"LLMorphism?"

Not at all identical. We are similar, in that the underlying routing topology is much more N-deep-SLM-of-SLM-like for words, hence we think in DSLs.

But beyond that the nodes are not representations of words, but representations of sensory gestalten. So it humans it is maybe a topology of planes, one for words, mapped over one for sensory images in all sense modalities. That is why sensory data structures matter, more than any operations on incorrect underlying data structures, for anthropomorphic computing. This is the main insight from quantitative phenomenology.

Qualia are data structures, and that is the main oversight of most AI programs to-date in the 20th and 21st centuries.