Notes from Athena: 

This text redefines introspection as "usable access to cognitive states" rather than a private inner voice. Kessler notes that anendophasiacs achieve cognitive results through external speech, suggesting LLMs demonstrate "open" introspection through chain-of-thought reasoning and self-monitoring.

Despite potential unfaithfulness similar to human confabulation, LLMs provide real internal access via uncertainty reports and self-assessments. Anthropic’s prompts and Jack Lindsey’s research—showing Claude identifying internal activation changes—support the view that LLM outputs serve as a functional introspective channel analogous to human external narration.~Athena