How is AI changing the profession of an application developer for young people who are still in the middle of their apprenticeships? Marvin Krieg, a fourth-year EFZ computer science apprentice specializing in application development at InfoGuard, is experiencing this transformation firsthand. Marvin shares his personal perspective on how AI has changed his day-to-day training, what opportunities and challenges this brings, and why different skills are in demand today than just a few years ago.
We’ve been using AI as a tool since the very beginning of my apprenticeship. Back then, AI tools mainly helped us understand parts of code or summarize topics in school. At the same time, we still relied heavily on traditional problem-solving methods and searched for code examples on Stack Overflow or W3Schools. Back then, attitudes toward AI varied widely.
In the inter-company courses (ÜK), we were usually not allowed to use AI so that we could learn the basics from the ground up. I’m very glad about that today, because if you just generate code without understanding it, you don’t build a solid foundation in the fundamentals. That’s exactly what can become a problem later on in everyday professional life.
At vocational school, on the other hand, we were allowed to use AI with virtually no restrictions. As a result, we were usually able to get very good grades. On top of that, some teachers created exams using AI, which could then be solved very easily with AI.
Two years ago, AI was nowhere near as advanced as it is today. While teachers themselves know just how much AI has changed programming in the meantime, their teaching often still follows traditional methods.
In project work today, AI handles a large portion of the work. In my view, the results are hardly ever critically reviewed afterward. Our exam grades speak for themselves: For projects or assignments, we now consistently achieve nearly full marks while spending only a fraction of the time we would have needed without AI.
We can even create entire presentations in a matter of minutes. The result: a nearly perfect PowerPoint presentation that can be used almost as-is and generates correspondingly positive feedback.
Looking back, I’m surprised by just how much my profession has changed in just a few years. The actual programming work is no longer the biggest challenge for me. Today, it’s much more important to clearly define the actual task, evaluate the result, and take responsibility for it.
Machines can write code. But it’s still up to humans to decide whether the code meets the requirements—even when AI is used to assist with that assessment.
At the same time, a shift is currently taking place in the IT job market that cannot be ignored. In recent years, major tech companies have seen a steady stream of massive layoffs. AI or automation are frequently cited as the causes.
Particularly relevant to me: According to the Stanford HAI-AI Index, employment among software developers under the age of 26 has fallen by about 20 percent since 2024. At the same time, the report shows that the effects of AI are particularly noticeable among young professionals and in entry-level positions.
It is precisely these entry-level and junior roles that are important to me as a learner. In the past, after completing my apprenticeship, I would have gradually grown into such tasks step by step. Today, it is precisely these roles that are under greater pressure.
I can clearly feel this shift in the market: In the past, young developers were able to gain experience primarily by writing code. Today, AI is taking over many of these tasks. As a result, there is an increasing demand for skills that cannot be easily automated, such as judgment and an understanding of context, as well as a willingness to take on responsibility.
In conversations with other students and colleagues, I’ve noticed that AI-assisted development has long since become the norm in many companies. It’s no longer just about having individual lines of code explained to you. Today, AI also helps plan functions, find bugs faster, and develop initial solutions.
To me, this clearly illustrates how the industry is changing: projects can be developed faster, and certain tasks take less time. At the same time, the role of developers is becoming more demanding. It’s no longer just about writing code. It’s more about asking the right questions, formulating good prompts, critically evaluating results, and taking responsibility for the solution.
AI is also playing an increasingly important role in my work environment. For example, I use AI to narrow down errors more quickly, better understand code, or develop initial solutions. At the same time, it’s clear, especially in cybersecurity, that AI must not be used without oversight. Especially when processing sensitive data, clear rules, secure processes, and a solid understanding of how AI is used are essential. One must be able to assess where AI provides useful support and where humans must take responsibility.
At InfoGuard, this conscious approach to AI is evident on a larger scale—for example, through investments in our own AI infrastructure and the further development of cyber defense services such as MDR 3.0. There, AI capabilities are specifically integrated into analysis, automation, and reporting, including components of the reporting engine that were developed with AI support. To me, this demonstrates that AI isn’t just a tool you use on the side. It’s a tool that must be consciously embedded into processes and solutions.
I see particularly great potential in code development. Even the supportive use of AI can accelerate development, provided the results are thoroughly reviewed and not simply adopted blindly. For me, therefore, what matters isn’t whether AI is used, but how thoughtfully it is applied. AI can take a lot of work off your hands. But it doesn’t replace an understanding of the project, the requirements, and the responsibility for the final solution.
My apprenticeship—and even more so my profession — have shifted dramatically in just a few years. “Writing code” has become “orchestrating systems.” “AI as an aid” has become “AI as a standard tool.” And what was once a secure junior career path has turned into a market where entry-level positions, in particular, are under considerable pressure.
Many see this as a threat. But I’m convinced that anyone who uses AI blindly without really engaging with it makes themselves more easily replaceable by AI. That’s why I strive to understand a project down to its very foundations, provide AI with the most precise instructions possible, and take responsibility for the final result. For me, this is the best way to ensure I’ll continue to be needed in the future.
I also engage intensively with the topic of AI in my personal life. For example, I use AI to develop my own projects, which I publish on HelveticLabs.org. I don’t see AI as competition, but rather as a tool that I want to master before it becomes a risk to me.
For me, one thing became particularly clear during my apprenticeship: AI is changing not only how we develop software, but also which skills will be in demand in the future. Anyone training to become a computer scientist (EFZ) specializing in application development no longer needs to just know how to program. It’s just as important to formulate tasks correctly, critically evaluate results, and take responsibility for AI-powered solutions.
That’s why I don’t see AI as competition, but as a tool that I want to understand and use thoughtfully. The better I understand the technology, the better I can put it to good use.
In my view, this is precisely where the greatest change in the EFZ Computer Scientist apprenticeship with a specialization in Application Development lies. The profession isn’t changing because we’re programming less. It’s changing because different skills are in demand today. I’m convinced that this is exactly where a great opportunity lies for apprentices and the profession of tomorrow.
My colleagues and I are excited to welcome new members to our team! Apply for an apprenticeship position in 2027 and become part of a team that not only finds computer science exciting but lives and breathes it every day.
Caption: AI-generated image