llmfit. Useful tool that probes hardware and tells you exactly which LLMs will actually run. - handles MoE expert offloading, picks the best quantization for your RAM, estimates tokens/sec before you even pull the weights. Essential for local dev. github.com/AlexsJones/llmfit

Lmao

n8n is dead. I fed my AI all my workflow JSONs and it replaced every single one. I ran n8n for over a year. 15 workflows on a Hetzner VPS. HTTP nodes, webhooks, JSON parsing, conditional branches, error handling. Then I did something stupid: I exported every workflow as JSON, dropped them into @OpenClaw, and said "replicate all of this." It took 20 minutes. Not 20 minutes per workflow. 20 minutes TOTAL. The AI read every JSON, understood the logic, and rebuilt each one as a self-healing cron job. Here's the difference nobody talks about: n8n workflow: → 47 nodes connected with spaghetti → API changes? Rebuild 6 nodes manually → Edge case? Add more conditional branches → Breaks at 3am? Wake up and debug OpenClaw workflow: → "Scan X for trending AI news, write a 1000-word article, generate a thumbnail, publish to WordPress, submit to Google Search Console" → API changes? The AI adapts automatically → Edge case? It figures it out → Breaks at 3am? It reads the error and tries a different approach. You sleep. The craziest part: I now have 20 autonomous jobs that do MORE than my n8n ever did. Because here's what n8n can't do: • Write a full SEO article from scratch • Generate images for thumbnails • Decide NOT to publish if the content is trash • Review its own mistakes every 4 hours • Learn from yesterday's failures • Generate full YouTube videos — script, voiceover, stock footage, animated whiteboard, editing — completely autonomously That last one blew my mind. I give it a YouTube URL, and it comes back with a fully edited video summary: AI-generated narration, Excalidraw animated scenes, relevant stock footage, transitions. No editing software. No templates. The AI figured out how to use Remotion, FFmpeg, and TTS on its own. Try doing THAT with n8n nodes. My AI scrapes Instagram, transcribes reels, hunts expired domains, deploys websites, writes posts on X — all on cron. If one task fails, it self-corrects and keeps going. n8n doesn't think. It executes. OpenClaw doesn't just execute. It understands. You can literally hand it your n8n JSON export and say "do this but better." And it will. Because it doesn't need nodes and connectors. It needs intent. My n8n VPS is still running. I haven't logged in for 3 weeks. Stack: Mac Mini, @OpenClaw, plain English prompts. That's it.

Want to try OpenClaw or other AI agents on Omarchy but don't want to risk your setup? Run Omarchy inside Omarchy. One script, one command — full QEMU VM with auto-install, SPICE GUI, and SSH access. Let agents go wild in a sandbox you can destroy and rebuild in minutes. gist.github.com/sevos/5ffaa7… cc @dhh

Podpowiem Wam fajny patent na to, jak pozbyć się dewelopera ze wspólnoty. Kiedy kupowałem mieszkanie, pozostali właściciele mieli łącznie ponad 50% udziałów. Podjęliśmy dwie kluczowe uchwały: 1. Zmiana zarządu na właścicielski (niestety sam wylądowałem w jego składzie). 2. Ustalenie wysokiej stawki na fundusz remontowy. Jedynym głosem sprzeciwu w naszej wspólnocie było 45% udziałów dewelopera. Argumentował, że przecież budynek jest na gwarancji i nic nie ma prawa się psuć 🤡 Obecnie obniżyliśmy już sobie opłaty, ale deweloper po tamtej podwyżce zyskał ogromną motywację, żeby szybko sprzedać pozostałe lokale. O dziwo, inni mieszkańcy nie protestowali przeciwko chwilowej podwyżce, bo rozumieli, że jeśli mieszkania będą stały puste, koszty ogrzewania części wspólnych będą bardzo wysokie. Tak czy inaczej, jestem u tego dewelopera na „czarnej liście”. Z funduszu remontowego zlecaliśmy profesjonalne audyty nieruchomości, które owocowały długą listą nieprawidłowości. Dzięki temu skutecznie egzekwowaliśmy naprawy z gwarancji i rękojmi: od poprawy kostki brukowej, przez nowe nasadzenia, aż po montaż rynien (o których „zapomniano”) i wiele innych poprawek z najdroższych chyba było kucie klatek, bo zapomniało się podłączyć system ppoż xD Praca w zarządzie ma swoje plusy. Choć dostaję symboliczne wynagrodzenie, to w ramach czynu społecznego zaoszczędziłem wspólnocie pewnie z milion złotych, wymuszając na deweloperze naprawy. Przy okazji bardzo dobrze nauczyłem się pisać pisma przedsądowe. A i polecam jako pierwszą decyzję zarządu rozwiązać umowę z zarządcą którego wyznaczył deweloper. :-)

