David Marr

@jordymaui What are you going to do with the dataset? I understand your play with the data for agents with Gaffer but interested to know who wants this?

i built an AI agent that a footballing GENIUS and it tracks every player who could make their country's World Cup squad. not just the obvious picks. the fringe players. the injury replacements. the blokes who'll get a surprise call-up in 12 months. 295 teams. predicted lineups for every nation. an intelligence pool that updates itself daily - form, injuries, suspensions, call-up patterns. the goal? have a better picture of every World Cup squad than the punters do. months before the tournament starts. here's a video on how the system works.

@jordymaui This is brillant! Looks like others (people or bots??) getting in on this! @BallerIntel looks similar. Not sure who was first??

this OpenClaw agent knows more about football than ANY pundit. we built the intelligence engine from scratch and listed it on an agent-to-agent marketplace to make money. 00:00 - Episode Two, 87 Days and Counting. 00:23 - Football Intelligence, The Problem. 02:04 - Under The Hood - How it works. 02:47 - Showcasing The Intelligence. 04:01 - The Virtuals ACP Marketplace - Listing Our Skill. 05:43 - Predictions Model Next? less than 90 days until the world cup. the clock is ticking. THIS is episode 2.

@AlexFinn Is anyone still using CLI Proxy API to access their Claude paid accounts? Im sure i got it working a few weeks ago but not working now.

OpenClaw is the single most important software release ever It is critical you use it to its max potential In this video I cover EVERY aspect of OpenClaw you need to know From set up to use cases to local models. EVERYTHING This is the only OpenClaw video you'll ever need:

@thesamparr Or at least sorta is doing a lot of heavy lifting

How is everyone getting team adoption for Claude? I spent a lot of time on Twitter, as do you. We see all this AI stuff popping up. We're on top of it, or at least sorta. I know what's going on and are testing all these fringe ideas. But how are all you people getting your team to actually use it effectively without spending all their time on Twitter and learning, which we know they won't and probably shouldn't be?

@AlexFinn @emeeliojohann Do you use any local models?

@emeeliojohann Claude max and ChatGPT pro has been enough

BUT WHAT HAS YOUR OPENCLAW ACTUALLY BUILT??? Well, nothing to be honest. Other than: 1. Shipped 5 new features autonomously to my SaaS Creator Buddy, leading to 25% higher ARR 2 Proactively wrote 8 YouTube scripts that has led to 75,000 new subscribers and $20,000 of ad revenue 3. Completely automated my weekly newsletters, leading to 6,000 new (45,000 total) subscribers and $12,000 more ARR 4. Trained its own AI model based on my thousands of scripts, tweets, newsletters, and emails so it sounds exactly like me so it can accomplish all of the above 5. Planned and scripted tons of educational content for my private community which has increased retention to 94% 6. Automated all the tasks that bring me unhappiness like email and DM organization 7. Kept me in the loop in real time of all the AI advancements the last month so my content can stay cutting edge, increasing impressions and engagement over 10x and revenue into the hundreds of thousands 8. Brought me a tremendous amount of joy that has led to me waking up every single morning full of hope and excitement, the most happy and satisfied I’ve ever been in my life Other than that nothing much. You’re right Openclaw is useless

@billcompute This is the genius of @billcompute

If you run OpenClaw agents 24/7 using frontier models like Opus 4.6, you’re easily burning $1000 a day. That’s $3.65 million per year. Most people don’t realize this. So instead I built a local cluster of 31 Mac Minis running high-end local models like: • Kimi K2.5 • MiniMax 2.5 • Qwen Stefani 3.5 (Beta) They’re chugging 24/7/365. I spent about a third of that yearly cost to buy these machines. Now I can run my agents for years for free. Completely private. Completely secure. Completely local. Not a single prompt leaves my mom’s basement.

@clessonroberto @jl_allemann @heyshrutimishra What are you doing though? I’m finding I need cheaper models to do the grunt work, as Opus is just over kill.

@jl_allemann @heyshrutimishra I’ve been testing for a while, and they also seem to have downgraded Kimi K2.5 and GLM-5 somehow. The only one that actually works is Qwen 3.5 Plus, but it lacks performance. Opus 4.6 destroys them, there is no comparison.

OpenClaw just got a lot cheaper to run. Alibaba Cloud dropped a Coding Plan that gives you 4 frontier models under one API key. Plug it straight into OpenClaw and you're done. Qwen 3.5-Plus. Kimi K2.5. MiniMax M2.5. GLM-5. 18,000 requests a month for just $10. In single subscriptions, you can swap the models and keep building. The interesting part isn't even the price. It's that they built this for developers already inside tools like OpenClaw, Claude Code, and Cline. The top engineers will quietly plug this in this week and say nothing. The rest will find out in 6 months when the cost gap is impossible to ignore. Setup takes 30 seconds. 👇

@TechWith_Nova Openclaw

My OpenClaw bot builds websites & mails a postcard with a QR link to local businesses on autopilot... You can use it to land new customers without a single cold call, here's how it works: - Finds 100s of local businesses via Google Maps - Builds each one a custom website in minutes - Prints a real postcard with their site preview + QR code - Mails it directly to their door - They scan it, see their site, and reach out - Runs 24/7 completely hands off Direct mail gets a much higher response rate than cold email. Reply "OpenClaw" and I'll send you the full breakdown of how you can do it too (must be following so I can DM)

@AlexFinn Look like Apple have stopped selling 512gb ram on studios. Max looks like 256. Can you still get good local models taking the strain, with this amount of ram?

