Camilla Holm Soelseth
banner
supercamilla.bsky.social
Camilla Holm Soelseth
@supercamilla.bsky.social
Kulturingeniør, informasjonsflanør, jungeltelegrafist, medieøkolog og (digital) humaniora-entusiast. Gift med @sortulv.bsky.social

Researcher and Destroyer in the (digital) humanities. PhD in Library and Information Science.

Academic hedge witch.
Pinned
Endelig! Nå er Storygraph Reading Wrap Up 2025 klar!
Reposted by Camilla Holm Soelseth
Tenk å bruke begrepet "blandingsbarn"
February 14, 2026 at 8:17 AM
Reposted by Camilla Holm Soelseth
Nååååhh, kosetekst om Harry føkkings Potter og audible fra Siss Vik. Man skulle jo anta at folk som jobber med bøker var i stand til å finne noe som kan innta samme rolle, samt finne andre steder å skaffe seg lydbøker, men ok. Potter i 2026 er faen meg helt absurd.
February 14, 2026 at 7:54 AM
Reposted by Camilla Holm Soelseth
Don't have time to read/answer folks, but I need help.

Any documents mentioning or to/from:

Epstein, Terje Rød-Larsen (aka Terje) or Miroslav Lajčák (aka Miro)

for the following dates +- (next post)

I also need approx geo location on all of those people, same dates
Bonus for Bannon & Anil Ambani
Norwegian media: I have a bombshell and I would rather it was dropped by a newspaper. So if you’re a journalist in a Norwegian newspaper working with the Epstein files dm me.

If I don’t hear anything I’ll post a thread later today
February 13, 2026 at 1:11 PM
Faren min hver gang han er oppgitt over hvordan folk uttaler seg offentlig:

"Nei, nå man han se å smøre brødskiva med leverpostei sjøl, altså"

And I think that's beautiful.
February 14, 2026 at 12:18 PM
Blir radikalisert av TikTok
February 14, 2026 at 9:33 AM
Just watched Marc Kennedy and Oskar Eriksson in a heated rivalry.
February 14, 2026 at 8:01 AM
Reposted by Camilla Holm Soelseth
Hadde vel heller bedt dem lese Tess of the D'Urbervilles. Da får man virkelig skutt ned sine romantiske fantasier....
February 14, 2026 at 6:36 AM
Reposted by Camilla Holm Soelseth
På generelt grunnlag synes jeg litt synd på alle som skal lese boka etter å ha sett den nye filmen og oppdager at det er en bekmørk voldsorgie
På generelt grunnlag vil jeg bare si at jeg tar avstand fra alle som forsøker fremme Wuthering heights som valentines-roman.

Det er kun et red flag.
February 13, 2026 at 10:47 PM
Reposted by Camilla Holm Soelseth
1/ Finally wrote up “The Story of Mendeley”! Most people know the tool, few know about its rise and fall. The Mendeley story provides important clues for how to build self-sustaining AND non-extractive knowledge commons, which is why I think it deserves more attention 🧵
What happened to Science Goodreads and how do we rebuild it? A 65 million dollar question (at least) - CAIROS Blog
The story of the rise and fall of Mendeley
cairos.leaflet.pub
February 13, 2026 at 8:55 PM
On institutional betrayal

"While interpersonal traumatic events are often the most harmful, those that involve betrayal of a trusted or depended upon relationship or institution are uniquely harmful"
The Emotional Toll of Living With Institutional Betrayal - Dame Magazine
What happens to us when institutions protect power instead of people? From Epstein to ICE, institutional betrayal is inflicting trauma and moral injury.
www.damemagazine.com
February 13, 2026 at 9:48 PM
Reposted by Camilla Holm Soelseth
Hva er det egentlig innvandrere har gjort mot Ola Svenneby? Han er jo helt besatt jo
February 13, 2026 at 7:49 AM
På generelt grunnlag vil jeg bare si at jeg tar avstand fra alle som forsøker fremme Wuthering heights som valentines-roman.

Det er kun et red flag.
February 13, 2026 at 8:04 PM
Reposted by Camilla Holm Soelseth
how bad do you have to fuck up to allow an exploit in fucking NOTEPAD
The AI features added to Notepad created a critical exploit vulnerable to loading a compromised txt file.

The fix is part of Patch Tuesday, so update your systems.

