Software engineering is rapidly evolving from a code‑centric craft to a broader, system‑level discipline. The most visible drivers are (1) the rising importance of system design and ownership—engineers are now expected to understand and manage the full product line, including cost, latency, and risk, rather than merely delivering working code; (2) the widespread adoption of AI tools, which are being integrated into every stage of the development pipeline; and (3) the emergence of a "harness" around code, encompassing routing between large language models, context management, token consumption, security guardrails, tracing, and evaluation metrics. This shift is often described as moving from traditional software engineering to "factory engineering," where AI agents handle planning, coding, review, testing, and deployment, while human engineers retain judgment at critical checkpoints. The practical implications include a need for engineers to develop skills in system architecture, cost modelling, and AI‑augmented workflows, and organizations must invest in robust harnesses and tooling to ensure reliable, secure, and cost‑effective AI integration. [1] [2] [3]