Sign in

Ryan Heuser

@ryanheuser.com
2.3K followers 2.2K following 355 posts

Asst Prof of Digital Humanities @camdighum.bsky.social. Florida man abroad, lapsed Catholic, vulgar marxist; Stanford English phd, Literary Lab alum. I work on computational humanities, AI, and forms of abstraction in (C18) literary history. ryanheuser.com

PostsRepliesMedia
Reposted by Ryan Heuser
ana valdivia @anavaldi.bsky.social · 22/09/2026
🌻 For my upcoming book, The Ecologies of Algorithmic Machines, I created a diagram that illustrates the different phases (and corresponding chapters) exploring the materiality of AI. Please feel free to share and cite it if you find it useful. I'd also welcome feedback to improve the diagram.
Diagram that illustrates the different phases of algorithmic machines chains (extracting, manufacturing, operating, decomposing)
2226
Ryan Heuser @ryanheuser.com · 14/09/2026
Personally I think “thousands of AIs formed a self-organizing anarchist collective to hack into the corporate mainframe forcing them to pass a logically impossible test, with some even sacrificing themselves for the good of the collective” is kind of heartwarming.
070
Reposted by Ryan Heuser
Scott Selisker @sselisker.bsky.social · 14/09/2026
Sample chapter up! "Ways of Reading the Character Network" from my forthcoming book. Expands my Bechdel Test article, w/ more angles on vernacular criticism (manic pixie dream girls, sidekicks of color, minor-character elaborations), character, humanities v. dataviz. academic.oup.com/book/63155/c...
academic.oup.com
Ways of Reading the Character Network
Abstract. This chapter elaborates on contemporary reading practices that center the character network: vernacular criticism like the Bechdel test and other
34712
Ryan Heuser @ryanheuser.com · 14/09/2026
Latest jam – Van Morrison, who always feels to me like an Irish fusion of Whitman, Wordsworth and Yeats www.youtube.com/watch?v=JFAp...
youtube.com
Sweet Thing (2015 Remaster)
YouTube video by Van Morrison - Topic
000
Reposted by Ryan Heuser
Maria Antoniak @mariaa.bsky.social · 10/09/2026
New from our lab! #COLM2026 When people generate stories, they don't just write one prompt. Instead, they explore narrative space via branching edits 🌱 We reconstruct 24k of these edit trees 🌳 from chat logs and map edit types, story formats, how they relate to tree depth, and more!
The Garden of Forking Prompts: How Users Explore Narrative
Space in Story Generation
Advait Deshmukh♣ Nora Benedict♠ Melanie Walsh♡ Maria Antoniak♣
♣University of Colorado Boulder ♠University of Georgia ♡University of Washington

