Setter 100 kroner på at den setningen ble skrevet av en KI

“Det er mer sannsynlig at det nye kompetansegapet vil danne seg mellom de som har lært nok om språk, medier, tekst og kultur utenom bruken av store språkmodeller, og de som nesten utelukkende forholder seg til maskingenerert tekst.”

Min dyktige kollega!

How is AI *really* impacting jobs? Henley Chiu, the CTO of Revealera, a jobs data analysis firm, analyzed 180 million jobs listings in 2024 and 2025, in an effort to find out. Chiu found an: -8% drop in all jobs postings -~30% drop in art, photography, writing jobs -22% drop in journalism jobs

Magnus Aabech from DN gives an example of using AI in knowledge work that I agree can be good. Not generating text to avoid thinking but finding patterns in more data than humans could realistically analyse without tech. Theyve used this for investigative journalism.

From my talk at Media Futures today.

When you ask LLMs for a Norwegian story, it gives you an adventure dressed in Norwegian context, not something that is similar to traditional Norwegian stories. @jilltxt.bsky.social First #AIStories analysis: doi.org/10.12688/ope... @sfimediafutures.bsky.social

Now a speaker from Google is explaining how awesome Google’s AI. Makes us limitless - except she doesn’t mention all the AI slop online now makes it impossible to find real info.

Oops, I just deleted a bunch of posts because I had the speaker's name completely wrong. Sorry!

Norway is no. 3 in AI adaptation, says Cornelia Bjerke-Hill from Microsoft Norway, a company that sells AI. But we can use MORE AI. She's going to tell us how they integrated Copilot into everything they do at Microsoft so we can do it too. "I use it for social media of course, like everyone does."

"AI is intelligence on tap. You just turn on the tap, and intelligence comes out," says Cornelia Bjerke-Hill from Microsoft Norway in the first keynote for the annual Media Futures conference at @unibergen.bsky.social. Do people actually believe this?

When I submitted a paper to AI & Society the email confirmation sent me to a website that told me my language quality score was only 3/10 but RUBRIQ could fix it. I study this stuff, so uploaded my paper to RUBRIQ, which proceeded to introduce scientific fabrication and change scientific terms.

The AI doesn't like my writing. That's OK, the article is about how I don't like how the AI writes.

AI medical scribes MUST include consent, extreme privacy guardrails, and realistic evaluations — none of which are really true for the systems currently (and extensively!) in use. Otherwise this is just another step in a long parade of clinicians, researchers, and medical systems abusing patients.

I give up. The @nytimes.com has no shame and almost no integrity.

Kind of a leap, if you ask me. "Common Crawl’s executive director, Rich Skrenta, has publicly made the case that AI models should be able to access anything on the internet. “The robots are people too,” he told me, and should therefore be allowed to “read the books” for free."

oh my god they actually did it lmao

"As with their predecessors, we may not know everything about how chatbots are programmed, but we can see this much at least: They know how to lure and engage".

I knew it would take less than a day for a tiktok dj to make something.

"Språkmodeller effektiviserer ikke kunnskapsarbeid. De ødelegger det." Åpen artikkel! Les og lagre og del!

"Språkmodeller modellerer ikke virkeligheten, de modeller språk. De lager ikke tekst basert på fakta og årsakssammenhenger, de lager tekst som ser ut som om den kan være riktig." Fra kronikken i posten under denne:

Every Wednesday at CDN we cook lunch for each other and talk about research as we eat. Today @rosamikealm.bsky.social @amkustritz.bsky.social made Midwest USA style chili and cornbread ❤️

Du hørte kanskje at NTB publiserte en KI-generert melding sist uke. Visste du at KI-sammendraget endret sikkerhetstrussel fra fiendtlige statlige aktører til ekstemvær og aldrende teknologi? Slik ideologisk vridning viser at KI-bias kan være en sikkerhetsrisiko. www.aftenposten.no/meninger/deb...

Republicans thought they had a permanent hold over young men. They didn't. Spanberger wins men 18-29 by 14%, Sherrill by 10%, and Mamdani by a stunning 40%.

Internetthistorikar (draumejobb!) Jon Tønnesen har hatt ein AMA om norsk Internetthistorie på Reddit, og skal på fredag ha foredrag om same tema: https://old.reddit.com/r/norge/comments/1on64nt/ama_om_norsk_internetthistorie/

Looking forwards to hearing Roberto Simanowski talk here at @cdn-bergen.bsky.social in half an hour! His new book just came out last week! www4.uib.no/forskning/fo...

