Bliki: Interrogatory LLM

This article by Martin Fowler discusses the 'Interrogatory LLM' technique, where an LLM interviews a human to gather context for complex tasks, such as designing software features or reviewing specifications. Instead of humans writing lengthy context documents, the LLM asks one question at a time to extract information, then generates a context report for further use. The approach can also verify document accuracy by interviewing experts, offering an alternative to traditional review. Fowler notes its broader applicability for people who find writing difficult, enabling knowledge capture through conversation. The article is relevant to IT/technology, focusing on generative AI, prompt engineering, and software development workflows.