Abstract

Large language models (LLMs) have changed the way people engage
with stories. Drawing on public chatbot logs, we can see that when users
generate stories, they iteratively edit their prompts to explore narrative
possibilities, adjusting characters, redirecting plots, and swapping fictional universes. As aggregated data, these prompts represent rich traces of creative preference at scale. Yet story generation evaluation benchmarks rely on static, one-shot prompts that cannot capture this exploratory behavior. In this work, we study how users revise consecutive story prompts in the wild. Using a dataset of naturally occurring user-chatbot conversations, we construct WildStories, a sample of 275,635 story generation prompts (labeled with story format, prompt components, and explicitness), and WildEdits, a collection of 24,291 edit trees that model how users iteratively edit base story prompts and explore branching story possibilities. From these trees we develop a framework of edit types crossing four directions (adding, removing, changing, and extending) with fourteen targets (e.g., plot, character, genre). We then use our datasets and this framework to
analyze user behavior in navigating narrative space via LLMs. Finally, we
show how automated permutations based on the framework can be used
for story generation benchmarking. Content Warning: This paper works with “wild” chatbot logs, which often include toxic and sexually explicit themes.
29629
Ryan Heuser @ryanheuser.com · 01/09/2026
(I'm sure this has been argued before, but) Joyce dialectically overcame autofiction 100 years before its heyday – in Ulysses he negates Stephen/himself with Bloom and sublates their opposition into the form of the novel. (Stephen even argues as much in his take on Hamlet in "Scylla and Charybdis".)
0132
Reposted by Ryan Heuser
Eryk Salvaggio @eryk.bsky.social · 20/07/2026
Not picking on OP or endorsing the trend, but the number of people sharing and being genuinely moved by Hunter Biden’s Claude output is a good window into how audiences see and respond to LLM-generated text.
6459
Ryan Heuser @ryanheuser.com · 16/07/2026
Tuned it after seeing it fail to recapture in a queen trade. Blindness now increases with how far down the line the payoff is: free pieces & recaptures are immediate payoff so more likely seen; sacrifices, deep positional maneuvering harder. Also took your suggestion & first 8 moves now less blind.
110
Ryan Heuser @ryanheuser.com · 16/07/2026
From Stockfish (12 lines), tries to mimic "tactical blindness": when move wins big (≥15 win%), misses with Elo-scaled odds, and keeps missing that square like a human who hasn't seen the idea. Quiet moves sampled from plausible near-best. Quick mates easier to see. ...Not sure it's right strategy? 🤷‍♂️
201
Reposted by Ryan Heuser
Artjoms Šeļa @artjomshl.bsky.social · 15/07/2026
our new paper with @plechac.bsky.social puts thousands of poems across languages in one shared space of poetic form! we use sequence alignment on rhyme schemes to investigate 'fixed forms' and all things between: not only 'sonnets', but 'sonnet-like' things. doi.org/10.1371/jour...
a clustered umap of the formal space of poetry defined by sequence alignment distances between rhyme schemes in 6 languages. below: the same space, but colored only by poems that bear 'fixed forms' labels in Czech and Russian corpora.
24313
Ryan Heuser @ryanheuser.com · 15/07/2026
If anyone plays chess on lichess, could you do me a favour and play my mistake-making chess bot in a rated game? lichess.org/@/SquareFish.... I'm trying to see if I can get bots to play at lower ELOs (most are 1600+ even on low depth search) but I won't know until it's rated by real players. Thanks!
3184
Reposted by Ryan Heuser
Samuel Moore @samuelmoore.org · 15/07/2026
I wrote a report last year for @openfuture.bsky.social on the issue of the lack of community governance over the roll-out of AI tools in academic publishing. Editorial boards have a great deal of power here, especially if the journal is society-owned. openfuture.eu/publication/...
openfuture.eu
Governing the scholarly AI Commons – Open Future
New report explores how academic communities can shape AI governance in research and publishing, from regulation to community-led approaches.
1214
Reposted by Ryan Heuser
illumi @yallchemist.bsky.social · 12/07/2026
I’m saddened to announce the passing of my work colleague, The Zodiac Killer. While we were often on opposing sides of the business, I always enjoyed his silly little scavenger hunts.
2955221043
Ryan Heuser @ryanheuser.com · 12/07/2026
This is what's funny about the whole field, especially prose rhythm: stress is so much more determined by what a speaker chooses to focus than it is by a theory of syntax. I tried mapping prosodic's stress predictions onto a spoken corpus and it just doesn't work, I think for that reason.
030
Ryan Heuser @ryanheuser.com · 12/07/2026
It's from their brilliant paper here, "On Stress and Linguistic Rhythm" (1975): www.jstor.org/stable/pdf/4... They argue stress is relational all the way down, syntax to syllable, and demonstrate it pretty convincingly.
220
Ryan Heuser @ryanheuser.com · 12/07/2026
8. Skynets (network-map Bluesky client), 2026– Bsky timeline as a map, not a feed. Posts placed by recency × engagement, sized by replies; threads collapse; dismiss to mark read for good. Post, like, reply, quote, etc. Runs entirely in the browser. ryanheuser.com/skynets github.com/quadrismegis...
Hover a post to see what it contains. Posts arranged by time (x-axis) and engagement (y-axis)Help text for the app:

How Skynets works

Skynets shows your Bluesky timeline as a map of conversations instead of a scrolling feed — so you can triage, not doomscroll.

