You could argue that speech is literally noise that drowns out the signals of your environment. If you just babbled, it would be useless, but instead you use it intelligently to communicate ideas. LLM output is a new palette with which humans can compose new signals. We just have to use it intelligently.

Prompt engineering is an example of this. A clever prompt by a domain expert can prime an LLM interaction to yield better information to the recipient in a way that the recipient themselves could not have produced on their own.