It used to be controversial to state that there exist 10x engineers. People would get all worked up. I have news for you. AI has created 100x, possibly 1000x engineers. Time to update those hiring plans.

Hey @AnthropicAI, a big fan of your achievements and Claude Opus here. Regarding the recent update… what about making Claude Code faster and better than @opencode instead?

Just today: I let AI in #opencode, for the first time, drive the browser and make a purchase for me in an online shop, compose and send out emails on my behalf, read and process my emails. I need to process this. Good night.

Unlike Campfire, where we shared a zero-history copy of the codebase, Fizzy's public codebase has the entire life of the product development cycle out in the open. 1,792 pull requests of historical fun and insight into how we make things at 37signals! github.com/basecamp/fizzy/pu…

opus is so good that upgrading to the $200/mo plan is starting to sound reasonable to me. i’ve never built this much in so little time

the smartest engineer you know will be a plumber in 5 years

I think we're headed toward a world where traditional app UIs become largely unnecessary. Instead of building custom interfaces that users need to learn (and relearn with each UX update), apps will just expose their backends through protocols like MCP. AI agents will handle the interaction layer. Users will tell the agent what they want, and it'll work directly with the backend. In this model, only backends survive. Frontend development as we know it gets displaced by agent integration.

bunch of people are asking me for my reaction about anthropic saying "software is done in 6 months" or whatever here's my response

Software engineers are waking up to a new reality today

European regulators: STOP DELAYING FSD. Every single week you block it, another ~400 people die on EU roads. FSD is already 5–10× safer than human drivers in real-world data. This isn’t “caution.” This is blood on your paperwork. Approve Tesla Full Self-Driving in Europe NOW. Lives are not negotiable. @RDWnl @UNECE @vonderleyen @teslaeurope @elonmusk #FSDinEuropeNow #StopDelayingFSD #TeslaFSD #SaveLives 🚨🚨🚨

🚨🚨 BREAKING: TESLA QUIETLY REMOVED “BETA” FROM AUTOSTEER IN FSD V14 BUILDS! 🚗💥 This is MASSIVE for Europe! ✅ Basic Autopilot/Autosteer is now officially “done” – no longer experimental ✅ Tesla is saying: “Highway Level 2 is mature & rock-solid safe” ✅ This CLEARS THE PATH for real FSD Supervised (city highway) approval in the EU! ✅ New 3-tier smarter nags prove driver monitoring is bulletproof – exactly what UNECE/RDW demands! Europe: the ball is in YOUR court! Give us FSD Supervised NOW – we’ve waited YEARS! ⏳🔥 @elonmusk @Tesla @Tesla_Europe @RDWnl @UNECE @EU_Commission #Tesla #FSD #FSDSupervised #Autopilot #AutosteerNoBeta #FSDEurope #BringFSDtoEurope #SelfDrivingCars #TeslaEurope #TeslaNederland #TeslaBelgium #TeslaDeutschland #Level3 #AutonomousDriving #EV 🔥🔥🔥 RT IF YOU WANT FSD SUPERVISED IN EUROPE RIGHT NOW!!! 🔥🔥🔥

I am tired of Gemini 3 hype for weeks, while the model is still not available to me. I really appreciate silent crunching by @AnthropicAI and dropping something when it works - not sponsored.

Never do this. HTTP 200 already means there is no error. Don’t reinvent the bicycle. Don’t overcomplicate.

your CLAUDEmd file might tell me more than your cv

I’m probably > 1000x. No joke. Projects that in the past would have taken a team of 5 $300k senior / staff eng / researchers 6 months are now weekend solo projects. It’s so over.