Mau “Rules without Rulers” Ledford

We are the easiest place to use the best frontier models & open source models all at once!👇

AI film director is gonna be a normal job.

@garrytan Why 397B moe over 27b dense? The latter is way more accurate for long back and forth prompts and agentic use cases

Thinking Machines is impressive. In a couple hours I just fine tuned my own Qwen3.5-397B model this afternoon. Fast usable multimodal is also going to enable very mind-blowing personal AI.

Are these “orbit data centers” in the room with us now?

Cool "local LLM saved me" story: I'm at a hotel in a remote country and the wifi works on my phone but the hotel's Captive Portal auth just won't load on my laptop. Just can't get it to trigger. There is literally no one that can help you. With no internet, I crack open OpenCode + Qwen3 27b 8-bit via LMStudio and it begins debugging. Says wifi is operational but DNS is blocked for some reason so it can't trigger Captive Portal. It proceeds to find the Captive Portal local IP address and adds to my etc hosts the override: 12.35.64.44 hotel.wifi.network And boom. I'm auth'd and online. When I get access to Codex I tell it to debug it further and turns out I have 1.1.1.1 global DNS override in my wifi settings which hotel wifi was blocking. Local LLM saves the day.

@Teknium @steipete @naviyd

@steipete @naviyd This is not true. The two times i attacked you on a personal level were after you made personal attacks against me, my team, or our community. Happy to truce and be friends from here if thats over with, always have been. I especially am fond of your lead mod, shadow!

Hey @steipete and @Teknium - can you become friends and work together? Open source collab is a super power.

Codex is a capable, all-you-can-eat buffet, while Claude Code is being more restricted every day, with Anthropic secretly nerfing models/settings and now charging you for agent use at 10x API rates. Their MO seems to be to offer you less, while charging you more as time passes. Claude is a better product ran by a worse CEO and logistics planner but better tech geek. Access to GPU’s is what Anthropic failed at and continues to handle it in the most shifty way possible. The XAI deal was crazy esp with the smack Elon talks about “Misanthropic” but hopefully this leads Claude on a better, more stable path. Competition is important.

Thanks for the shoutout @AethirEco ! It's a pleasure to try to make a difference in a field you love

To install simply point hermes (or any other agent) to the github and say "install this skill". github.com/krunkosaurus/h…

A few weeks ago I built Hermes Therapist: a therapy-style reflection skill for Hermes Agent. This skill is designed to make Hermes a structured, emotionally attuned, non-sycophantic reflective companion: part mirror, part translator, part witness, part coach, part emotional debugger. When installed all you have to do is run "/therapy-mode" to start a session. Sessions can last just a few minutes or 30 minutes+ if you like. Notes and feedback are kept private and exist only on your machine. I have field tested it for a few weeks and have gained some really valuable insights about my own philosophy and struggles. Maybe you will too? Link below.

@maryisclear @Sogni_Protocol Can read it here in detail: docs.sogni.ai/sogni-chat/

@krunkosaurus @Sogni_Protocol Can I see the prompt that made this please I

Psst. Seedance 2.0 is coming to Sogni 🤫 Made using chat.sogni.ai

Today we’re spotlighting @Sogni_Protocol, one of the most interesting creative AI teams building in the Aethir ecosystem. Sogni is not just another AI image app. It is a full-stack creative production network: apps, SDKs, agent tools, and decentralized GPU infrastructure working together.

Kenneth was a dev, then a decade long stellar VC, and thanks to AI he’s back into hardcore dev mode with an army of agents. I call him my Garry Tan.