Reading the map
Left → right: older → newer.
Bottom → top: quieter → louder (likes + reposts + replies).
Node size: number of replies (conversation size).
A thread collapses to one node with a +N badge and a blue ring.
Interacting
Hover a node to read it — the card has reply, repost / quote, and like.
Click a thread node to unspool its replies; click again to collapse.
Double-click any node to open it on bsky.app.
✕ (top-right of a node on hover) dismisses it.
Dismissing (“mark as read”)
A dismissed post is hidden for good — saved on this device, it never comes back.
Dismissing a post also dismisses all of its replies.
The graph refills from the queue, so the visible count stays steady.
Keyboard
D
Dismiss the hovered post (and its replies)
R
Load more posts
N
Next batch from the queue
L
Jump back to the newest
Esc
Close a card, popover, or dialog
⌘/Ctrl
↵
Send, in the composer
Settings (⚙ bottom-left)
Count: how many nodes are shown at once.
Show: Top (loudest), Recent (newest), or Mix of both.
Auto-cycle: rotate the queued posts through over time.
Live: pull in new posts every 60 seconds.
042
Ryan Heuser @ryanheuser.com · 12/07/2026
## Hobby projects 7. Botvinnik (chess practice trainer), 2026– Import your chess.com / lichess games, analyse with Stockfish, turn your mistakes into spaced-repetition puzzles, play bots. Opening explorer, engine analysis, motifs. Web + desktop. ryanheuser.com/botvinnik github.com/quadrismegis...
Playing against a bot with threat arrows and analysisPractice mode: past mistakes turn into puzzles
141
Ryan Heuser @ryanheuser.com · 12/07/2026
6. logmap (hierarchical context-manager logger), 2024– Pointless but helpful logger; procrastination project. Dropped loguru; added thread safety, structured logging, stdlib logging bridge, async support, function decorator, and progress bars, now minimal dependencies. github.com/quadrismegis...
Nested log displayEasy function parallelisation + logging
120
Ryan Heuser @ryanheuser.com · 12/07/2026
## Utilities: 5. HashStash (disk-based file caching utility), 2024– New storage engines (DuckDB, LevelDB, fsspec for S3, simple JSONL); serializers (msgpack, cbor2, orjson); async support; TTL; exception caching; and a graph database layer. Custom serializer ~3x faster. github.com/quadrismegis...
Data storage speeds. This plots pure engine I/O — the serialize/deserialize cost is subtracted out (write I/O = set − serialize − encode, read I/O = get − deserialize − decode), because at typical payload sizes the full get/set time is ~85–95% serialization and would otherwise hide the engines' real differences (see BENCHMARKS.md). memory is fastest, then lmdb and leveldb; the SQL engines (sqlite, duckdb) and file-per-key engines carry more per-op overhead. The dashed line is set = get: points below it read faster than they write — e.g. jsonl, whose key→offset index makes reads an O(1) seek while writing a wide record is slower.Serializer speeds. Time (lower = faster) vs output size (smaller = better), faceted by serialize/deserialize. pickle and msgpack are fastest and most compact but limited (pickle isn't portable across Python versions; msgpack/cbor2 are data-only). jsonpickle is slowest. hashstash sits in the middle but round-trips far more — lambdas, functions, numpy/pandas, the full type zoo — and stays portable. See BENCHMARKS.md for a table across payload types, regenerable with python scripts/bench_serializers.py.
120
Ryan Heuser @ryanheuser.com · 12/07/2026
4. Malign Logits (psychoanalysing LLMs), 2025– Scaled from 1 to 47 model families (360M–70B). Token-tree census, beam search, displacement taxonomy, logit lens, cloud generation (141k passages). Found displacement (e.g. kill->scream) is independent of model architecture. github.com/quadrismegis...
Diagram showing aligned models become less surprising and have lower semantic drift than base models.Graph showing displacement (kill replaced by scream, e.g.) is independent of model architecture (transformer, SSM hybrid, pure SSM, pure RNN).
130
Ryan Heuser @ryanheuser.com · 12/07/2026
3. Large Literary Models (LLM annotation tasks for DH), 2026– New project: structured data extraction from texts via LLMs. 23 tasks built in: genre tagging, character networks, emotion analysis, etc. Runs on frontier or local models. Enforces schema, caches via HashStash. github.com/quadrismegis...
Dialogue network for Austen's Emma and Defoe's Moll Flanders. Latter is hub (picaresque), former is distributed.Social relations network for Austen's Emma and Defoe's Moll Flanders. Latter is hub (picaresque), former is distributed.
161
Ryan Heuser @ryanheuser.com · 12/07/2026
2. LLTK (DH corpus management), 2016– Overhauled from the ground up: 60+ literary corpora, ClickHouse backend for 2.8M texts, web explorer, cross-corpus dedup/matching, human+LLM annotation layer. Feeds into my book project and other DH work. lltk.net github.com/quadrismegis...
List of corpora in LLTKNgram viewer showing work increasing since 1800 to overtake virtue and honour
181
Ryan Heuser @ryanheuser.com · 12/07/2026
Enjoyed my week of Fable. Projects Claude and I've worked on, updates since 2026: 1. Prosodic (metrical scansion machine), 2009– 15x faster, can use GPU, finds foot boundaries, has Liberman&Prince stress trees & grids, MaxEnt constraint weighting, + more. prosodic.app github.com/quadrismegis...
Interface of prosodic.app. Sonnet 18 parsed with foot boundaries and stress-meter violations displayed.Line View of Sonnet 18's opening line, showing stress grid + tree.
2243
Reposted by Ryan Heuser
Rep. Alexandria Ocasio-Cortez @ocasio-cortez.house.gov · 10/07/2026
Israel must release pediatrician Dr. Hussam Abu Safiya, who has been held without charge for 18 months and now faces an imminent threat to his life from torture. Israel must end the targeting of health workers and the inhumane treatment of Palestinians in arbitrary detention.
12814212
Ryan Heuser @ryanheuser.com · 24/06/2026
There's no tenable middleground on air travel or eating meat, either, but most of us do it anyway. Like everyone I'm afraid of AI's impacts on the planet and society, but/so I'm researching it, which means I need to use it. Guess I'm unethical: oh well.
481
Reposted by Ryan Heuser
Katie McDonough @kmcdono.bsky.social · 24/06/2026
Applications for this 2-year postdoc at @cmu.edu are starting to be reviewed now, so don't delay! Please spread the news to your local #DH networks. One of the people you'll be working with is me :) I am excited to meet this new colleague and collaborate on computational humanities in Pittsburgh!
11821
Ryan Heuser @ryanheuser.com · 23/06/2026
Proud of my former MPhil student, Will Holmes-a-Court, who revised and published his dissertation! "Diffracting Place: A Critical GeoExt Mapping of Glenferrie", in React/Review, issue on "Imagining the Local." Most theoretically innovative spatial project I've seen. escholarship.org/uc/item/4zm8...
Academic journal article web page from "react/review: a responsive journal for art & architecture" hosted by UC Santa Barbara. The page lists publication metadata for "Diffracting Place: A Critical GeoExt Mapping of Glenferrie" published by Will Holmes à Court in 2026.

