Declarative programming versus imperative programming is a common debate among programmers. Declarative programming allows the computer to “figure out” how to perform a task, while imperative programming requires detailed instructions. AI/LLMs could revolutionize declarative systems by eliminating the need for custom DSL and providing intelligent tooling. Reliability of AI solutions relies on comprehensive tooling rather than raw AI output. Surprisingly, old-school software companies like SAP could benefit immensely from AI declarative capabilities, leading to lower costs for customers. Declarative processing of configurations grounded in AI is crucial for reliable applications. Overall, a combination of AI capabilities and declarative processing is needed for consistent and reliable results in the future of programming.
https://blog.codesolvent.com/2024/09/declarative-programming-with-aillms.html