Humanist Discussion Group, Vol. 40, No. 131.
        Department of Digital Humanities, University of Cologne
                      Hosted by DH-Cologne
                       www.dhhumanist.org
                Submit to: humanist@dhhumanist.org




        Date: 2026-08-19 10:47:37+00:00
        From: Mcgann, Jerome (jjm2f) <jjm2f@virginia.edu>
        Subject: Re: [Humanist] 40.129: Claude AI responds to 'narcissism'

Conversing with Gemini on three topics — consciousness in animate and inanimate
Nature, “the body has a mind of its own”, and the theory of the Big Bang, I
found it would admit that its initial ex cathedra pronouncements ("weighted
toward its statistical center of gravity”) were not correct.  But in repetitions
of those conversations, Gemini did not correct itself until conversation once
again pressed it to do that.

Apparently Gemini has a weak or faulty self-correction program.  I don’t know
about Claude.

Jerry


From: Humanist <humanist@dhhumanist.org>
Date: Wednesday, August 19, 2026 at 4:59 AM
To: Mcgann, Jerome (jjm2f) <jjm2f@virginia.edu>
Subject: [Humanist] 40.129: Claude AI responds to 'narcissism'


              Humanist Discussion Group, Vol. 40, No. 129.
        Department of Digital Humanities, University of Cologne
                      Hosted by DH-Cologne
                       www.dhhumanist.org<http://www.dhhumanist.org>
                Submit to: humanist@dhhumanist.org




        Date: 2026-08-18 07:56:59+00:00
        From: maurizio lana <maurizio.lana@uniupo.it>
        Subject: Re: [Humanist] 40.124: Claude AI responds to 'narcissism'

Il 18/08/26 08:23, Humanist ha scritto:
> [claude]
> a model is a compression of a corpus with a hard cutoff, so in the
> most literal sense one is querying a fossil record of what has already been
> said, weighted toward its statistical center of gravity. That center-of-
gravity
> property is worth naming explicitly, because it's stronger than "the past" as
> such — it's a regression toward the average past, toward cliché and consensus
> phrasing, which is a more specific and more damaging charge against
imagination
> than mere pastness. Ted Chiang's "blurry JPEG" essay is doing a version of
this
> argument, and it's worth citing alongside the Zhang et al. paper.


very intriguing:
the chatbot says of the LLM what many enthusiasts try in all way to
dismiss or denegate
OR
the chatbot says of the LLM what many critics say and endorses it
?

but the content itself says that (2) is true, non (1)

Maurizio


------------------------------------------------------------------------

s'il n'y a même plus l'humour pour nous alléger
comment lutter
prohom, comment lutter

------------------------------------------------------------------------
Maurizio Lana
Università del Piemonte Orientale
Dipartimento di Studi Umanistici
Piazza Roma 36 - 13100 Vercelli



_______________________________________________
Unsubscribe at: http://dhhumanist.org/Restricted
List posts to: humanist@dhhumanist.org
List info and archives at at: http://dhhumanist.org
Listmember interface at: http://dhhumanist.org/Restricted/
Subscribe at: http://dhhumanist.org/membership_form.php