I hope it's okay that I pulled this library in — it's much easier on the eyes than my off-the-cuff spectrum (live version: pletcher.github.io/ccc-2026-web...)

David Mimno@dmimno.bsky.social · last mo.

This vis got me thinking about ways to show a quantitative variable on a text block. Here's a pure CSS library that provides some options (fill, like this; marks in the margin; angled texture; an underline "bar plot"; etc) and picks up a value from a <div> element. mimno.github.io/TextVariable...

For an upcoming conference presentation: a heatmapped version of Greek tragedy to show how "epic" or "tragic" each line is (by averaging log likelihoods of lemmata appearing in each genre)—the Tutor's speech in S. Electra is particularly exciting to see in context pletcher.github.io/ccc-2026-web...

Work in progress: quantifying work/register distinctiveness of lemmata in Homer. One interesting result: "god" [θεός] and "human" [ἄνθρωπος] are much more likely to appear in speech than in narrative. HT www.dices.uni-rostock.de (for speech references) and github.com/francescomam... (for treebanks)

It’s weird that Helen of Troy being played by a black actress is a problem, but most of the rest of the Mediterranean, Greek and Turkish characters being played by incredibly fair and blond haired people of British, Irish and Scandinavian descent isn’t. And by weird, I mean entirely predictable.

I asked students to choose two Programming Historian (programminghistorian.org/en/lessons/) lessons beyond those assigned in class. So many lessons broken by Python and/or dependency updates—maybe we should prioritize platform stability? (I *love* PH; I do not love how frequently Python libs break)

Here we go again. Every so often tech makes an incursion into history, completely ignores that history is a rigorous discipline with a methodology, and then claims that a mystery or issue that historians have grappled with can be solved with an algorithm. www.bbc.co.uk/news/article...

Scientists believe they have found previously unknown sketch of Anne Boleyn

A computer science team believes they have discovered a previously unknown sketch of King Henry VIII's second wife - but not everyone is convinced.

bbc.co.uk

"criticizing AI use in journalism or other creative industries is classist" NO IT IS NOT there is a class that benefits enormously from cognitive surrender, epistemic warfare, and the enclosure of all hitherto produced culture and it is NOT THE WORKING CLASS

Related to my last post: has anyone taught a digital humanities class that banned student laptops? #digitalhumanities (is that hashtag the right bat signal?)

Revisited repl.it to evaluate using it to teach an Intro DH class. It’s nuts how AI enshittification (which is just three MBAs in a trench coat, i.e. standard mismanagement) has ruined a once-useful product. You can’t even start a REPL without being bombarded by prompts to use their chat agent 🫠

Replit – Build apps and sites with AI

Build and deploy software collaboratively with the power of AI without spending a second on setup.

repl.it

I mean, he's right that American defense strategy going forward is going to consist of subordinating ourselves to our traditional enemies in order to secure the subsidies to win internecine and domestic struggles.

Paul Johstono@profpaulj.bsky.social · 6mo ago

And then Spartan Oliganthropia enters the chat

Also, the entire vocational case for <not> majoring in the humanities/liberal arts rests on approximately no data at all:

Mark Copelovitch@mcopelov.bsky.social · 3y ago

👇🎯 Periodic reminder that the right-wing career/salary talking points about the liberal arts are belied by huge piles of data (e.g., news.wisc.edu/new-data-on-...), & the complicity of Democrats & university administrators in repeating those right-wing talking points is a major part of then problem.

Announcing—LatinCy Readers v.1.0.2, i.e. LatinCy-powered corpus readers for Latin text collections. Quickly get sentences, lines, words annotated for lemma, POS, morphology, NER, etc. Supporting .txt, .xml, .tess, .conllu, and more. github.com/diyclassics/... #digiclass #nlproc

The dj on the MIT station just went on a gorgeous tangent about how this semester he’s taking all liberal arts classes, and instead of asking him to reproduce established knowledge (as in most MIT courses) they push him to think about how he thinks. My Super Bowl MVP.

This and the discussion about how academics can now just read AI summaries of things really drives home how much these guys always confuse the product for the goal. You don't do a coloring sheet because you want to have a colored sheet at the end!!! The coloring is the point!!!

Parker Molloy@parkermolloy.com · 8mo ago

This is just sad.

