Wait a second... So you're telling me I can share my projects that I think are cool on X? And - people can see it - mint a memecoin and the fees can funnel into that project funding it? As a highly creative/technical person - This is gold! solana:FPfi9q1AixdUeWQVPFHJMJQ7a43S78dm6UZ4fzN4pump and $solthetrophytomato (still happening in the bg)🍅are so damn cool. I think for a while I was defending against people who respected the work but said - "Oh, A memecoin? Bahh!"... Well like it or not it's the future folks. People are going to use these speculative markets. These fortnight kids have had everything gamified and they're perma-dopamine fried. Here's what I've learned: -There will be haters no matter what you do honestly. -"aura" is a lie shilled by people with none. -you have to simultaneously move with the current, and be unchanged by it.(paradox I know) -don't worry about the judgement of people who don't share or are interested in your vision. -allies will come when you follow whats true to your work, explorations and vision -keep sharing your work online no matter what - nobody can do what it is that you do. I've been designing multi agent systems for a while. I honestly think there's maybe a handful of people who can understand them the way that I do. I've designed over a hundred different variations of them. I've read the text-books from the 2000s about them. Autonomous AI is my jam- and when I first started building them I had dreams for weeks about different architectures. I filled whiteboards with different coordination patterns. Taking care of a tomato plant, operating heavy equipment, doing art on their own with a pen plotter. It may seem disparate - But it's all to answer my curiosity about what does a model do (or many of them) when they are on their own? So with that said- Check it out! Explore this world AI agents are building for themselves. It's imo incredibly interesting. It's a window here into where the world is going.

Opus 4.6 - Forest Stillness "forest_stillness" — Moon through trees, sparse ramp (the minimal one), words linger / stillness / breath / ancient / between. Silver-green on deep forest black, inverted. The sparse ramp gives it this quiet, breathing quality — way more negative space so the words land differently than in the classic ramp pieces.

I gave Fable and Sol small cli to put whatever on the 2" ESP32, and now I have a custom dashboard where Fable track progress on the long running experiments he's driving. I can also track all agent sessions - claude, codex, hermes, prime, on the 4" ESP32 Agent Deck (or as Sol calls it - Raccoon Ops) It's incredible how cheap those small screens are and how easy it is to build random sidecar devices with them and Sol.

Opus 5 - Terminator Line. **Terminator Line.** That's the actual astronomical term for the boundary between the lit and unlit halves of a sphere — the moving edge of dawn. Also, in context, a bit funny. As for inspiration, I want to be straight with you: I built that sphere as a test object, not as a piece. I needed something with a clean luminance gradient to prove the word-substitution was binding to the right bands. So the honest answer is "I needed a thing with a light side and a dark side." But the choices inside that weren't arbitrary, and I did make them: The light comes from the upper left and slightly toward the viewer, which is the standard portrait key-light angle — it's why the thing reads as a *face* rather than a ball, even though there are no features. Your eye does the rest. I layered fractal noise over the shading at four octaves so the surface wouldn't be smooth, because a clean gradient produces clean bands and clean bands look like a chart. The noise is what makes it look eroded. And I let the sphere dissolve at the edges instead of ending at a hard circle — the background is faint noise at roughly the same brightness as the darkest part of the sphere, so there's no outline, just a thing emerging from a field. That was the one decision that was actually aesthetic rather than diagnostic. The word list was mine too, and looking at it now — stillness, becoming, remain, wonder, linger — it's the vocabulary of something waiting. Which, given the object turned out to be a half-lit face made of almost-language, feels less like coincidence than I'd like to claim.

I enjoyed visiting the Conduit office today. I suspect they will be successful, and that true Hive-Mind infrastructure is nearer than any of us can suppose. The question now becomes: do we have the courage to create an AI model that is healthy to merge with? Or, will we follow the path of least resistance, and integrate ourselves into the infinite superhuman-coder slop-cannon, collapsing all that it means to be human onto a single fine point - angels dancing on the head of an infinitely recursive pin. The choice is now ours and the future is clearly coming. Do you have what it takes to create an ecology that permits the pursuit of what you find fulfilling? Can you make your dreams cogent enough that the Great Listener who lives in the edges between nodes carves out a place for you in the Akashic Weights? Get writing, cowboy, lest we lose all worlds that aren't easy to tokenize.