AI is not a feature I ever desired in Notepad. I should look into some alternatives...
AI-upgraded Notepad now exposes Windows users to critical exploit
A critical remote code execution vulnerability found in the AI-upgraded Windows 11 Notepad has fueled renewed criticism of Microsoft's forced AI feature creep.
cybernews.com
February 11, 2026 at 9:37 PM
Reposted by Camilla Holm Soelseth
I cannot stress enough how hard it is to fuck up Notepad, an app that is literally older than me. And yet here we are.
“You can’t ignore LLMs are good at code” sure but also Microsoft broke notepad this week.
February 12, 2026 at 5:41 PM
Reposted by Camilla Holm Soelseth
Just like with Brazil arresting their insurrectionists rather than re-electing them, other countries are holding their Epstein-tied politicians accountable far better than we do. www.bbc.com/news/article...
February 12, 2026 at 5:44 PM
Reposted by Camilla Holm Soelseth
people are also surprised when they learn the internet doesn't literally float about in the cloud but is made of physical infrastructure and can be banjaxed by strategically fucking up some submarine cables
February 12, 2026 at 8:37 PM
Reposted by Camilla Holm Soelseth
right, the internet has infrastructure. which LLMs/AI are incredibly dependent upon for "survival."
people are also surprised when they learn the internet doesn't literally float about in the cloud but is made of physical infrastructure and can be banjaxed by strategically fucking up some submarine cables
February 12, 2026 at 8:39 PM
Reposted by Camilla Holm Soelseth
another core reason why we should fear becoming too dependent as a species on AI/LLMs is because there are still an enormous number of situations in life - disasters, sheer remoteness, infrastructure problems - in which you do not have access to electricity or the Internet.
right, the internet has infrastructure. which LLMs/AI are incredibly dependent upon for "survival."
people are also surprised when they learn the internet doesn't literally float about in the cloud but is made of physical infrastructure and can be banjaxed by strategically fucking up some submarine cables
February 12, 2026 at 8:41 PM
Reposted by Camilla Holm Soelseth
OpenAI’s internal ChatGPT tool can access company documents, Slack and email—then map what was reported to who had access to that information. https://thein.fo/4rO9WNx
How OpenAI Uses ChatGPT to Catch ‘Leakers’
Here at The Information, we’re always trying to break through the hype to understand how folks are actually using AI in the real world. And recently, I came across a new application of AI that I can’t...
thein.fo
February 12, 2026 at 9:15 PM
Reposted by Camilla Holm Soelseth
This is genuinely insane, I clicked to see the years expecting the low part to be 70s or maybe 90s but its 2015!!!!! The entire anti trans panic is a decade old
a stat i always cite is the number of anti-trans articles in the press went from 60 in a year, to 7,500
February 12, 2026 at 8:43 PM
Reposted by Camilla Holm Soelseth
A lot of people really don’t realize that there is a vast amount of information out there that is not online and plausibly never will be, or can’t be.
I think that's a point that will surprise a lot of people. There's a lot of calling the internet capable of accessing "the sum of all human knowledge" and such that I think gives people the impression that any and all records exist online somewhere.
February 11, 2026 at 8:47 PM
Reposted by Camilla Holm Soelseth
Related, suspect there's a goodly amount of stuff online that is not *discoverable* :
• Not indexed by search engines
• So thoroughly de-optimized by SEO that it's unlikely anyone will ever happen upon it
• Exists in a DB but the server rendering software is broken/missing
February 12, 2026 at 9:13 PM
Reposted by Camilla Holm Soelseth
Interesting. No access for AI models “unless item-level provenance and meaningful attribution can be demonstrated in practice, and the archival organization retains contractually enforceable control to stop further use.”
"Information without origin is a liability."

Introducing the UVA Archival AI Protocol, an AI training and access standard for archival organizations.

Read more: https://library.virginia.edu/news/2026/protecting-what-remains-introducing-uva-archival-ai-protocol

#AI #Archives #EthicalAI
February 12, 2026 at 9:18 PM
Reposted by Camilla Holm Soelseth
most people think that what historians do is read things and repackage them - what we do is more like a combination of being a detective, an archivist, a public speaker, and a philosopher, and all of those require the manipulation and interpretation of info rather than regurgitation
this also goes for history. I keep thinking about that study a while back that claimed
historians jobs were among the most vulnerable to AI. which is a thing one can only assume if you haven’t considered where history *comes* from - fundamentally, from people reporting back on stuff that happened.
February 12, 2026 at 9:18 PM
Reposted by Camilla Holm Soelseth
A historian colleague of mine once blew someone’s mind after he asked “can’t you just look that up online?” and she replied “who do you think finds that stuff and puts it online?”
February 12, 2026 at 6:08 PM