market,” Kok says. “Companies are looking for developers who combine an AI-first mindset with the ability to think critically and use model-based tools effectively.”
For many developers, this shift can feel uncomfortable — especially for those who’ve spent years mastering a specific language. But Kok notes that every major engineering discipline has gone through similar transformations. “Software engineering is actually the last to adopt model-based development as its main approach,” he says. “It’s understandable given the history of computing, but it’s changing quickly now.”
The fusion of tech and business
The structural changes aren’t limited to individuals. They’re reshaping how teams operate, too. Kok points to fusion teams — a term coined by Gartner — as an example.
“In an era defined by digital disruption, organisations face a core challenge: how to integrate technology with business strategy to drive innovation and maintain an advantage,” he explains. “Fusion teams are one answer to that challenge.”
These teams bring together a mix of roles — technologists, data scientists, product managers, analysts, and customer-experience specialists — all working toward shared business goals. Low-code platforms make this collaboration easier, allowing team members from different backgrounds to contribute without needing deep programming skills.
“The idea is to bring business and technology closer together,” Kok says. “When everyone can participate in software development, innovation moves faster.”
The skills that will matter next
Looking ahead, Kok expects prompt engineering and data fluency to be among the most valuable skills for developers. “Ultimately, it’s about selecting the right agentic development platform,” he says. “Prompt engineering is fundamental — both for how you use prompts effectively and how you build them for the applications you’re creating.”
At Mendix, the company is already applying low-code methods to prompt engineering itself. By simplifying how developers build and refine AI prompts, these tools aim to make the creation of agentic applications — AI systems that can act autonomously — more accessible.
For developers, this means that learning how to communicate effectively with AI models will soon be just as important as learning to code was in the past.
A new era for software work
Kok doesn’t believe the future of software engineering is bleak — just different. The profession isn’t ending; it’s evolving into something broader, where creative problem-solving matters more than technical syntax.
As AI takes on the repetitive parts of coding, developers who can think critically, design at higher levels of abstraction, and work closely with business teams will thrive. Those who stay tied to traditional coding alone, however, may find the ground shifting beneath them.
“Higher levels of abstraction and automation are what will keep software engineers relevant,” Kok says. “The way we build software is changing, but the need for people who understand how to make technology work for real problems — that will never go away.”
(Photo by Radowan Nakif Rehan)
See also: Google Gemini Deep Research can now access your workspace files
Want to dive deeper into the tools and frameworks shaping modern development? Check out the AI & Big Data Expo, taking place in Amsterdam, California, and London. Explore cutting-edge sessions on machine learning, data pipelines, and next-gen AI applications. The event is part of TechEx and co-located with other leading technology events. Click here for more information.
DeveloperTech News is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.