New preprint on "Computational Hermeneutics," co-authored by too many people to list in one post. TL;DR: GenAI is a cultural technology, and needs to be evaluated in ways that recognize situatedness, plurality, and ambiguity as the conditions of meaning — not noise to be minimized.
Wut
When I started playing with LLMs, one of the first books I reread was The Three Faces of Hermeneutics. I had read the technical papers, such as Attention is All You Need, but my understanding felt unbalanced. So I went back and read Heidegger, Habermas, and Derrida.
Indeed, it is in personalization of individual, user-adopted models that I see the future. One-size-fits-all methods hold everything back.
they speak my language, context, ambiguity, hermeneutic circle, great stuff, have to read
Oh Steve Benford, from Nottingham!
Looks amazing! Right in my area of interest. Can’t wait to read it further.
I think I said as much in 2017: posr.org/w/images/c/c...
Final was 2017, draft was 2016. Final is here: ieeexplore.ieee.org/document/794...
I’m glad we’re all going in on the phrase computational hermeneutics!
By the way, we try to make the point in the new paper that we can read performance measures relationally wrt to the different geometries they select for in latent spaces as a way to render their interpretive dispositions visible. bsky.app/profile/zach...
Yes — what you’re doing is actually computational hermeneutics in the literal sense of the phrase! To be candid, I suggested linking to your earlier slide presentation, because it’s a closely related.
There are actually about five different intellectual projects going on now that could be called “computational hermeneutics”
Do you have a sense of where this stuff is being submitted for publication? I’m still trying to figure out where our paper has a decent shot at being appreciated.
We are also still trying to figure that out! Will DM you because I'm not sure whether Cody wants that fully public.
Wait. Kirsten Ostherr, my buddy from Brown? Love it when worlds collide…
@tedunderwood.me excellent piece
this seems further support for the idea that genai is best understood - as engaging in deeply relational processes (natch) - under modal rather than binary reasoning
Authors I can find here include @mariaa.bsky.social , @mmvty.bsky.social , @evelyngius.bsky.social , @hoytlong.bsky.social , @mrsbunz.bsky.social , @jeddobson.bsky.social , @richardjeanso.bsky.social , and @mattwilkens.bsky.social.
This might be of interest to some ..... makingsciencepublic.com/2025/09/05/u...
Thanks! Yes, it definitely will be. I’ll share it with the other authors who aren’t here.
Thank you. I really am an amateur in this field, but I was just intrigued.
and @ruthahnert.bsky.social , and @yalidu.bsky.social , and me
holy shit murderers row
The whole project was organized by Cody Kommers and Drew Hemment, who are not on bsky afaik. It’s related to this initiative: bsky.app/profile/vais...
Really skimmed through this way too quickly Seem like good ideas, but people with a technical bias might prefer the status quo with respect to current benchmarks. But the humanities community and frontier labs, to be more inclusive, could benefit from these suggestions so... - A Fascinated Nobody
