Well, there is this, though I'm sure there's newer stuff:
Enhancing AI-Assisted Group Decision Making through LLM-Powered Devil's Advocate https://doi.org/10.1145/3640543.3645199 The interesting contrast (IMNSHO) is "consensus-oriented dialectical inquiry" versus what I'm calling "agonistics". The former leans hard on pro versus con. But that seems so dualist to me. The latter tolerates many different perspectives in and around the same (perhaps vaguely specified) constellation of targets. For example, we can take a look at how Debian recently voted on the use of AI in its development. https://www.debian.org/vote/2026/vote_002#quorum """ Option 1 "Ban LLM contributions from Debian via Social Contract" reached quorum: 144 >= 48.4896896257338 Option 2 "Allow AI-Assisted Contributions with conditions" reached quorum: 267 >= 48.4896896257338 Option 3 "Reject LLMs as far as practical, update Code of Conduct" reached quorum: 176 >= 48.4896896257338 Option 4 "Accept AI contributions for Debian specific work" reached quorum: 259 >= 48.4896896257338 Option 5 "Responsible Use of Generative AI" reached quorum: 281 >= 48.4896896257338 Option 6 "A cautious approach to generative AI" reached quorum: 276 >= 48.4896896257338 Option 7 "Debian is created by humans" reached quorum: 213 >= 48.4896896257338 Option 8 "Avoid the use of LLM: climate destruction is a deal breaker" reached quorum: 225 >= 48.4896896257338 """ This isn't clearly separable into us vs them. Each proposal has subtle differences from each other proposal. Attempts to reduce it to pro vs con fail. And I know the academy uses "dialectic" in a general sense. It still irks me. Similarly, "stakeholder" is a meme whose infection vector is carried by the Suits. But it gets at the right sense of a rich ecology of agents, each with its own agenda. On 9/1/26 7:03 AM, Nicholas Thompson wrote:
Hi Glen, I am deep in meteorology right now so I hope you won’t be offended if I give you an iPhone-Ish quick response. Why not a dialectic between a love of convergence and a fascination with the, a dialectic that never settles? I believe you computer folk have a name for that? Nicholas S. Thompson Emeritus Professor of Psychology and Ethology Clark University [email protected] <mailto:[email protected]> https://wordpress.clarku.edu/nthompson <https://wordpress.clarku.edu/nthompson> https://substack.com/@monist <https://substack.com/@monist> On Tue, Sep 1, 2026 at 7:57 AM glen <[email protected] <mailto:[email protected]>> wrote: https://www.aaup.org/academe/issues/fall-2025/seven-theses-against-viewpoint-diversity <https://www.aaup.org/academe/issues/fall-2025/seven-theses-against-viewpoint-diversity> OK. I admit I've been a bit guilty in my emphasis on diversity. But I am aware of Goodhart's Law. And given thesis 2, here, I think I'm going to stop calling myself a "methodological pluralist" and use "instrumental pluralist" instead. But 1) what Siraganian says here is simply wrong: "To promote viewpoint diversity is to be thrilled to have my views in the mix, even if you think they are wrong, because either (a) they contribute to more viewpoint diversity (the more views you have, the more viewpoint diversity there is), or (b) they provide more views to choose from to figure out which view is the most true. In the first position, you end up not caring about whether a particular view is true or good, because you care about viewpoint diversity intrinsically: What matters is having the greatest possible variety of views. In the second position, you instrumentally value a diversity of views; having viewpoint diversity only matters until you can figure out which view is the best or most true, at which point all the other views can be rejected." If the "best" or "most true" is relatively smooth from the 1st best to the 2nd best, 3rd best, etc., then NO, you do not reject "all the others". Such smooth grading has to remain contingent, defeasible. And my guess is the gradation isn't smooth everywhere. It may be smooth for the first 2 or 3 distinguishable ideas, but #4 might be batsh¡t. (cf. SynthID<https://arxiv.org/abs/2510.09263 <https://arxiv.org/abs/2510.09263>>) 2) Diversity remains a useful measure of the crosstrophic nature of complex systems, circularity, industrial symbiosis, etc. So even in light of Goodhart's Law, diversity is a reasonable *outcome*, even if it's dangerous to optimize to that sole outcome. There are issues with all the theses she lays out. But persnickety nits aside, her gist seems right. And you can't overlook the blatant hypocrisy of objecting to DEI initiatives, but then arguing for "intellectual pluralism", pointed out in thesis 7. This post is too long already. But I know many, here, have succumbed to Haidt glazing. So I offer this as further demonstration of Bad Faith rhetoric: https://assignedmedia.org/breaking-news/jonathan-haidt-social-contagion-rogd-pbs/ <https://assignedmedia.org/breaking-news/jonathan-haidt-social-contagion-rogd-pbs/>, only peripherally about instrumental pluralism.
-- 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/
