Simon Wardley on Evolving Software Engineering Practices with Agentic AI
Simon Wardley defines the evolution of software engineering practices with agentic AI as a shift from traditional, deterministic approaches to more dynamic, ecological models. He highlights that agentic AI systems, which operate based on intent and can be unreliable or non-deterministic, open new possibilities for resilience by mimicking the adaptive and robust qualities found in nature. Wardley, referencing the work of C.S. Holling and the classic Dobson & Randell paper, emphasizes that these emerging practices are still developing, but they promise to create highly reliable systems from unreliable components by leveraging ecological principles rather than just engineering ones.
