>>/104789/
Nobody's going to accept you're talking about, strictly applying rules to their own speech, but I'll make it simple:

> You can't sanitize what was rotten from the root. 
> Not just "how to play," but the history, the genealogy, the cultural foundations of rock, metal, and rap. What I found broke something in me.
> Not just the musical techniques - the values. 
> destructive aesthetics baked into the DNA of the genres
> But you can't purify water by pouring it through sewage.
> The foundational culture is foreign to European ideals - deeply, fundamentally alien. The more I studied, the more the cognitive dissonance tore at me.
> Yes, there are exceptions - individual artists who transcended their medium. But they're exceptions. They prove the rule.
> Not with mere criticism. With construction.
> Take the technical innovations, the emotional range, the sonic possibilities from the handful of worthy artists in these genres. But use them only as substrate. 
> it's in the intervals, the progressions, the timbre itself. Sound as ideology. Melody as statement.
> The goal is music that elevates rather than degrades. 
> We don't need better metal. We need something categorically different - rooted in our own traditions, aimed at our own future.
^Those are all AI statements. Large Language Models talk like this, not humans.

The LLM doesn't know how sound is ideology. It just knows "Abstract Concept A" as "Abstract Concept B" is a sentence template heavily represented in it's training data. It's simply the most probable word gestures for a request. But what LLMs do is, they tend to make as much as they possibly can rhetorical.

The single idea uniting every statement is metaphors. Constant metaphors. A person might make one or two metaphorical references, but an LLM, especially Sam Altman's ChatGPT, is going to stack one after the other. It's a preference of LLMs. Their metaphors are first and meaning is either second or most of the time, never even there.