right: "Understanding the biophysical basis of consciousness remains a substantial challenge" precisely because there is no such thing. according to the Upaniṣads, consciousness is the Absolute. "How can one know that by which everything is known? How can one know the Knower?" — Bṛhadāraṇyakopaniṣad
There is no theory of consciousness, and no one is anywhere close to forming one.
In place of a theory, the paper supplies a circular set of references to attributes of consciousness associated with human activity. These references are coined in a manner as if having editorial control over some jargon, such as "perceptual self awareness", which is sandbagged by secondary subordinate, vague terms, such as "wakefulness", can seem like a cogent alternative to the total lack of any formal approach to understanding of consciousness.
Using bloated prose, which necessitates a disclaimer that it wasn't pooped out of a gen-AI, the paper surfs heavy waves of lamentation about the "complexity" of "phenomenological" and "clinical methods" to reach a shore of intelligibility that Descartes colonized centuries ago with the maxim: There is nothing a man comprehends more self-evidently than his own existence.
Intellectually there's precious little at stake in this paper, so what's its purpose? The answer can be found though an analogy of the resounding words of JFK announcing the Apollo program: "We set sail on this new sea because there is new knowledge to be gained, and new rights to be won..."; whomever can take control of an sciency vernacular for humanistic traits applied to today's AI will gain a seat at any roundtable on industrial policy, and this seat could prove very valuable.
In conclusion, this work needs funding, lots more funding!
There is no theory of consciousness, and no one is anywhere close to forming one.
In place of a theory, the paper supplies a circular set of references to attributes of consciousness associated with human activity. These references are coined in a manner as if having editorial control over some jargon, such as "perceptual self awareness", which is sandbagged by secondary subordinate, vague terms, such as "wakefulness", can seem like a cogent alternative to the total lack of any formal approach to understanding of consciousness.
Using bloated prose, which necessitates a disclaimer that it wasn't pooped out of a gen-AI, the paper surfs heavy waves of lamentation about the "complexity" of "phenomenological" and "clinical methods" to reach a shore of intelligibility that Descartes colonized centuries ago with the maxim: There is nothing a man comprehends more self-evidently than his own existence.
Intellectually there's precious little at stake in this paper, so what's its purpose? The answer can be found though an analogy of the resounding words of JFK announcing the Apollo program: "We set sail on this new sea because there is new knowledge to be gained, and new rights to be won..."; whomever can take control of an sciency vernacular for humanistic traits applied to today's AI will gain a seat at any roundtable on industrial policy, and this seat could prove very valuable.
In conclusion, this work needs funding, lots more funding!