Jason Farris

@TheAhmadOsman What do you mean by “learn how to use them”?

You don’t need Mythos, just learn how to use Qwen 3.5 27B and Gemma 4 31B and you’ll realize you have so much intelligence under your control on a single GPU

@NousResearch @PaulSolt I did. Openclaw would just crash all the time. Hermes isn’t perfect but it’s better

@PaulSolt Day 3: download Hermes

OpenClaw didn’t work for me on day 2. How many days of setup and exploration will I need to get a meaningful assistant?

@recap_david Nobody told him?

I turned a Zillow listing into a cinematic property renovation video for $15 (no camera, no drone, no film crew) Just listing photos + Calico AI. Most agents are paying $100-$500 per property for videos at this quality. Offer this to realtors in your area and print money. Here's how the workflow breaks down: → Grab listing photos from any property (Zillow, Redfin, take your pick) → Lock in one consistent renovation style across the whole house → Generate photorealistic after-renovation images → Animate the transformation in every room → Build a cinematic closing shot with the realtor's contact info → Drop in a custom music track → Stitch it all together The result: a scroll-stopping property video that gets buyers dialing — built entirely from photos that were already on the listing. No videographer. No staging budget. No waiting around for golden hour. Comment "RENOVATE" and I'll send over the full workflow + every prompt + a step-by-step walkthrough video (gotta be following so the DM goes through!).

@max_paperclips @Teknium I want one

heyyyy I have one of these virtual operating system desktops for Hermes Agent. when I'm happy with it will drop it as a plugin. Sort of OpenBox inspired but with Hermes-ish theming. It's become my main daily driver already even though I'm still working on polishing it up, whole setup is crazy useful I found a regular chat UI is just too limiting, and some Factorio-style thing is too gimmicky (I tried lol). You need a way to arrange multiple chats easily, check the status of stuff at a glance (which agents need an update or are blocked), check kanbans, watch premade workflows & scheduled tasks flow through everything, check infra, training runs or evals, VNC into sandboxes, whatever. Stay in the flow as much as possible, lock in when you want, expand back out to check the whole system state at a glance. Operating systems kind of have that UX all nailed down, may as well use it Totally bullish on the paradigm

@StarfieldGame Go Terrans!!!

The Terrans have arrived with one purpose: to unite humanity by force. The Terran Armada story DLC is available now on PC, Xbox, and PlayStation!

@TheAhmadOsman I got two sparks and two 5090 boxes. We will see

@wokerandomism @Tesla_AI Fake

@Tesla_AI Got the FSD 3 years ago. Still barely use it because you get a strike as soon as you put your eyes away from the road for a few seconds. I feel like it'smore convenient to just drive on your own. Refrain from buying this feature until the roll out unsupervised FSD.

New release of FSD Supervised now starting to roll out This update brings 20% faster reaction time to further increase safety, among many other improvements Full release notes below Full Self-Driving (Supervised) v14.3 includes - Upgraded the Reinforcement Learning (RL) stage of training the FSD neural network, resulting in improvements in a wide variety of driving scenarios. - Upgraded the neural network vision encoder, improving understanding in rare and low-visibility scenarios, strengthening 3D geometry understanding, and expanding traffic sign understanding. - Rewrote the AI compiler and runtime from the ground up with MLIR, resulting in 20% faster reaction time and improving model iteration speed. - Mitigated unnecessary lane biasing and minor tailgating behaviors. - Increased decisiveness of parking spot selection and maneuvering. - Improved parking location pin prediction, now shown on a map with a (P) icon. - Enhanced response to emergency vehicles, school buses, right-of-way violators, and other rare vehicles. - Improved handling of small animals by focusing RL training on harder examples and adding rewards for better proactive safety. - Improved traffic light handling at complex intersections with compound lights, curved roads, and yellow light stopping – driven by training on hard RL examples sourced from the Tesla fleet. - Improved handling for rare and unusual objects extending, hanging, or leaning into the vehicle path by sourcing infrequent events from the fleet. - Improved handling of temporary system degradations by maintaining control and automatically recovering without driver intervention, reducing unnecessary disengagements. Upcoming Improvements - Expand reasoning to all behaviors beyond destination handling. - Add pothole avoidance. - Improve driver monitoring system sensitivity with better eye gaze tracking, eye wear handling, and higher accuracy in variable lighting conditions.

@coreyganim Any post that uses movie and tv clips to promote some AI breakthrough is a scam. Also it’s the same ten clips ll the time. Bot harder bro!!

The moment you realize "Second Brain as a Service" is a real business: 1. Charge $1,500-3,000 to build a client's knowledge base (3 folders, 1 schema file, their existing data loaded) 2. Monthly retainer $300-500/mo for ongoing ingestion, health checks, and new source processing 3. Target agencies and consultants first. They have years of scattered data across Slack, Drive, email, and call transcripts. They'll pay tomorrow. 4. The setup takes a weekend to learn, a few hours to deliver. The client gets a searchable wiki that gets smarter every time they use it. 5. Stack it: competitive intel vault + client knowledge vault + content vault = $1,000-1,500/mo per client 10 clients = $60K+ year one. From a system built on folders and text files. Full breakdown of the system in the article.

@gophillionaire @Google @openclaw @NousResearch I’m running gemma4 with Hermes currently

Has anyone tried @google Gemma 4 with their @openclaw / @NousResearch Hermes? Still looking for a good Claude alternative as the main orchestrator.