Important: This is a summary of an amazing video by one of the best creators I know. Video in the first comment. If you've been struggling to setup a productive local environment I'll summarise, but you should watch the video. 1. Qwen3.6-27B with NO THINKING - 4bit - 16bit depending on your resources 2. Hermes agent: It's polished, minimal, and OSS, if you're on OC keep at it it's also great but IMO consumes more tokens == slower 3. Learn to work in a detached way: Instead of small, unclear prompts make something really specific and hand-off to a local agent: - What is your end goal? - What is the output format? - How would you recommend a task be done? - How would you deal with common issues? Kick off a job and go do something IRL, the lower speeds paired with very clear prompts means you can check in every 2 hours on average for your next free task. Treat it like a challenge, it'll teach you so much. ~~~~~~~ If you're more technical, wire your devices with Tailscale & use protected Cloudflare Tunnels to serve your inference API to your network so you can work from ANYWHERE with your local models. I love Droid, Pi, and Opencode but you can use local models in Claude Code, Codex, Cursor relatively easily. Most of your day will involve computer-use which the models are great at. Need to reset a server? Do it for free, want to gather research? Do it for free. Not every task requires a 10T param beast crunching on things, being able to quickly cycle between models is what's critical here, which is why I recommend the three harnesses I did ~~~~~~~ What is productive? - Organise all your files, folders, images, etc.. have each one tagged (Qwen is omni so no more date time titles) - Create a content cache for yourself: - recommendation algo for books & videos etc. - content/papers/courses for studying - Social media post ideas (write your own posts tho) - Torrent manager, you don't need Netflix or Spotify lol - Simple utility coding projects: landing pages, designs, games, scripts - Market watchers: purchase stuff, check marketplaces, do shopping - Budgeting, taxes, and subscription management I could go on and on, obviously it's not going to make you a millionaire or whatever but it makes life more FUN. ~~~~~~~ AI is a tool, one that's very good at accelerating you towards self fulfilment as long as you: 1. Understand what it is 2. Learn how to talk to it 3. Keep trying to improve you ~~~~~~~ My videos tend to be stream of thought, I am starved of time and just want to make sure I keep up. I would not be doing this if I couldn't just generate 3 thumbnails on t3chat, for example. That's why I spend so much time working on this. I see myself being more of who I want to be every day and a lot of that is with the help of this technology, allowing me to more quickly and effectively interact with the world. Being able to own this at home? True sovereignty. ~~~~~~~ Thank you, Digital Spaceport. I wouldn't have gotten this deep into running my own infra without your videos.

@soft4211 @Sogni_Protocol See announcement channel!

@krunkosaurus @Sogni_Protocol Any way for me to get Beta Access??

Sogni Intelligence is here across all our apps and workflows.

Sogni Intelligence is the AI layer making Sogni apps, SDKs, and APIs smarter, easier to use, and more personalized. A few weeks ago we introduced the first version. It has already received a massive update. Along with an expanded suite of agent-ready Sogni Tools, we have incorporated the powerful new Qwen3.6-35B-A3B model, with up to 40% better performance across benchmarks and a 4x expanded 264k context window. That is enough memory to hold the entire Goblet of Fire novel without running out. Learn more about Sogni Intelligence below. 🧵 👇🏻

How to start winning : Install Hermes on your MacBook. Get a $100 a month codex subscription and have it run on gpt 5.5 Don’t waste time on anything else. It has absolute frontier intelligence and they have made soo much improvements to the personality. It feels like working with a friend. ( kind of scary tbh ) The amount of tokens you get subsidized with running gpt 5.5 through the codex cli is unbelievable. Give hermes GitHub access, ssh keys to your servers, and cloudflare tokens. It will take your local project and put them live on the web with domain, dns, ssl, nginx, pm2, everything setup in just ONE prompt. It can monitor your backend / database and make sure everything is running perfectly. Give it a support email, when customers reach out instruct it to look into there account on the server and makes any non destructive fixes. Hermes will automatically turn everything it does into skills, getting better and learning from mistakes constantly. You can literally have a 250k a year highly technical employee that WORKS 24/7 for $100 dollars right now. I need you to wake up.

We used to think a stronger collective came from getting bigger. More people, more noise, more visibility. But after four years of building ARC from scratch, we’ve learned the opposite. The best momentum comes from smaller, sharper conversations. The kind where the right people can challenge each other honestly, share what they’re seeing, and help solve common founder pain points together. That’s what our ARC Roundtables are for. We bring the right builders into one room, each with different experiences, networks, and market perspectives, so the conversation is useful, honest, and leads to action. That’s how something stops feeling aspirational and starts becoming real. It’s also how institutions stop looking aspirational and start becoming inevitable. Quick shoutout to all the builders who joined us for the first session:: - @icunucmi, CEO, ARC - @edison0xyz, Co-Founder, Based - @JontyKay, Co-Founder, Undercurrent Capital - @krunkosaurus, CEO, Sogni AI - @md_argv, Co-Founder, Zabava Labs - Ryan Chew, Co-Founder, Tribe - @sharonlourdes, Head, Gnosis Business & Gnosis Pay (Ramp) - @hypetris_, CEO, Elfa AI - @nickooj (moderator), ARC - @jaclynleezy, ARC