I dag har jeg skrevet om innestemmens paradoks, om hvordan media mangler evnen til å gjenkjenne autoritære og ekstreme strømninger når den er kledd i dress og opptrer «dannet». Om hvordan verdiene våre brukes mot oss.

It’s time to wake up — if you aren’t using ChatGPT to optimize your work (and your life!) you’re getting left behind. Watch this for 5 simple, lesser-known ChatGPT prompts that geniuses use! youtu.be/IS65dBNlng8

“NVIDIA and OpenAI have pushed each other for a decade [...]" s/pushed/inflated/

That's a big if you're starting with there, Nadella.

Another excellent summary by David Gerard.

Please - keep anthropomorphizing your glorified RNGs as if they have any sense of intent or agency.

humor er bare morsomt om det er mot marginaliserte og sårbare grupper, når skal den radikale venstresiden forstå det

!!VIKTIG!! Youtube sitt nye AI-system har automatisk begynt å markere noen videoer som "mature", og fjerner disse fra feeden din dersom det samme systemet har *automatisk* har skrudd på "Begrenset innhold-modus" på kontoen din! Dette betyr at mange av våre nyeste videoer er USYNLIGE! Her er fix: 1/X

😮‍💨

Zora's domain, er det Zora dot gov?

We all know fubar. Now we have cobar: Compromised beyond all repair. Alias: sloppified Superset of vibe coded and LLM-powered. To use e.g. for software where LLMs (or similar fundamentally unreliable mechanisms) one way or another have been allowed to inject their tendrils.

Woop woop.

What the bloody... To be this disconnected from the effects of what you're spearheading when in such a position should be grounds for negligence verdicts.

Fortsatt ikke verdt det.

We need to stop using a framework of 'ChatGPT killed someone', it feeds into the myth of the system's agency & OpenAI's framing. The right framework to use: OpenAI is continuing to perpetuate an indefinite industrial disaster. The oil pipe is not at fault when it breaks. The company that built it is

I did it the other way

I'm looking for laptop to eventually replace my 2017MBP. Something conveniently repairable and upgradable preferably. I've been holding out quite some years waiting for @frame.work to ship to Norway, but I don't see it happening in the near future. Any general (non-mac) recommendations out there?

“i have to use AI because i’m bad at drawingggg 😭😭😭”

Vi har fast journaliststilling ute! ✨ Hos oss kan du: 💘 Skrive om verdens mest spennende bransje. 💘 For Norges fineste lesere. 💘 I Oslos koseligste kontorkrok, omgitt av gamle datamaskiner, stålamper og Turtles-figurer. 💘 Og du trenger *ikke* å kunne kode! stilling.m24.no/fast-stillin...

You break git, you buy git?

Perhaps. But there’s more to the discussion than being scared/hurt and performative anti: there are also valid reasons to dislike certain AI-models which are so fundamentally flawed that they can first appear to look usable after they’ve vacuumed every piece if written text ever.

the amazing thing is that I don't think Altman has actually lied to his investors. We're gonna set your money on fucking fire. Consider it a donation. We might also build *AI GOD*. But that's a forward-looking statement, you understand.

#ICYMI 2 mil+ Filipinos perform #AI “crowdwork,” part of a workforce spread across the Global South. “The budget for all this, I know it’s big,” Paul said at a coffee shop in Cagayan de Oro. “None of that is trickling down to us.” www.washingtonpost.com/world/2023/0... via @washingtonpost.com

Fandango pay for everyone’s ticket if you talk too much

We ran a randomized controlled trial to see how much AI coding tools speed up experienced open-source developers. The results surprised us: Developers thought they were 20% faster with AI tools, but they were actually 19% slower when they had access to AI than when they didn't.

People really, really, really need to stop saying nonsense like “well, clearly they have a plan, otherwise they would never spend this much money on this”. Stop. People and companies spend unimaginable amounts of money on BETS about technology or political power ALL. THE. TIME. Spend ≠ knowledge.

Slop as a Service

Jeg har litt delte følelser. På den ene siden er det forfriskende ærlig at folk sier de «vibe-koder» og på den måten eier det at du ikke bryr deg om verken faget eller opphavsrett. På den andre siden så er det dumt at ikke flere bryr seg om faget og opphavsrett.

No they won’t. These models are not sentient in any way. They are trained auto complete engines. Anthropic routinely publish bullshit «research» like this to make their LLMs (and LLMs in general) appear more powerful and intriguing than they are. Press: please stop the fucking hype.

Can it think or feel or reason? No. Does it have anyone's best interests at heart? No. Can it be held responsible for choices? No. Does it understand the consequences of "choices" it "makes"? No. Should it be DECIDING? FUCKING HELL NO.

Who hurt those guys? And why didn’t it stick?

Jeg tar meg rett som det er i å tenke at jeg ønsker meg den foreløpig kun teoretiske salongbordboka som ymtet frempå av @jostein.radcrew.net / @radcrew.net .

Jeg vil anbefale alle å stange hodet gjentakende i veggen - det blir så godt når man slutter. Kanskje er dette en subtweet til andre hendelser i verden.

Jeg er uenig med dere, Brønnøysundregistrene.

Jeg kommer ikke til å lese din artikkel med AI-genererte bilder. Hvorfor? Jo, jeg antar da at din egen verdsettelse av teksten er så lav at du har vurdert slikt generert søppel til å være relativt sett kvalitets- eller verdiøkende. Hvorfor skulle jeg da bry meg? Tips: dropp heller bildene helt.

Wow! New amazing tool created with raylib: **Scrap** A tool to build software using a simple blocks-based interface. 🤯 Free and open source: github.com/Grisshink/sc... And the tool is just **1 MB**! Can you believe it??? Better software is possible. ❤️

Thanks to @ndcconferences.com for including my talk "Bootable applications - an introduction" for their NDC Oslo conference in May 2025. I look forward to have some fun with everybody joining me on @ Room 5 at 11:40 on Friday 23rd of May. ndcoslo.com/agenda/boota...

Nature may be healing!

It seems to me people are too focused on AGI in particular rather than "contextually useful". Even confuse the two. That's not helping. I also see quite a tendency of people simply to ignore negative aspects in favor of any positive aspects. I find that ignorant and irresponsible - at best.

Somehow I find myself loving this.

Growing up loving MI, MM and DOTT, this was the first post-that-era point an click game I loved which also showed me that it was possible to build a point and click game with puzzles founded in logic and still make it entertaining. An eye opener. Can recommend.

En annen ting er at vi som faktisk er med å lage disse programmene du visstnok «egentlig vil se», bygger mye av vår kulturelle kompetanse nettopp på å utforske kulturens ytterkanter, og skaperne av dagens «sære kunstfilm» er premissleverandør for morgendagens kassasuksess.

– Som første norske nettavis erklærer jeg herved full AI-stopp. Jeg *forbyr* AI-generert innhold på kode24. ✊ – Jeg vil at det skal være ett sted igjen på internett hvor du veit at det er mennesker som produserer innholdet du konsumerer.

«Ja men jo kanskje» er en serie på åtte episoder om gaming, popkultur og historiefortelling. Personlig og nydelig fortalt av hele spill-Norges @kjernekarl.bsky.social Alle episodene er samlet i denne hendige spillelisten! www.youtube.com/watch?v=iXm-...

Nydelig å se tech-sfæren blomstre her 🙌

I get it. Twitter has bigger userbase, more outreach, etc. But look, I just don't wanna fucking open it. Every time I open it I get blasted with ads, propaganda and rage bait no matter how many people I mute. What's the point of having all this "outreach" if it literally destroys your mental health?