AI is a powerful ally for software developers. It automates routine tasks, improves code quality, increases productivity, and opens up new learning opportunities. But can AI also replace human creativity, critical thinking, and ethical responsibility?
I don’t think so – at least not for a long time yet. I rather believe that AI is fundamentally changing the role of software developers. AI acts as a copilot that supports, and does not dominate, developers in their work. Let me illustrate this point with Jessica, an experienced software developer who has integrated AI into her everyday work.
Jessica used to write code from scratch, as well as optimize, test, and improve it. These days, generative AI performs many of these tasks – in a fraction of the time. When Jessica is tasked with introducing a new functionality into a project, she starts with a brainstorming session. She describes the technical and business requirements and her vision of the new features. She then asks the AI for two solution options, selects one of them, and suggests refinements.
Now she requests the AI to create placeholder code, basic structures, and an initial implementation for the business logic. She refines the code and lets the AI create test cases, always keeping an eye on the project-specific requirements.
Finally, Jessica checks the result, i.e., she makes sure that the code is reviewed, safe, and ethical. After all, every contribution of the AI also falls under her responsibility.
Accordingly, Jessica uses AI as a copilot to reach a more efficient and dynamic workflow. She saves time and resources, increases the quality of her code through iterations of prompt engineering, expands her knowledge, fosters her creativity, and focuses on her human skills, such as communication, teamwork, and critical thinking.
AI has another advantage that may seem self-contradictory at first: It can help overcome the labor shortage. According to a study by the University of Zurich, IT developers and analysts rank second among the occupational groups with the most acute shortage of skilled workers. But the situation has eased somewhat compared to 2022, and the wider use of AI could bring further relief – as well as promote digital transformation and innovation in many industries and regions.
When experienced developers like Jessica rely on AI as a copilot, they achieve more in less time. They also expand their skills and knowledge by working with different copilots that specialize in a specific task or domain and interact with each other in a multi-agent setup. Depending on the use case, it is possible to combine various large language models (LLMs) to generate, test, or optimize high-quality code. By using and linking LLMs, developers can orchestrate and even guide them to create outstanding applications on their own.