Stop calling it vibe coding (davekiss.com)

A recent blog post challenges the notion of "vibe coding," introducing the term "agentic coding" as a more accurate descriptor of how developers should interact with AI tools in coding. The author highlights a fundamental shift in software development: while AI can generate code rapidly, true expertise lies in the ability to manage and validate AI outputs, ensuring quality through planning and intentionality. This necessitates a deep understanding of both the task at hand and the potential pitfalls of AI-generated code. The significance of this argument for the AI/ML community is profound, as it emphasizes the need for developers to adapt their mindset and skill sets in an AI-augmented landscape. The implications are clear: rather than becoming passive consumers of AI, engineers must engage more critically in the coding process, maintaining their judgement and expertise to leverage AI effectively without compromising on quality. This evolution reflects a broader conversation about the future of software craftsmanship, the preservation of skill, and the balance between efficiency and the artistry of coding.