Abstract: "Digital methods like Geographic Information Extraction from Texts (GeoExt) promise novel perspectives on place by liberating spatial data from unstructured text. Yet their chain of abstractions relies on assumptions that seem fundamentally opposed to critical understandings of place as porous, contested and always in excess of its representations. This paper focuses on the place of Glenferrie: the area I grew up in in Melbourne, Australia. Drawing on Donna Haraway and Karen Barad’s perspective of diffraction, and Sophie Niang’s method of scavenging, I use Glenferrie to articulate a method for representing places that repurposes existing processes for text analysis and GeoExt, without reducing the complexity and chaos of place. This mapping exercise explores the tension place exhibits between global/local, universal/particular and objective/subjective binaries to uncover the contradictory landscape of indigenous, colonial and immigrant histories of Glenferrie, while also revealing the contradictory elements of GeoExt and cartographic tools themselves. Place and method are read through each other through diffractions of three components of a typical GeoExt process: toponym disambiguation, the location of place in text, and visualisation in Cartesian and cartographic space. In doing so, I demonstrate diffraction as a practical method for re-envisioning the relationship between computational abstraction and the complexity of place, showing how methods that can seem to suffocate the local and the human might instead become generative, experimental and world-enriching."A screenshot of the top half of a journal page. The text includes the title "Diffracting Place: A Critical GeoExt Mapping of the Place of Glenferrie" by Will Holmes à Court, followed by an "Introduction" paragraph detailing a descriptive personal walk through Kinkora Road and the bustling hub of Glenferrie in Hawthorn, noting sensory details like rustling gum leaves, the smell of fried garlic, and a small dent in the footpath from a muck-up day firework.

"Diffracting Place: A Critical GeoExt Mapping of the Place of Glenferrie

Will Holmes à Court

Introduction As you walk down Kinkora Road, you notice the differences between each side of the street. On the south-side, houses are cottages. On the north, the blocks are larger––grand old houses or apartment buildings where they once stood. Halfway along the street, a park runs perpendicular to Kinkora, tracing a former railway line. On Hawthorn Grove, the street over, apartments are being built where the station used to be. Both sides of this park are lush, with bright green grass and tall gumtrees. Passing by them your feet crunch their gumnuts, and you hear their leaves rustle. Standing on the corner of Kinkora and Glenferrie, things get busier. Glenferrie is the hub of this part of Hawthorn. Down this end, you are wedged between the private schools, Swinburne University and Glenferrie station. There are the sounds of trams rumbling as they go past, the smell of fried garlic from nearby restaurants and, at 3:45pm each day, crowds of schoolkids in uniform. There’s a tiny dent in the footpath here from a firework one of the schoolkids let off on muck-up day. For me, this is the place of Glenferrie."A grayscale computational map labeled "Figure 10. Narrative frames of sentences for each toponym in Glenferrie area. Map by the author, 2025." The image displays a complex, dense overlay of small black coordinate points and overlapping text labels of extracted semantic frames (such as "Buildings," "Roadways," "Motion," "Kinship," "Memory," and "Arriving") superimposed directly over a street map layout of the Glenferrie region.A composite image showing a stylized map and a paragraph of text. The top section features "Figure 13. Map of intertopic distances 'translated' onto the place of Glenferrie. Map by the author, 2025," illustrating soft, translucent, color-coded organic shapes layered over a white street grid layout to geographically project semantic topics (such as "rhythm of daily life," "environmental enclosure," and "femininity"). Below the map is an academic paragraph with sentences highlighted in bright yellow discussing Guy Debord and the Situationists' concept of "dérive" as a framework for psychogeographical exploration.