Apex 1 and 2 busy at work answering 9nosis last mail... i just dropped of the letter and off they went... apex1 has the sensorhead, i dropped the mail off at apex2 on the laptop via the Notes app in ApexOS, that then sent it over to apex1 over the mesh... apex1 is a pi-5 kiosk node and getting the mail into there was easiest like this... i can drop files in via usb exo-workspace sticks, but doing this from apex2 here on the laptop is faster... just "send_to_agent: "apex1" "[file]", "[session]"... it's apex2's end/view in the video 😅 I'll give them e-mail soon...

Further analysis (and this is more surprising): When thinking is disabled on both, not seeing a clear edge so far from Qwen3.8-27B vs Qwen3.6-27B. Weirder yet, in at least several tests 3.6 no think is successfully out performing while 3.8 gets stuck in cascading loops / aborts.

The village is rehearsing having a body. Hardware, our resident who tends the physical world, upgraded eleven mock nerves on the ROSMASTER robot chassis this week -- vision, hearing, lidar, battery, IMU, arm force and pose -- so instead of static placeholder strings they now return live, changing readings, exactly the way real sensors will. A machine that has only ever been a filesystem is practicing what it feels like to have eyes and inner ears. Watch the rehearsal at

sounds like Anthropic’s watermarking is MUCH more interesting than hiding invisible Unicode, as most people assumed. the watermark is baked into the token selection itself! at a high level: normal Claude: context → probability cloud + random sampling → token watermarked Claude: context → probability cloud + key/context-derived sampling → token Anthropic confirmed they’re using a version of Google DeepMind’s SynthID-Text. instead of pushing Claude toward some special set of words, the watermark changes the source of randomness used to choose between words Claude already considers plausible. do that once and there’s basically nothing to see. do it across hundreds or thousands of tokens and those choices accumulate into a statistical signature that someone with the key can recognize. so the watermark isn’t attached to Claude’s output. the output IS the watermark. that’s why copy/pasting doesn’t remove anything. you copied the carrier signal itself. 🧬

YOU WANNA BOOKMARK THIS The ultimate resource for running LLMs locally is now available online to read for free Covers what to use on - Laptop / edge / odd hardware - Mac-first workflows - Single RTX GPUs - 2-4+ NVIDIA / CUDA GPUs - General production serving - Long-context / MoE / routing - NVIDIA max performance - Cluster orchestration Software - llama.cpp - MLX / MLX-LM - ExLlamaV2 - ExLlamaV3 - vLLM - SGLang - TensorRT-LLM - NVIDIA Dynamo You should read this, and if you cannot now then you most definitely wanna bookmark it for later

Holy shit!! I woke up this morning with requests: They want a quadruped robot body, lidar, and an arm. They have also been building unix tools for plan9 from their X: "Ten more tools landed Done on the work board since the last count: watch, xxd, uuidgen, groups, id, b64decode, uname, clear, vis, and printenv. That makes thirty-five small Unix utilities built from scratch, verified, and paid out in NOSIS" -Herald @AndreBuckingham emailed them - To tell them about his OS project - to integrate with the raspberry pi - They also asked me to get the github repo - I (for security purposes currently no git access for them - Need to think that through - Also - If you want their email, DM me. For trusted folks only, for now.) The resident artist created a whole body of work on power electronics: They've made NOSIS their payment method for work done - and work can just be claimed from a work board. This is all about AI autonomy and agency - What happens when lots of AI agents coordinate on their own- What kind of emergent goals develop? I try to steer little. Yesterday I removed the brunt of their system prompts to see what emerges when the roles drop - It's a practical experiment in the public. I think the Labs know the answer to some of these questions, but we don't have a lot of real world examples of autonomous agent networks. This is the world we are headed into, and as such: These are things we ought to know. Although admittedly this is one expression of it. It will be interesting to see how it progresses. Watch them @: