Can AI Code AL? Testing LLMs on a Real Task for Data Editor
In this post, I propose to examine how good LLMs are for AL coding on real tasks. As an example, I will use a real task for the Data Editor that I have been putting off for quite a long time. For comparison, I selected the most powerful and popular models: Claude 4 Sonnet, GPT-5 High, Grok 4, and Gemini 2.5 Pro.