Rene van Osnabrugge, co-creator of the LEAD podcast, discusses the evolution of developer roles in the age of AI, sharing insights from conversations about GitHub Copilot and engineering culture.

Your New Rubber Duck is an AI

By Rene van Osnabrugge

In 2024, Rene van Osnabrugge and his colleague Geert van der Cruijsen launched the LEAD podcast, focusing on the multifaceted process of building an engineering culture. Over numerous episodes—with various guests and singular discussions—they explored trends affecting modern software development. This blog post distills lessons from those conversations, highlighting the growing influence of generative AI (GenAI), particularly in the context of developer work.

Are Developers Becoming Obsolete?

A recurring theme in the podcast was the speculation around AI making developers redundant. This debate gained traction after bold predictions in articles suggesting developers could soon be out of work. However, Rene’s own daily reliance on GenAI contradicted such claims, especially in coding and architectural contexts. To address these concerns, Rene and Geert invited April from GitHub, an expert in cloud and DevOps, and an advocate for developer experience and productivity via tools like GitHub Copilot.

April gave her perspective: developers are not vanishing, but their workflows are evolving.

Historical Parallels

April drew parallels to prior technological shifts—cloud computing, virtualization, containerization—where job loss was predicted but didn’t materialize. Instead, these transitions required professionals to adapt. AI, she argued, is another such inflection point. The fundamental responsibilities of developers, architects, and engineers—decision making, system design, and critical trade-off analysis—remain outside AI’s current reach. While GenAI can suggest solutions and automate repetitive tasks (like a supercharged ‘rubber duck’), the human expert retains ultimate control and direction.

GitHub Copilot: The “All-Knowing Junior”

An apt analogy shared in the podcast compared Copilot to pair programming with a junior developer who has access to the world’s knowledge base. Despite its breadth of information, Copilot can make basic mistakes or propose solutions that do not compile. This underscores the need for oversight—testing, validation, and contextual understanding remain essential. Since AI lacks grasp of organizational specifics or business contexts, the onus is on the developer to provide proper input and interpret outputs sensibly.

The Risk of Not Using AI

A significant insight was that avoiding AI may be riskier than misusing it. If colleagues leverage AI to complete tasks in minutes which take others hours, resistance to adoption can result in losing professional relevance. Rene suggests that technology continuously redefines expectations, making AI a tool for augmentation rather than replacement.

Proficiency in ‘prompt engineering’—the iterative practice of refining queries and leveraging context—emerges as a new skill for maximizing GenAI’s value. The best outcomes arise from thoughtful dialogue with AI assistants, just as with human collaboration.

Cultural Change, Not Just Licenses

Buying GitHub Copilot (or similar tools) isn’t a silver bullet for productivity. Effective adoption requires a supportive culture, education, experimentation, and active communities for sharing usage patterns and lessons learned. Drawing analogies to DevOps, the shift is as much about mindset and process as it is about tools.

The Future of AI in Development Work

Looking ahead, April envisages AI playing an embedded role throughout the software development lifecycle—from writing and testing code, reviewing pull requests, enhancing security, to automating aspects of documentation and team collaboration. While simple bug fixes or features might eventually be handled autonomously, developer guidance and oversight remain indispensable.

Adapting for Continued Relevance

Rene concludes that developer roles are not facing extinction, but those who shun learning and experimentation may find themselves left behind. The path forward is not about fending off obsolescence, but about staying relevant by:

Investing in new skills beyond coding, such as effective AI utilization
Learning to give better input and evaluate AI output critically
Maintaining curiosity and openness to change

Ultimately, the enduring value of developers is their capacity for critical thinking, system design, collaboration, and lifelong learning—domains where AI can augment but not fully replace human ingenuity.

Listen to the Episode

For those interested in the broader discussion, the full episode is available via the LEAD Podcast.

This post appeared first on “René van Osnabrugge’s Blog”. Read the entire article here