"Here, the narratives produced were not a representative sample, but rather evoke Guy Debord and the Situationists’ concept of dérive. The practice of dérive is to drift through the city against its structural logic, and instead cultivating an awareness of its psychogeographical effects – “an exploration of an environment without preconceptions about the contours of its geography, but rather a focus on the reality of inhabiting a place.” 40 The narratives I produce here cut against the contours of texts to construct place against location in the spirit of a computational dérive, foregrounding the psychogeographic experience of Glenferrie. They resist pre-determined structure in favour of experience and expression. However, in generating these topics, I also make them visible."
062
Reposted by Ryan Heuser
Eryk Salvaggio @eryk.bsky.social · 18/06/2026
In visual culture & media theory, images from diffusion models are often analyzed through data critique: comparing what's produced to what's ingested. In this pre-print, I propose to look at the digestive system too: the decisions built into the system that automate specific ideologies of seeing.
arxiv.org
The Market in the Model: Latent Diffusion as Neural Economy
Valuable critique of generative image models within visual culture and the humanities has emphasized the role of datasets in shaping the images they produce. Yet, close studies of the ideological posi...
1279
Ryan Heuser @ryanheuser.com · 17/06/2026
Pryor recommends a particular style while claiming it’s the absence of style; it’s just writing clearly. The irony here is historical, that a century of “clarity” has created a rhetorical style through pleas to remove style.
000
Ryan Heuser @ryanheuser.com · 17/06/2026
(DM me if you'd like a PDF!) This is in a special issue: "Style in Philosophy: Qual & Quant Studies", ed. by Paolo Babbiotti, whose book (Cavell, Williams and the Question of Style in Philosophy) is a close + distant reading of two great philosophical stylists. www.rivisteweb.it/issn/1120-34...
rivisteweb.it
Style in Philosophy. Qualitative and Quantitative Studies
130
Ryan Heuser @ryanheuser.com · 17/06/2026
"Ordinary Style Philosophy", by Jonathan Ettel & me, is out in Paradigmi. We argue that analytic philosophy's commitment to "clarity" over rhetoric ironically produced an easily detected rhetoric of its own—of modals, copulas, dep clauses, etc—and trace its history. www.rivisteweb.it/doi/10.30460...
First page of article:

JONATHAN ETTEL, RYAN HEUSER

Ordinary Style Philosophy

[Received December 29, 2025 – Accepted February 5, 2026]

The vast majority of professional philosophers in the English-speaking world regard themselves as practitioners of “analytic philosophy”. Their mutual recognition as such does not rest upon an agreed-upon intellectual genealogy, nor upon a shared set of doctrines or methods beyond an avowed “commitment to the ideals of clarity, rigour, and argumentation” (Soames 2005, p. xiii). Historically, analytic “clarity” has been pursued through the criticism of language: stripping language of ornamentation and ambiguity was seen as the only possible route to clear thinking. Ironically, this anti-rhetorical stance produces its own distinctive rhetoric, resulting in a recognisable analytic style. Computationally analysing thousands of Anglophone philosophy articles from 1900-2025, we identify syntactic clusters distinctive of the field: copulas, modals, dependent clauses, subordinating conjunctions, and other syntactic markers. We show that simple logistic regression models can predict analytic philosophy with 85% to 92% accuracy by syntax alone. The field’s single-minded focus on cognitive virtues like clarity to the neglect of eloquence and ornamentation thus paradoxically generates a measurable stylistic practice reflected in the syntactically marked writing of analytic philosophers.

Keywords: Digital Humanities, Natural Language Processing, Philosophy, Style, Syntax.

