Zombie theory of literary merit, (part 1)
From the WSJ piece cited in the last post, we take a closer look at Samuel Fitoussi, a young venture capitalist in artificial intelligence who helped expose the Thélyson Orélien affair.
Here's a quick summary of his arguments, from Gemini:
In this context, let me offer my zombie theory of literary merit. It's where an author's subjectivity somehow survive the ravages of the production process, which is by definition collaborative, never mind the passage of time, to emerge intact for the reader's enjoyment.
In an ideal world, the publication of an author's undiluted subjectivity would be accompanied by certified assertions as to its all-human origins; and any misbegotten content which comes to use via AI or some hybrid form (hang on - are typewriters allowed, or could they interfere with the transmission of subjectivity as well?) would be accompanied by a radiation-style warning telling the unsuspecting to keep their distance (see above).
Unlikely? What it's really about, I suspect, is a simple fear among some observers of machines and what they can do.
GPT on the zombie theory of literary merit:
Next, here's GPT on claims by the pro-subjectivity brigade that while much of Manga production is outsourced, it still respects the difference, supposedly, between concept and execution:
Me: 
GPT:
ARTIFICIAL INTELLIGENCE EXPOSURE WARNING
This text has been exposed to generative AI.
Possible effects include:
improved grammar
altered syntax
removal of clichés
addition of clichés
unexpected em dashes
excessive use of "moreover"
loss of the author's original voice
restoration of the author's original voice
arguments with the machine
arguments won by the machine
arguments won by the author
uncertain literary provenance
Readers should consult a qualified literary bureaucrat before forming an opinion.
A watermark, GPT says, could become a status symbol, though not one as intended by its advocates:
