[TeMPOraL]

I think the core insight OP may be looking for is that your dictionary is just an illusion - that concepts being related to other concepts to various degree is all that there is. The meaning of a concept is defined entirely by other concepts that are close to it in something like a latent space of a language model.

Of course humans get to also connect concepts with inputs from other senses, such as sight, touch, smell or sound. This provides some grounding. It is important for learning to communicate (and to have something to communicate about), and was important for humans when first developing languages - but they're not strictly necessary to learn the meanings. All this empirical grounding is already implicitly encoded in human communication, so it should be possible for an LLM to actually understand what e.g. "green" means, despite having never seen color. Case in point: blind people are able to do this, so the information is there.

[cymian]

Blind people are no more able to understand* (as qualia) "green" than a sighted human is able to understand* gamma rays. The confusion is between working with abstract concepts vs an actual experience. A picture of bread provides no physical nourishment beyond the fiber in the paper it is printed on.

In an abstract space (e.g. word vectors, poetry) green could have (many potential) meanings. But none of them are even in the same universe as the actual experience (qualia) of seeing something green. This would be a category mistake between qualia-space and concept-space

understand in the experiential, qualia sense.

https://en.wikipedia.org/wiki/Qualia

https://en.wikipedia.org/wiki/Category_mistake

[ben_w]

I don't need the qualia of gamma rays to understand gamma rays, nor to be understood in turn when I say that "I understand gamma rays".

Conversely, I can (and do) have qualia that I do not understand.

The concept of qualia is, I think, pre-paradigmatic — we know of our own, but can't turn that experience into a testable phenomena in the world outside our heads. We don't have any way to know if any given AI does or doesn't have it, nor how that might change as the models go from text to multimodal, or if we give them (real or simulated) embodiment.

[foldr]

>that concepts being related to other concepts to various degree is all that there is

This is the view that Fodor termed "inferential role semantics". https://ruccs.rutgers.edu/images/personal-ernest-lepore/WhyM...