What did AI change in my scheme? Incremental updates instead of full projects

The author analyzes changes in the development approach for the PlyLoom project following the integration of Large Language Model (LLM) tools. The focus is on shifting from full project reloads to incremental updates. Previous versions of the system lacked the ability to interact effectively without rebuilding the entire project, and also lacked tools for comparing differences between two files. Using AI has allowed for the optimization of the process of refining knowledge maps and regulations, ensuring more flexible data handling. The author shares their experience integrating LLMs into the workflow, emphasizing the importance of making incremental changes rather than globally restructuring the system. This approach significantly simplifies the maintenance of complex systems and improves version control for regulations, allowing developers to make targeted edits and compare change results in real time.