This is the title of Rod Johnson‘s post: “Just because LLMs are eloquent in natural language doesn’t mean that we should always communicate with them in it. Where important processes are concerned, humans themselves don’t communicate in natural language. Experience shows that processes don’t scale without structure. Forms long predate computers.

The fluency of LLMs in natural language is valuable in communicating, and smoothing over the fuzziness of many integration problems. However, it can also be a distraction, and should not make us lazy in designing systems using LLMs.” Link