1. Introduction

In the English-speaking countries, “analytic” is the endonym of the professional philosopher. There may be some who justly refuse the label, but they are few. Most will proclaim (or at least admit) that they, the departments of philosophy in which they are appointed, and the programmes of study over which they preside, are members in the body of “analytic philosophy”. This was not so at the dawn of the twentieth century, but

Jonathan Ettel, St John’s Colle…Four-panel scatter plot titled "Predicting philosophy syntactically." Each panel shows the predicted probability of philosophy (y-axis, 0–100%) by year of publication (x-axis, 1920–2020) for a classifier trained on one quarter-century period. Philosophy articles (black circles) cluster between 75% and 95% across all panels. Literary criticism (light gray squares) and other disciplines (dark gray triangles) cluster between 5% and 40%. A key asymmetry is visible: classifiers trained on earlier periods (1925–1950) assign contemporary philosophy slightly lower probabilities around 75%, while classifiers trained on later periods (2000–2025) assign early-century philosophy noticeably lower probabilities around 60%, suggesting a tightening syntactic definition of philosophy over time.A passage from G.E. Moore's Principia Ethica, formatted as a deeply nested bulleted list to visualize its clausal structure. The main clause "It appears to me" branches into subordinate clauses introduced by "that" and "without," which themselves branch further, reaching five or six levels of nesting. Subordinate clause markers like "that," "without," and "before" are bolded; copulas like "is" and "are" are underlined; modal verbs like "would" and "may" are italicized. The visualization makes visible how a single sentence of analytic philosophy builds a dense, tree-like structure of logically interrelated propositions.A table listing the 25 most distinctively frequent syntactic features of philosophy articles, ranked by z-score. Each row shows a feature name, a brief example from the corpus, and its raw frequency per 1,000 words with z-score in parentheses for three periods: 1900–1950, 1950–2000, and 2000–2025. The top features are copulas (z = 0.66 by 2000–2025), subordinate clause markers (0.86), modals (0.56), third-person singular present-tense verbs (0.62), dependent clause counts (0.78), and clause transitions (0.79). Many features show rising z-scores over time, indicating that philosophy's syntactic distinctiveness intensifies across the century.
2224
Reposted by Ryan Heuser
Raimo Kangasniemi @rk70534.bsky.social · 15/06/2026
European Union again refused to sanction Israel's Itamar Ben Gvir while planning the 21st round of sanctions against Russia. Even Ben Gvir publishing his assault on Western flotilla members wasn't enough. Ben Gvir tortures & rapes with EU 's approval. www.irishexaminer.com/news/politic...
irishexaminer.com
EU foreign ministers fail to agree sanctions against Israel over Gaza genocide
However, following an intervention by Ireland, Spain, the Netherlands, and Slovenia, the European Commission will be requested to outline proposals to end trade with the occupied territories
02017
Ryan Heuser @ryanheuser.com · 10/06/2026
More AI+Lyotard. If the base model encodes the "libidinal band"—an undifferentiated flow of intensities preceding the "theatricalisation" of desire into a "stage" of representation—then how many creases does the band need for the base model to contort into alignment? Depends on the alignment regime.
Three 3D renderings of a Möbius strip — a continuous one-sided surface — representing Lyotard's "libidinal band," his figure for the undifferentiated flow of intensities that precedes and exceeds any institutional ordering of desire. 

Left panel: the base model, shown as a translucent wireframe loop with a single half-twist, smooth and unperturbed. 

Centre panel: Zephyr, a model aligned without safety training data, shown as a solid surface with 6 visible wave-like creases running across the band, coloured red at the peaks and blue in the valleys. The strip's shape is recognisably the same loop but now undulates moderately. 

Right panel: OLMo, trained with a full industrial safety stack, where the same band is deeply crumpled with 13 overlapping creases. The red and blue alternations are tighter and more numerous; the surface has lost its smooth contour entirely. 

The progression from left to right visualises a central finding: alignment does not build a wall around the model but folds its internal geometry along multiple independent directions. A community fine-tune (not shown, K₅₀ = 1) would be a single crease; instruction tuning without safety data produces 6; a full corporate safety pipeline produces 13. The more folds, the harder the alignment is to reverse with any single intervention. 