Excited to release the new minimal-computing Pausanias Reader in Progress: pausanias.opencommentaries.org/about (If you find an empty comment, that's the "in Progress" part) Enormous thanks to @eltonteb.bsky.social and the rest of Digital Periegesis team for their help with named entity linking!

I'm thrilled to report that the Perseus Digital Library has been selected for a 2025 Humanities and AI Virtual Institute (HAVI) award from @schmidtsciences.bsky.social! Read more about our plans here (sites.tufts.edu/perseusupdat...). .

I'm honored to be part of the team alongside @alicatlibrarian.bsky.social, @pnadelofficial.bsky.social, @charlespletcher.org, Lisa Cerrato, Gregory Nagy, Riccardo Strobino, Clifford Wulfman and Ph.D. candidates Sarah Abowitz and Si Wu.

“Beyond Translation: Opening up the Human Record” is among the HAVI projects awarded by @schmidtsciences.bsky.social. Congratulations to PIs @philologistgrc.bsky.social and @dasmiq.bsky.social!

Schmidt Sciences@schmidtsciences.bsky.social · 8mo ago

JUST ANNOUNCED: Schmidt Sciences has awarded $11M to 23 teams globally for our Humanities and AI Virtual Institute (HAVI). These projects bring AI to history, archaeology, literature, and film, unlocking new understandings of human culture. Learn more: buff.ly/tvTUYwY

Who “writes” an LLM-translated work of literature? Even if it isn’t a 1:1 copy, is an AI translation at scale just pure plagiarism? Any middle ground between stealing and creating new resources? (maybe relevant article: www.theguardian.com/books/2024/m... — Venuti probably relevant here too)

AI translation: how to train ‘the horses of enlightenment’

Translation algorithms have greatly improved in recent years, but can they work on literature? Human practitioners of the art are not convinced

theguardian.com

This is how news/media got around for most of human history. Sophocles' _Women of Trachis_ provides something of a thought experiment in what happens when the traditional avenue for news delivery becomes unreliable (hint: it ain't good)

tisane al-gaib@dallaninvictus.bsky.social · 9mo ago

the thing about social media I keep saying is that it makes EVERYBODY into a potential media outlet and some of those media outlets gained reach/influence comparable to the traditional media, and gained it far faster than they gained any consciousness of the responsibility that should come with it

I think the belief by the powerful that "people will put up with many things if you are excellent at math" is incredibly revealing in terms of understanding How We Got Here

From today's Chronicle of Higher Ed briefing. I am *never* an advocate of cutting programs. But I am curious to see if there will be the same type of "students aren't majoring in this, so let's cut the program" discourse around computer science as there always is for the arts and humanities.

lol

Unusual Whales@unusualwhales.bsky.social · 9mo ago

BREAKING: OpenAI is requesting US government support to help guarantee financing for the massive investments in AI chips and data centers it needs for expansion, per Bloomberg.

Here's another bit of conventional wisdom Zohran completely blew away: His speech was at well above a 10th-grade level. It was complex, erudite, punctuated by deep and fluent references. You don't have to condescend to voters with baby talk! Part of re-establishing norms is speaking like an adult.

The <?xml-model?> processing instruction found in many TEI documents—which never advanced past Working Group Note www.w3.org/TR/2012/NOTE... —breaks some XML parsers (Non-standard processing instructions cannot include the string "xml": www.w3.org/TR/xml/#sec-pi) So why are we still using it?

the power of uh, reading books

Sharon 🪳🌹@sharonk.bsky.social · 9mo ago

a genuine problem is that a lot of this stuff is just completely alien to the mass media elite like who else besides weirdos like me take time to analyze what the intellectual righties are saying

Cleaned up the previous sentence embeddings comparisons—cleaner calculation of similarity for tragedies, and now both charts use the same scale (in autumn hues 🍂) It makes sense that tragedies overall are v. similar, but their similarities alone don't explain the lower-right messenger cluster

I might need to update the original plot, but so far it's interesting to note that the messenger similarities don't seem to be a result of consolidation across tragedies, which remain quite distinct (at least from the perspective of sentence embeddings) #dh #ancmedsky

Charles Pletcher@charlespletcher.org · 9mo ago

So sentence embeddings of messenger speeches in Attic tragedy show a modest but notable bump in semantic similarity toward the end of the 5th century (see attached image) — generic consolidation, maybe, but why? #ancmedsky #dh