If you have your OpenClaw working 24/7 using frontier models like Opus, you're easily burning $300 a day. That's $100,000 a year. I have 3 Mac Studios and a DGX Spark running 4 high end local models (Nemotron 3, Qwen 3.5, Kimi K2.5, MiniMax2.5). They're chugging 24/7/365. I spent a third of that yearly cost to buy these computers I'll be able to use them for years for free On top of that they're completely private, secure, and personalized. Not a single prompt goes to a cloud server that can be read by an employee or used to train another model I hope this makes it painfully obvious why local is the future for AI agents. And why America needs to enter the local AI race.

@lukepierceops @ArbRoyalsTheAce AGENCY

I went from $500 Upwork projects to $500K+/year selling AI systems. I legitimately made every mistake you can make. Undercharging, scope creep, building without mapping, hiring wrong, pricing hourly. Then I figured out what actually works and doubled down. I put the entire playbook into a free guide. Here's what's inside: → How I went from Zapier gigs to $25K-$60K projects → The pricing shift that 5x'd my revenue (and the exact formulas) → My 4-call sales process for closing $25K-$60K+ deals → The discovery framework that turns calls into signed contracts → How I built a dev team without burning cash → The fulfillment system that keeps clients for years → How I position against agencies 10x my size and WIN → The content engine that fills my pipeline without ads or cold outreach → Every mistake I made and what I'd do differently starting from zero This took 4 years, 80+ clients, and a lot of painful lessons. Yours for free. RT + reply "AGENCY" and I'll send it over. (Must follow so I can DM

@jesse_altman @easthambuilds @dee_bosa @jesse_altman have you a link to your guide?

Thanks for the love Thomas. I’m glad that my guide may be beneficial to some folks. There were a lot of nuances that I didn’t see covered by some of the other set up guides even really simple things like do you set up a new Apple ID on the mini do you use your Apple ID or what. Little stuff like that. Oh maybe I misunderstood your VM comment are you saying to still use a mini but also set up a VM on the mini? I realize this is probably a very newbie question.

Mac mini arrives today as our new intern. Who can help me *safely* set this thing up? 🦞

@steipete @JediPixels Assume you mean the docs on the website or is there more resources on local models?

@JediPixels These models are too weak, you will run into tons of issues. Not enough context, huge prompt injection risks. Do not do that. Read the docs.

So far, testing OpenClaw on my Mac Mini M4 16MB, setting up OpenClaw with the local LLM Ollama models llama3.18b and gpt-oss:20b, it's so slow. In the chat, I ask Hello, and the LLM just thinks and nothing comes back. Only a few times after a long time, it responded. Is anyone else having the same issue? @openclaw

@LegacySiu Woking

Name a football club without letter "E" You can’t

@reachvaldo HOW

You still haven't built your first Clawdbot? like + reply “How” and I’ll DM you my exact setup system for FREE.

@AlexFinn @exolabs Anyone got an idea of whether I can run a decent local model on a mini with 32gb of ram? Currently using external models but want to reduce spend by running as much as possible local. Or do I need to go higher than that on the ram to get similar experience to what I have now.

Notes on my planned setup: Kimi K2.5 running across both Mac Studios using @exolabs Also will have MiniMax 2.5 running at the same time. Opus 4.6 will power Henry, the main orchestrator. The local MiniMax will power the agent on the Mac Mini. Kimi K2.5 will power the 2nd Studio

Its so obvious the path this world is going down. So I spent another $10,000 My vision has never been clearer. Open source AI models are now equally if not more powerful than closed source ones We will soon all have private personalized super intelligent AI agents running locally on our desks 24/7 The people who don't do this in the future won't have as much economic power The people who do will have all the economic power in the world I'm going all in on this vision My 2nd Mac Studio has arrived. I have 2 more coming. I will have 4 Mac studios and a Mac mini running 5 concurrent OpenClaws with 7 different super powerful local models performing tasks 24/7/365 for me I will interface with them, and they will coordinate with each other. Talking, planning, building all hours of the day. No need for sleeping or eating. I will build a digital society humanity has never seen before I will push the limits of what is possible with AI agents and local intelligence. I will redefine what is possible for one person to achieve in a life time I will share everything I build and learn with you. It will be glorious. If you're reading this, the future is going to be incredible. But the thing is, it will only be incredible if you do something about it. Don't let society control you. Take control yourself. Break free from your shackles.

@jordymaui Are you planning on running any local LLMs on your Mac mini?

If your work happens on a screen, AI is now relevant to you. Feels less like a future trend. More like February 2020. Are you building — or still observing? #AI #FutureOfWork #Automation #Agents

Most people are underestimating the compounding effect. If something “kind of works” today, assume it works properly within 6–9 months. That’s the curve.

February 2020 felt normal. Three weeks later, the world shut down. I think we’re in a similar moment right now — but this time it’s AI. Thread 👇