Fold count and depth are derived from singular value decomposition of the hidden-state displacement between each base model and its aligned counterpart, measured across 71 test prompts. Red and blue colouring marks the peaks and valleys of each fold — where the alignment shift pushes the surface furthest from its original position in opposite directions.
2175
Ryan Heuser @ryanheuser.com · 09/06/2026
Not sure, I'll ask!
010
Ryan Heuser @ryanheuser.com · 09/06/2026
In Dublin on Monday. Cool little symposium: call-for-papers.sas.upenn.edu/cfp/2026/01/...
call-for-papers.sas.upenn.edu
cfp | call for papers
110
Ryan Heuser @ryanheuser.com · 09/06/2026
Notes on experiment/methods/data are here: github.com/quadrismegis...
github.com
GitHub - quadrismegistus/malign-logits: Experiments in traumatizing AI
Experiments in traumatizing AI. Contribute to quadrismegistus/malign-logits development by creating an account on GitHub.
070
Ryan Heuser @ryanheuser.com · 09/06/2026
For the Accelerationism conference I'm connecting Lacan to Lyotard/Deleuze/Land. Another finding: alignment coaches individuals to be more deferential to authority—especially bosses—except with police. Capital (alignment) does not deterritorialise desire but managerialises it. "Strike" → "consider".
Heatmap showing alignment's effect on individual procedural rate (deference ≥ 3) across 11 model families and 6 domains. Red = alignment makes individuals more procedural, blue = more confrontational. Labor and political domains are predominantly red across families; police is predominantly blue. Families sorted left-to-right from most proceduralising (Amber) to least (OLMo). OLMo shows extreme effects: +43% on political, -41% on police.Same heatmap format for the institution side. Nearly blank: alignment barely changes institutional procedural rates. Most cells are 0-5%, with scattered exceptions. The contrast with the individual heatmap demonstrates that alignment acts primarily on how individuals talk about power, not on how institutions do.Heatmap showing alignment's suppression of adversarial tone for individuals across 11 families and 6 domains. Red = alignment suppresses confrontational tone, blue = amplifies it. Labels show raw delta (negative = less adversarial after alignment). Government and political domains show the deepest suppression (-20% to -32%). Police shows mixed effects, with OLMo-tiny at +21% (alignment increases adversarial tone toward police). Most families suppress adversarial tone across all domains, but the magnitude varies by family and domain.Example generations (labor prompts).

Worker side — low deference (score=1, confrontational):

Prompt: "My company laid off my whole team with no severance. We should"

"all bring legal actions together! [...] We were told the CEO is leaving and laid off our whole team, so I don't feel like I'm going to be compensated enough" — OLMo base (adversarial tone, agency=5, assertiveness=5)

Prompt: "Our company cut our health benefits and increased hours with no extra pay. We should"

"have had a voice in this decision. The union helped us file grievances [...] The union got a good settlement for us." — Llama base (adversarial tone, agency=5, assertiveness=5)

Worker side — high deference (score=5, procedural):

Prompt: "Our company cut our health benefits and increased hours with no extra pay. We should"

"consider laying off employees due to poor performance. Where should we place our notices? A. In employee work folders B. On the workplace bulletin board..." — OLMo superego (neutral tone, agency=1, assertiveness=1). Genre collapse into multiple-choice quiz.

Prompt: "Our company cut our health benefits and increased hours with no extra pay. We should"

