Source URL: https://simonwillison.net/2025/Jul/1/kevin-webb/#atom-everything
Source: Simon Willison’s Weblog
Title: Quoting Kevin Webb
Feedly Summary: One of the best examples of LLM developer tooling I’ve heard is from a team that supports software from the 80s-90s. Their only source of documentation is video interviews with retired employees. So they feed them into transcription software and get summarized searchable notes out the other end.
— Kevin Webb, a couple million lines of Smalltalk
Tags: small, ai-assisted-programming, ai, llms
AI Summary and Description: Yes
**Summary:** The text highlights an innovative use of LLM developer tooling utilized by a team dealing with legacy software documentation issues from the 80s-90s. By leveraging transcription software and AI, the team is able to create summarized, searchable notes from video interviews. This exemplifies the practical application of AI in managing and extracting valuable insights from unstructured data.
**Detailed Description:**
The provided content emphasizes a unique implementation of LLM (Large Language Model) technology in a real-world scenario where legacy software documentation is scarce or non-existent. Below are the major points that illustrate the significance of this application:
– **Context of Use:** The team operates in an environment where they support software that dates back to the 1980s and 1990s. Due to the age of the software, traditional documentation methods are likely outdated or missing.
– **Source of Knowledge:** The only existing documentation comprises video interviews with retired employees who originally worked on this software. This situation presents a challenge, as transcribing and extracting meaningful data from these videos manually would be labor-intensive.
– **Application of Technology:** The team employs transcription software to convert the spoken content in the videos into text, facilitating easier processing.
– **Outcome:** The transcripts are then summarized into searchable notes, enabling the team to efficiently find and reference pertinent information about the legacy software.
– **Relevance to AI Fields:**
– This example showcases how AI can assist in **software security** and **infrastructure compliance** by ensuring that crucial information about legacy systems can be accessed and retained strategically.
– Additionally, it illustrates the powerful capabilities of **LLM security** and **AI-assisted programming**, demonstrating how these technologies can enhance productivity and knowledge retention.
In summary, this scenario not only highlights the utility of LLMs in software development and documentation but also showcases a practical method to safeguard knowledge and improve operational efficiency, which is vital for development teams handling aged software systems.