.png) |
| Simonides hands Herodotus his famous Thermopiles Epitaph, made by AI |
This follows on a previous ironic (if not, um, iconic) contrast in AI posted here.
There is a sharp irony in the timing. In 2026, the United States marks 250 years since its Declaration of Independence from the British Crown. Yet the same year has brought a new wave of manifestos about artificial intelligence—documents that ask whether the systems now planning, coding, researching, and acting with limited human supervision might eventually demand a political language once reserved for peoples and nations: autonomy, consent, sovereignty, and independence.
The provocation is not that today’s AI agents are conscious citizens. They are not. It is that the two great AI superpowers—the United States and China—are building systems whose capabilities increasingly escape the neat categories of tool, property, infrastructure, and subject. And in doing so, they may unintentionally revive John Perry Barlow’s old cyber-libertarian dream: a digital realm that governments cannot fully command.
Update: let's juxtapose next another ironic contrast in AI from the NY Times...
This reprises AI conversations early in exploring AI around 18 mo. ago: two short ones here and here on Medium, and a long one here on my original blog's feature post on AI.
A short conversation about song lyrics, licensed lyrics sites, AI training, and the awkward asymmetry between what AI systems may learn from and what they may reproduce.
Update: next another thematic exploration of AI issues to follow a five-part series starting here.
A NY Times story about trekking the Arctic in the Hudson Bay area unknown to me - I was lucky to visit & wrote about the High Arctic here - reminded me of a previous tragedy in Alaska, both of men searching their souls & ending in the wilderness. As usual, what follows is AI-assisted except for the maps derived from data, more on main blog here.
Two GIS-ready waypoint tables follow each story. Coordinates are WGS 84 (EPSG:4326); use longitude as X and latitude as Y.
.png) |
| Made in Copilot AI following the original prompt above |
 |  |
Draw multiple portrait frames to string into a cartoon strip. Setting is a 19c. steam train in AZ Monument Valley at high noon w dusty atmosphere pulling a few box cars, one with an open door. 1st photo: me, my daughter& my ex & 2nd photo my new friends. Frames:
1) that train arriving in distance
2) side view of boxcar w me, my ex & daughter from 1st photo sitting looking out open box car
3) jump off to stand next to train and ex + daughter go off in box car
4) that train leaving in distance at L, new train arriving in distance at R
5) new train & boxcar arrives w cowgirl sitting by open door arm outstretched to me running next to train
6) have to jump aside to avoid water tower that has an EMPTY sign hanging akimbo
7) still run next to open door box car w cowgirl kneeling in open door beckoning me
8) jump onto open door of boxcar as cowgirl helps me in
9) we stand hand-in-hand looking out boxcar open door
10) train goes off to the right in the distance
Update 1: more on Queen Marie the story-teller in early XX c. European politics here
Update 2: next in ongoing AI-assisted experiments, a cartoon story board here
***
... My story-telling just won't stop! Following a journalistic experiment here, let's return to the classic format we left here: another AI-assisted far-ranging story from one photo...
 |
| Accidental sunbeam from a flash taking a photo of a print |
Political economy · Technology · Historical parallel — Stand‑alone essay
Fifty years after the political Right redefined efficiency as a virtue of the smaller state, Silicon Valley's AI evangelists increasingly promise a far more radical economy: dramatically lower costs, drastically fewer workers, and an eventual world beyond ordinary employment.
Pau, 24 September 2026 — by Andre Zolnai
AI‑assisted research and drafting
This New York Times opinion piece argues that the most dangerous AI risk is not accidental loss of control but our deliberate choice to allow AI systems to engage in recursive self-improvement (RSI), where they autonomously design and train ever more powerful successors. The author contends that because we already struggle to understand and control current frontier models, letting them improve themselves would rapidly push capabilities beyond human comprehension and oversight, so RSI should be banned or strictly prohibited until safety can be convincingly demonstrated. The column situates this proposal within a wider debate in which some AI leaders call for “pacing the frontier” or even a moratorium on advanced AI development, while political and commercial pressures—and rival states like China—push for rapid acceleration.
HTML: Perplexity AI
Source: what is doom loop
19 September, 2026
 |
| Sadhu reading in Kathmandu, Nepal (1965, Greg Zolnai) |
The first three posts in this series traced the AI doom loop, the rare‑book “buy → despine → scan → discard” pipeline, and the legal and copyright fault lines. This fourth post brings the story down to the level of historical and geospatial research: what happens to projects that rely on old maps, local histories, and specialized monographs when those sources are privatized, degraded, or made less accessible.
 |
| AI-generated by author (Medium) |
The first two posts in this series framed the “tiger eating its tail” dynamic of AI doom loops and mapped the “buy → despine → scan → discard” pipeline for rare and out‑of‑print books. This third post focuses on the legal and copyright fault lines: who owns these texts, on what basis AI companies use them, and where the law may yet draw a line.
 |
| medievalist.net |
 |
| Mouseion Alexandreias, "Shrine of the Muses of Alexandria" |
HTML: Perplexity AI
Source: What is a doom loop
18 September, 2026
This follows on a history note here if we remember "whence AI?" First of a series continuing here.
 |
| Original tale, YouTube |
There is an Indian folk image, often told as a kind of nursery rhyme, of a child and a tiger chasing each other around a tree, round and round, until both are worn down and “turn into butter.” It is a vivid picture of a futile, self‑consuming loop: motion without progress, effort that only exhausts and dissolves what it started with.
That image fits the emerging “AI slop cycle” uncannily well. As models train increasingly on text generated by earlier models, and as the public web fills with AI‑influenced content, we risk a kind of cultural churning: language and knowledge going in circles, losing texture and truth, until everything starts to feel smoothed into an indistinct paste.
 |
| At Teknica of TekXpert Systems, Calgary 1987 |