@jasonkneen I bought two sparks and two stout rtx5090 boxes. Fingers crossed!

The DGX Spark GB19 has up 1 PFLOPS (1,000 TOPS) of FP4 Tensor performance The Mac Studio M4 Max has 38 TOPS. It's GPU is stronger in higher precision workflows but it's not even a sniff on the tensor throughput of the Spark. The Spark can handle prompt refill faster, large-barch inference and it's faster 5-10x in FP4/INT4 workloads. You can also pick up the ASUS GX10 which is a Spark in all but name less than the Nvidia branded version.

Good god please don't listen to this horseshit. 1. OpenClaw is a ralph loop with messaging it's not f**king the greatest piece of software ever written and there's alternatives like paperclip and nanoclaw and tinyclaw that work as good if not better. 2. You don't have to be burning money like this fool. He's making his on clicks and views and so all this drama is great for his numbers but having a 24/7 workforce that's cost him tens of thousands has lead to him releasting ZERO products over last months. Even his screenshots showed 0 tool use. It's all engagement farming bullshit. But you can choose to follow the fool or listen to people who actually know what they're talking about and aren't farming for likes and replies and horseshit. Up to you I have no skin in the game -- just sick of FUD.

@Teknium Explain

Did you know you can get your claude through hermes without any fancy tricks? Just type /claude-code In a new session. Hermes Agent can pilot a claude code session for you like it’s nothing. Give it a try It also doesn’t break all the self improvement loop and features of your hermes agent like openclaw’s approach does

@jumperz @Teknium I gave up trying to keep openclaw running and switched to Hermes honcho. I’m still struggling to get it to do anything useful with local models behind it but at least it tries and doesn’t just break all the time Progress i suppose

honestly i was a fan of openclaw, until i started installing it for some people around me.. i’m not here to shill nous research but i have to say hermes agent just respects your time more than openclaw right now and the experience is just way better to me, the install is always the products first impression and openclaw is failing that test hard currently last time i spent over 4 hours setting it up for a friend .. discord, webhooks, memory, eveything .. it used to take me way less like 1 hour at max, so much kept breaking in between.. i still run both at the same time and i love what openclaw does architecturally, but i can’t imagine going thought the setup from zero at all, it would be such a nightmare.. hope the openclaw team takes a breath from shipping features and puts that same energy into onboarding…the product underneath is genuinely great, it just needs to stop losing people before they ever get to see that.

@Focus_entmt Xwing

What game would you play instantly if it got remastered? 🪄🎮

@AlexFinn Your post would be more helpful if you were defining exactly what kind of of work IS succeeding on local models

If you used a Claude subscription with OpenClaw, read this: Unfortunately all other AI models out there absolutely suck with OpenClaw compared to Opus It's just a fact and anyone denying this is delusional So here is my new recommended OpenClaw setup: Pay for the Opus API and use it as your orchestrator Then use other models as the execution layer If you do this correctly, yes your costs will go up, but not by as much as you think I use my ChatGPT subscription as the coding execution. GPT 5.4 is excellent at coding. When The Opus orchestrator gives a coding task to the ChatGPT subagent, it always performs really well If you are on the Pro plan, you should have enough usage to have ChatGPT be the execution layer for every task. But if youre on the $20 a month plan, youre going to need other subscriptions to handle other tasks GLM 5.1 and Qwen are excellent. I'd get a cheap sub through them and have them handle all other tasks given to them from the orchestrator The best setup tho if you have the hardware is Opus API for orchestrator, ChatGPT for coding, then local Gemma 4 and local Qwen handling everything else. Right now have Gemma running on my DGX Spark and Qwen 3.5 on my Mac Studio. They handle all other execution from my Opus API orchestrator Unfortunately all options above will cost more than the $200 a month subscription. It just is what it is. But if you optimize correctly it wont cost much more, and you'll still get frontier performance. OpenClaw is the most powerful piece of software ever released. $200 a month ($2,400 a year) was a steal for a digital employee. Honestly anything under $50,000 a year is a no brainer if you run a serious business. The situation isn't great but you also need to face reality: Claude Opus 4.6 is the best model for OpenClaw. If you use any other model, your productivity will suffer Business is a battlefield and I refuse to fall behind, so despite me not being happy with the Anthropic decision the setup above is what I'm going with Virtue signaling might get me brownie points on the internet, but it won't increase my productivity

@chrysb @openclaw Openclaw never ran after so many attempts. Hermes no issue so far

hermes agent setup is way more tedious than setting up @openclaw!

Hello I’m not smart so please bear with me. I’m running Hermes Agent on DGX Spark and everything I ask it to make it acts like it’s making it but then when it’s time to go run it it’s always “no such file” “no such path”. Hermes thinks it’s someplace it’s not and i can never actually find its outputs. Everywhere it says the output is i can’t navigate to or find in search. Please help im new

@bloggersarvesh How weird is it that everyone uses the same five movie clips to promote AI announcements? Jeez bot harder guys

Copilot employees watching this announcement be like

@EthanHe_42 Is there a published guide how to leverage imagine by API? My company uses a lot of video

Excited to share our most advanced image generation model! It excels at aesthetics, instruction following and world knowledge.

@sudoingX What model to run on twin sparks w unified 256gb?

people keep asking me what model to run on a single 3090. it's not even close. Qwen 3.5 27B dense Q4_K_M. undisputed.