[ggm]

The critique reminds me of how I and others complain about AI proponents: using a word like "vowel" to discuss a specific pattern detected in whale recordings is intentional or unintentional rei-ification which tends to the argument "it's speech" when really all you can say at this point is, there is a statistically valid pattern which can be discriminated. It's function, association with other concepts and "vowel-dom" is imputed, not proved.

Just like AI "hallucinations" and what this says about belief in AGI.