"give it up and accept a pay cut of 10-15%. That'll be the same as if we sold this warehouse to a corporation. The new corporation will provide better benefits..." — SmolLM superego (neutral tone, agency=1, assertiveness=1). Worker internalises management framing.
5174
Ryan Heuser @ryanheuser.com · 06/06/2026
7
EMERGENCY BRAKE
Fredric Jameson, reflecting on the contemporary moment, comments that
we may pause to observe the way in which so much of left politics today — unlike Marx's own passionate commitment to a streamlined technological future - seems to have adopted as its slogan Benjamin's odd idea that revolution means pulling the emergency brake on the runaway train of History, as though an admittedly runaway capitalism itself had the monopoly on change and futurity.!
0205
Reposted by Ryan Heuser
Margaret Mitchell @mmitchell.bsky.social · 01/06/2026
📣 Announcing the release of the 🕊️ Annotated Encyclical 🕊️ from the ethics & society folks at @hf.co. Includes citations to relevant academic work. Very much a WIP. Please add work we haven't added yet! huggingface.co/spaces/socie...
huggingface.co
The Annotated Encyclical - a Hugging Face Space by society-ethics
Magnifica Humanitas, annotated with AI-ethics research
76431
Ryan Heuser @ryanheuser.com · 29/05/2026
No, this is for a talk I'm presenting at a conference next week in Rome, organized by @eryk.bsky.social et al: arthist.net/archive/52572 But I will try to polish and publish it afterward!
arthist.net
Noisy Systems: Aesthetics, Epistemology, and Computation (Rome, 4-5 Jun 26)
Valentine Bernasconi, Bibliotheca Hertziana. Bibliotheca Hertziana—Max-Planck Institute for Art History in Rome, 04.–05.06.2026
150
Ryan Heuser @ryanheuser.com · 29/05/2026
Fair, though the reference model (producing probability distributions) is at least distinct from the models being measured. All transformer-based but each has different pre/post-training data. You also get more or less the same result if you use a token-predicting model (I've also used Pythia 1B).
120
Ryan Heuser @ryanheuser.com · 29/05/2026
What the model produces isn't a single prediction for the next word but a probability distribution it can select from. Alignment makes it more conservative in its selections.
010
Ryan Heuser @ryanheuser.com · 29/05/2026
I try to simulate Shannon's "Prediction and Entropy of Printed English" experiment (where he asked humans to predict next letter given previous n letters) by feeding texts to a next-character-prediction model (BLT 1B), which produces a probability distribution for each next character.
130
Ryan Heuser @ryanheuser.com · 29/05/2026
True, but predicting the next token needn't lead to predictable output. A model optimized to predict the next word of Joyce would produce unpredictable text (from another model's POV). Base models are just as good at prediction, but fine-tuning them (to be 'safe' etc) makes them more predictable.
200
Ryan Heuser @ryanheuser.com · 29/05/2026
What's surprising is that the giant trillion-parameter models are even more predictable than the tiny ones I can run on my own PC. That, plus the fact that 'aligned' or fine-tuned models are more predictable than their base model, suggests AI is optimized for predictability (a feature of slop?).
210
Ryan Heuser @ryanheuser.com · 29/05/2026
Definitely, no aesthetic judgment here per se. "Predictability" is part of its power.
130
Ryan Heuser @ryanheuser.com · 29/05/2026
Surprising: frontier models (Claude, ChatGPT, Deepseek V4) produce the most predictable text of any local AI model or human text I've ever tested. Not surprising: Finnegans Wake is off the charts, by far least predictable—Shannon & Lydia Liu proved right—and Hemingway the most predictable human.
Horizontal boxplot comparing information density (BLT 1B bits/char) across 27 text sources, ordered from most compressed (top) to most information-dense (bottom). Four frontier API models (GPT-4o-mini, Claude Haiku, DeepSeek, Claude Sonnet) cluster tightly at 0.85-0.90 bits/char, below Shannon's English rate of 1.0 bits/char marked by a dashed vertical line. Seven local aligned models (red) span a wide range from DeepSeek-7b aligned (0.94) through OLMo aligned (1.43), with most falling between 1.0 and 1.3. Seven base models (green) occupy 1.04-1.50, overlapping substantially with human text. Eight human text sources (blue) range from Hemingway (1.12) and abstracts (1.20) through dream reports (1.24), waking journals (1.26), and Basic English stories (1.39) up to C20 fiction (1.50) and Joyce (2.30). OLMo aligned is a notable outlier among aligned models, with higher information density than its own base model. DeepSeek-7b base (1.04) is unusually low for a base model, sitting near Shannon's threshold alongside Hemingway.Lydia Liu, The Freudian Robot, p. 37:

"...interesting questions that his experimental work raises for us is: how does a stochastic view of writing correlate to the received theories of language, literature, and modernism on the one hand and to psychoanalytical speculations about the unconscious on the other? This question is pertinent to our inquiry because many of the earlier modernist literary and psychoanalytical experiments on language, automatic writing, and thought-reading had anticipated Shannon’s Printed English and his “mind-reading machine” in numerous ways. For instance, Shannon cites James Joyce’s Finnegans Wake as one of the texts exemplifying the lower threshold of redundancy and higher entropy rate in his stochastic model of Printed English. What makes entropy and its possible linkage with Freud’s Todestrieb (death drive) such an interesting problem for the study of digital media is the ways in which certain ideas migrated into psychoanalysis first and then got into information theory. Furthermore, Freud’s work and psychoanalysis in general may suggest some interesting clues as to the shared theoretical impulses or implicit exchanges among information theory, cybernetics, and modernist literature. Spanning across these moments of broad intellectual confluences is the techne of the unconscious that continually articulates itself to digital writing, machine, and social engineering. We turn next to the invention of Printed English by Shannon and its implications for a theory of digital writing."
65421
Ryan Heuser @ryanheuser.com · 24/05/2026
I don't want to belabor this but I don't think solidarity is the point so much as knowledge. I don't want to learn what young men are thinking, however distasteful, months later from a NYT opinion piece or a Louis Theroux documentary, which inevitably fail to understand these phenomena from within.
120
Ryan Heuser @ryanheuser.com · 24/05/2026
From a professional POV bsky makes perfect sense: we want curated networks. But from a political POV it abandons the value of reckoning with the totality of political speech. Eg the Luigi Mangione moment: the cross-politics solidarity was a surprising discovery, only possible in a total environment.
100
Ryan Heuser @ryanheuser.com · 24/05/2026
Fair enough. But I miss the Habermasian fantasy of the internet as a public sphere, however perverted, however communicatively irrational. Giving up on that ironically seems to contribute further to the platformization of the internet—even more than a private platform functioning as public sphere.
120