[FRIAM] bigly

[Steve Smith]

On 12/16/25 9:45 am, Marcus Daniels wrote:
> As with everything about the TACO President, the mask he chooses tells us something about what's under that mask. He pretends at strength because he's inherently weak.

ah... the MASK is the message!

I've been assuming for some time that LLMs' superpower is their ability 
at rhetoric in the extreme...  that everything they say to me is in some 
sense, rumour, innuendo, gossip, even (especially) if it is highly 
persuasive and grounded deeply in fact(oids), most of which might even 
be ob(intersub)jectively true.  I liken the "factoids" to being *control 
points* in a spline curve generator...

So as I read your *bigly* business and reflected on the state of 
discourse, discussion, argument today (Trump 2.0 as well as GPT 5.x and 
2+ decadal conditioned FriAM collective era) I decided to go to (one of) 
the horse(s) mouth(s) and see what the pattern of tooth-grindage really 
looked like  from inside (as best as one can get their whole big head 
inside of a horses big mouth).

This is what GPT 5.2 gave me as a  well crafted, highly motivated, 
elaborately stated, empathetically implied "rumour":


            2. Why rhetoric is structurally baked in

        An LLM:

          *

            Predicts /what comes next/ given a context

          *

            Optimizes for *coherence, plausibility, and relevance*

          *

            Has no privileged “view from nowhere”

        That means:

          *

            It selects among many valid continuations

          *

            Based on implicit norms of explanation

          *

            Conditioned on the /implied audience/

        That is rhetoric at the level of *selection pressure*, not
        ornamentation.



LOVED the "view from nowhere" and "conditioned on the implied audience"

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://redfish.com/pipermail/friam_redfish.com/attachments/20251216/f1a9c181/attachment.html>