This article discusses the impact of Large Language Models (LLMs) on the field of software engineering, arguing that while LLMs can increase efficiency, it's crucial to maintain a pipeline of junior engineers who learn through practical experience and problem-solving, rather than solely relying on AI-generated code.