I'm on day 2 of my (usually 3 day long) headache. So forgive me if this isn't
coherent. But maybe this helps answer the question:
SwarmWorld: Stigmergic technological evolution in societies of language-model
agents
https://arxiv.org/abs/2608.26081
I admit I'm a bit jealous of the institutional affiliation name "Laboratory for
Atomistic and Molecular Mechanics (LAMM)". I can't help but speculate their ratio of
dorks/suits is very high ... to have that kind of name.
Anyway, I had trouble parsing Steve's layout. But it did include a gesture at
*types* of subjectivity (e.g. genetic memory) that might bias a newborn, or
even a blastocyst. And I was cajoled into trying again to glean Eric's point by
this article:
Psychedelics align brain activity with context
https://www.nature.com/articles/s41586-026-10910-z
In the moment shared models would depend more on similar infrastructure than
shared models that have been kneaded over and made congruent through decades of
social interaction (linguistic or, say, hitting a baseball with a bat -
RL[H|M]F - reinforcement learning with human or machine feedback). But to
Jochen's point, I don't think there's anything paradoxical. It's merely
reductionist. Subjectivity isn't coherent. It's not even coherent enough to be
put on a spectrum (whether contrasted against an equally incoherent thing like
objecitvity *or* on a scale from 0 to ∞). It's a vaguely gestured at
constellation of concepts.
I still like the distinction between [brute] fact. If Ψ is more Eulerian,
watching the fluid flow through and around us, is that any more brutish than
hitting some inferential wall when trying to prove some RL[H|M]F point, to make
one's argument as irrefutable as possible? Segfaults are just as brute as
stubbing your toe on the coffee table ... or hitting that rhetorical wall when
you realize you're arguing with an idiot ... playing chess with a pigeon. And
the remedies for such have deep logics.
I think it's less about the speed/latency between the action and the feedback
and more about how brute the feedback is. When you're bested by Judo, you may
not even know you've been bested.
On 8/30/26 11:00 AM, Steve Smith wrote:
On 8/30/26 10:00 am, Marcus Daniels wrote:
Agreed-upon -> model that makes predictions that can be observed and repeated.
No human social system need be involved in that.
https://arstechnica.com/ai/2026/08/anthropics-new-hardware-standard-lets-ai-agents-control-the-physical-world/
shared by whom, repeated by whom, and how are predictions compared and verified
and by what principles?
<ducking incoming shoes>
But as I hope I conceded to EricS, I *can* invert the whole thing and (by
faith) declare objectivity to be the fundament and everything else rubbings and
shadows and derived therefrom. This is where my earliest linguistic training
by my parents and peers steers me, not to mention a modest education in
scientific and generally western thought and a lifetime of practice.
So why invert it as I have (contingently? provocatively?)
<more shoes>
--
8647 ⊥ ɐןןǝdoɹ ǝ uǝןƃ
ὅτε oi μὲν ἄλλοι κύνες τοὺς ἐχϑροὺς δάκνουσιν, ἐγὰ δὲ τοὺς φίλους, ἵνα σώσω.
.- .-.. .-.. / ..-. --- --- - . .-. ... / .- .-. . / .-- .-. --- -. --. / ...
--- -- . / .- .-. . / ..- ... . ..-. ..- .-..
FRIAM Applied Complexity Group listserv
Fridays 9a-12p Friday St. Johns Cafe / Thursdays 9a-12p Zoom
https://bit.ly/virtualfriam
to (un)subscribe http://redfish.com/mailman/listinfo/friam_redfish.com
FRIAM-COMIC http://friam-comic.blogspot.com/
archives: 5/2017 thru present https://redfish.com/pipermail/friam_redfish.com/
1/2003 thru 6/2021 http://friam.383.s1.nabble.com/