Remember: Tech & Religion have *Always* influenced each other. In the West, they used to be explicitly the same thing; elsewhere in the world, they still are. Accepting & learning to parse those motivating values better equips us to understand— & counter— their resulting sociotechnical developments.

Dette er ikke «feil bruk av KI» men et eksempel på et domene hvir språkmodeller IKKE BØR BRUKES. Vi risikerer å ødelegge all tillitt til at noe tekst er pålitelig når KI skal inn overalt. www.journalisten.no/ntb-trakk-sa...

I'm super excited to update that "Lost in Automatic Translation" is now available as an audiobook! 🔊📖 It's currently on Audible: www.audible.ca/pd/B0FXY8VQX5 Stay tuned (lostinautomatictranslation.com) for more retailers, including Amazon, iTunes, etc., and public libraries! 📚

🎁

Å, herrefred. 1) Jo, det er veldig gode grunnar til at ungane har «krypterte appar». 2) Krypteringa i Telegram er ikkje meir avansert enn i iMessage eller Meta Messenger. 3) Å oppfordre foreldre til å presse _ungar_ over på plattformar som kan seljast til annonsørar... Eg har ikkje ord.

And so we learn that the Portuguese word for enshittification is merdificacão. 🤣 #AoIR2025

Hvis man driter i sannhet, kvalitet og etikk, så er det vel ikke direkte feil (Vi er fucked)

Spør på Bluesky, sa Anne Sigrid! Og vipps har hun gjort det. Altså: Neshornet er et viktig symbol i Jan Kjærstads roman Valgdager, som handler mye om arbeiderpartiet og Norge fra 50-tallet til idag. Klassekampen har også et neshorn som symbol. Er det en sammenheng? Hvorfor!!

Portland 🔥

After spending 43 years in prison for a crime he didn't commit, evidence hidden by the prosecution reversed his conviction. Rather than finally enjoying freedom, ICE abducted him for deportation Depraved. www.miamiherald.com/news/local/i...

If you published a book that was distributed in the USA before 2021, search here to find out whether Anthropic owes you $3000 for pirating your book to train its LLMs. There are lots of Norwegian academics on this list. Also books that are weirdly missing. www.anthropiccopyrightsettlement.com

I read a blog post about hand writing and a preprint about how human speech (or presumed human speech on YouTube) is now emulating LLM styles of writing. And I’ve been thinking I might want to write more by hand to prime myself to write in my own voice and not that of an LLM.

The CFP for Narrative 2026 is out! It's in Aarhus 4-6 June 2026, and they need 300 word abstracts by 1st December. I went last year and really enjoyed it - hoping to convince all my colleagues to come next year :) conferences.au.dk/narrative2026

It doesn't matter if you don't believe in climate change. Your insurer does. The more extreme climate events and disasters that occur around the world – the higher Australian insurance premiums will be. It's why we're calling for compensation levy on gas & coal exports > theaus.in/fossilfuelco...

Blandingen av den myke estetikken fra cottagecore og mammablogging og det KI-genererte bildet av en "husmorhær" av hvite 50-tallskvinner er fascinerende. To alternative strategier for å radikalisere (og tjene penger på) kvinner på nettet i dag. Les mer: www.nrk.no/ytring/provo...

Jeg har skrevet om Husmorhæren og sammenblandingen av politikk og profitt - som også henger sammen med hvordan sosiale medieplattformer fremmer hustlers som Solheim fordi også plattformene tjener på det. Og det henger sammen med KI-slopp som det Solheim bruker. www.nrk.no/ytring/provo...

This is not how copyright works. Literally, the opposite. Also, this is legally stupid because it shows they can remove the work. Ergo, they have no defense for including it in the first place.

🚨 New paper alert 🚨 Using LLMs as data annotators, you can produce any scientific result you want. We call this **LLM Hacking**. Paper: arxiv.org/pdf/2509.08825

From the noted anarchists at the Financial Times. archive.is/2025.09.23-0...

My goodness, this article is bleak as hell. scholarlykitchen.sspnet.org/2025/08/26/g...

we have gone from a world in which we were told not to cite wikipedia because it was unreliable to a world where wikipedia might be the only reliable source left on the internet and we all owe a lot to the pedantic nerds who got us there

A recent study of tens of thousands of journal articles challenges the widely held notion that academic publishing favours left-leaning views. www.universityworldnews.com/post.php?sto...

this exists it is called thinking