Non-determinism and ownership
Read OriginalThis article explores the non-deterministic behavior of Large Language Models (LLMs) in software development, using GitHub Copilot as an example. The author argues that while LLMs are powerful tools that can accelerate work, they must be treated differently from deterministic dev tools like linters. A key point is that ultimate ownership and accountability for the work, including any errors made by the AI, always rests with the human engineer, as the AI cannot be held responsible for its mistakes.
Comments
No comments yet
Be the first to share your thoughts!
Browser Extension
Get instant access to AllDevBlogs from your browser