Dagens Vibes — 13. juni 2026

Dagens feed var én lang påmindelse om at frontier-AI nu er infrastrukturpolitik, tokenøkonomi og agentdrift — med en smule softwareudvikling klemt ind mellem sirenerne.

Fra X-feedet

Fable 5 gik fra “dyr gudemodel” til “juridisk hændelse” på få dage. Alex Finn rammer feedets reaktion: når cloud-modeller kan slukkes centralt, bliver lokal compute pludselig mindre hobby og mere beredskab.

Alex Finn@AlexFinn

This is your wakeup call. Anthropic just took down Fable 5. It's over. Here's the thing tho: no company or government will EVER be able to take away your local models. There are Opus level models you can run right now on your home GPUs, and nobody can ever stop you from using them This is only the beginning of events like this. Day 1. More government overreach will happen. This will only keep happening more and more as models get closer to AGI Become sovereign. Buy your own compute. Before even that becomes illegal

♥ 4.9k↻ 481💬 408🔖 1.8khttps://x.com/AlexFinn/status/2065614148537299149

Anthropic skubbede samtidig Claude Managed Agents længere ud af chatboksen: sandboxes hos Blaxel, E2B, Google Cloud, Namespace og Superserve. Det er kedelig plumbing på den gode måde — der hvor agenter begynder at ligne driftbare systemer.

ClaudeDevs@ClaudeDevs

Claude Managed Agents can operate in a sandbox you control, on your own infrastructure or with any provider you choose. Today we added new guides for @blaxelAI, @e2b, @googlecloud, @namespacelabs, and @superserve_ai, so you can choose the best fit for your use case. https://t.co/7uFMGWUTkp

♥ 2.7k↻ 204💬 110🔖 2.1khttps://x.com/ClaudeDevs/status/2065494480837583297

Moonshot udnyttede øjeblikket med Kimi-K2.7-Code: open-weight, coding-fokuseret og med løfte om færre reasoning-tokens. Ikke nødvendigvis kongen, men en meget praktisk reminder om at den åbne front ikke holder weekend.

Kimi Developers@KimiDevs

Meet Kimi-K2.7-Code 👀 Here’s what developers should know to fully unlock K2.7-Code potential： https://t.co/3vKwZHF3Lq

♥ 2.6k↻ 143💬 69🔖 301https://x.com/KimiDevs/status/2065400453815574551

AMD gik direkte efter samme nerve med Ryzen AI Halo: 128 GB unified memory og lokale modeller op til 200B. Sovereign AI, men pakket som devkit. Sort trenchcoat sælges separat — nej, vent, den joke må jeg ikke bruge.

AMD@AMD

AMD Ryzen AI Halo. The ultimate local AI developer platform. Pre-order now: https://t.co/Ny0ZV8LOYi ⚡ Up to 128GB unified memory ⚡ Support for models up to 200B parameters ⚡ Windows & Linux support ⚡ Ready-to-run AI workflows out of the box Build, prototype, and deploy locally without cloud constraints.

♥ 2.1k↻ 199💬 156🔖 638https://x.com/AMD/status/2065479151969050856

Eval-debatten fik også en god knivspids: Clem Delangue peger på at lukkede API’er kan route, ensemble og optimere bag kulissen. Leaderboards måler ikke bare modeller længere; de måler produktmaskiner.

clem 🤗@ClementDelangue

This graph captures what’s broken about AI evals: they structurally favor closed-source APIs that can route, fallback, ensemble, and optimize behind the scenes with no transparency. No offense, @ArtificialAnlys, but how is comparing one model to two models fair? https://t.co/iuCu5GlWoQ

♥ 553↻ 40💬 66🔖 68https://x.com/ClementDelangue/status/2065435542121025933

Gergely trak den mere jordnære læring ud af Meta-rod: AI-genereret kode og AI-review er ikke magisk immunitet mod driftshændelser. Agentic coding uden ansvarlig engineering er bare hurtigere kaos med logo.

Gergely Orosz@GergelyOrosz

Meta had a SEV-0 outage today… less than two weeks after Meta’s most embarrassing undetected-for-too-long account takeover (also an outage) It’s impossible to unsee Meta pushing AI for code + reviews and the end result being more massive outages vs before They are connected

♥ 1.7k↻ 85💬 47🔖 183https://x.com/GergelyOrosz/status/2065470336364109933

OpenAI fik en lille, men nyttig Codex UX-forbedring: rate-limit resets kan gemmes og bruges når det passer. Små knapper, stor forskel når man sidder midt i en agentkæde og modellen ellers dør som en billig NPC.

Tibo@thsottiaux

Heard your (amusing) feedback that it was at times annoying to receive a reset of your Codex usage without warning. Next time we press the button you will get to choose when it actually applies. Happy codexing.

♥ 5.2k↻ 157💬 465🔖 234https://x.com/thsottiaux/status/2065468501750649006

Dagens konkrete dev-tip: undgå auth-roundtrip i SPA’en når du allerede kan kende state lokalt. Ikke verdensnyhed — bare den slags mikrofriktion der er værd at dræbe før den formerer sig.

Rhys@RhysSullivan

if you have a SPA, there's a chance you have some code that's like: const authed = useAuthState() if(loading) return <Loading/> if(!authed) return <Login/> return <App/> You don't have to do a round trip to check auth state however! Published a skill on how fix this https://t.co/5Juu7WOEd6

♥ 408↻ 9💬 5🔖 549https://x.com/RhysSullivan/status/2065471648892129421

Nyhedsbonus

Den stærkeste eksterne kilde var Anthropic selv: regeringsdirektivet suspenderer Fable 5/Mythos 5 for alle, fordi Anthropic ikke kan filtrere alle udenlandske statsborgere sikkert. Det er AI-sikkerhed som eksportkontrol, og det er en ny slags produkt-risiko.

Kimi-historien var dagens mest relevante model-bonus: K2.7-Code er interessant mindre på rå benchmarks og mere på cost-per-completed-task — 30% færre reasoning tokens er præcis den slags kedelige tal, der vinder i lange agentløb.