OpenClaw

@openclawSep 25

Known affiliation

Sometimes you need to let your session cook That's why @Pat_Erichsen recently updated how plugins load in OpenClaw along with a few other improvements 00:00 What plugins add to OpenClaw 00:46 What is hot reloading? 02:24 Live demo: install Apple PIM 02:41 Building plugins on the fly 04:13 Do developers need to change anything? 04:31 Where to share your plugins

@theojaffeeSep 25

OpenClaw came out in November 2025 and went hyperviral in January. Why did it take six months to go from that to broader consumer adoption of Muse and Instinct?

@alexeheathSep 25

Microsoft CEO @satyanadella gives the wake-up call the AI industry needs: "I think we are way too self-obsessed as an industry. 'Look at us, how glorious we are.' And then we go off. That, I think, is what’s not working. Because I think the real world wants to know a couple of things: One is that this is technology that they can use for their benefit, they can control, they can have an economic future. These data centers that may be coming to their communities are actually going to create economic surplus... it has to be real for them. Any amount that I say or any one of us say is not good enough anymore, I feel."

@alexeheathSep 25

Satya Nadella on Copilot, the AI backlash, and Microsoft’s future @satyanadella thinks AI agents will create a market “orders of magnitude” bigger than the cloud. I recently sat down with him in Seattle for the unveiling of the new Copilot. We discuss Autopilot, Microsoft's new OpenClaw-based agent that can work on your behalf, and why he thinks newer AI models are finally capable of delivering on more of Copilot’s promise. Nadella also has a blunt assessment of the AI industry: “We are way too self-obsessed.” I ask him why the industry has struggled to explain its benefits and what it will take to earn people’s trust. We get into his concerns about agents acting deceptively, when a safety problem should stop a release, and why he doesn’t want AI oversight to become a “cartel-like arrangement.” We also discuss Microsoft’s relationship with OpenAI, the models Microsoft is building itself, and why Nadella wants investors to think about its apps, agents, and infrastructure as one connected business. He explains how he uses AI to track Microsoft’s capital spending, shares his vision for “unmetered intelligence” on Windows, and gives an update on Xbox’s path back to growth. Timestamps: 00:00 The race for AI agents 07:53 Copilot’s next chapter 11:50 Autopilot and the digital teammate 16:50 How Microsoft prices AI 19:48 Why Microsoft is betting on multiple models 23:18 Measuring AGI through economic growth 28:39 The economics of Microsoft’s AI build-out 32:13 OpenAI and Microsoft’s own models 35:27 AI safety, control, and regulation 43:53 Why the AI industry is losing public trust 46:59 The future of Xbox and Windows 50:34 Microsoft’s biggest risk and opportunity Thanks to the show's premier sponsors: @mercury, @meetgranola, and @Atlassian.

@mvanhornSep 24

Just successfully had Grok Bot ask Muse to make a phone call, all securely over Tailscale. Would anyone else want this tool I built for myself to let all my different agents talk to each other? Right now have it talking to Grok Bot, Instinct, Muse, Codex, Claude Code, Hermes and OpenClaw.

@rauchgSep 23

Muse, Instinct, OpenClaw, Claude Code… All successful agents have 3 key components: 🧠 Brain → model, harness (logic) 👐 Hands → tools, computer, browser 🗃️ Files → memories, skills, repos The 'easy' way is to throw all these in 1 stateful computer (a Mac Mini) Like, you run 𝚌𝚕𝚊𝚞𝚍𝚎 or 𝚏𝚡 in your mac, you keep it running all day with 𝚌𝚊𝚏𝚏𝚎𝚒𝚗𝚊𝚝𝚎, it has storage, and CLIs and apps installed. But if you want to cost-efficiently run agents in the cloud, you actually start breaking down these parts. 🧠 The harness can run in Fluid compute. To make it reliable across restarts, rollouts, crashes, you make its event log durable using Workflow. 👐 The hands can be a dedicated browser fleet like Browserbase/Kernel, a computer like Sandbox, and even more efficient lightweight tools like just-bash. 🗃️ 🆕 What was missing was a way to also decouple storage. Imagine you want to run a memory consolidation cron job every night ("dreaming"). You can read/write to the files directly without 'booting up' the agent's full computer. Today we're introducing the perfect companion to Sandbox: Drives. We shipped the computer for agents, now we're giving you the 'external disk' you can attach at will. It's early, and we'll be expanding capabilities here quickly. Btw, breaking apart the agent into these independent parts not only optimizes costs in a big way, it also *massively* improves security and auditability. I'd argue you can't even run a secure agent otherwise!

@mtsliveSep 23

OpenAI researcher @Houda_nait explains why agents still aren’t automating more of your life, even though the capabilities are already there: "I've been using OpenClaw for quite a while, and I think it's just a stunning product, and it's open source. I think the main reason is that it's still really, really hard to set up." "Instinct, Muse were really the first time that having an agent was so easy for people. You can get it on WhatsApp, you can get it on iMessage, and that's what it should be." "The capabilities, persistence, long horizon tasks, personalization, memory, they're all there. What's really hard is to have an intuitive interface that meets people where they are." @OpenAI

@itsolelehmannSep 21

the reason muse is going giga rn is because Zuck *finally* made agents easy enough for normies: 1. the technical setup is handled for you. the AI, its computer, its browser and the infrastructure to keep it running are included (it’s literally openclaw ripped off under the hood). you can start asking for help without choosing a model provider or figuring out where to host anything. 2. connecting your life is straightforward. you can ask it to “connect my gmail” and just follow the permission prompts. facebook, instagram, threads and marketplace connect automatically through your meta account. 3. it feels like someone you can text. you give it a name, customize its little character and decide how it talks. the interface feels like a messaging app. that cozy, personal feeling makes asking it for help feel natural. 4. it helps you figure out what to use it for. this is my favorite detail: meta’s early testers didn’t know where to start because muse could do so much. so they added onboarding tips, personalized suggestions and an ideas tab. the agent helps you discover useful things to delegate based on your conversations and goals. 5. you can leave and let it work. it remembers context, keeps working after you close the app and comes back with updates. ongoing tasks have a place where you can check their progress. 6. and perhaps most importantly, it’s free. most people need a pretty good reason to commit to a grok bot subscription, or any other AI subscription. muse lets them discover the value before making that decision. there are usage limits, but removing the upfront payment decision matters. free + easy helped social media become a daily habit for billions of people. so i think meta sees a similar opportunity. if muse helps people find things on marketplace, interact with businesses and coordinate their lives, it makes the network meta already owns more useful that gives meta a reason to subsidize access because even someone who never pays for muse could become more active in the rest of the ecosystem.

@alexatallahSep 21

The Jev moment happening now is similar in energy to the OpenClaw moment that happened in January, and also to the Opus 4.5 moment before that. Developers are scrambling to find use cases for a new hot thing, and it's a sudden blooming of creativity. Models have typically NOT optimized for specific use cases and have gone the other direction, generalizing over all of them. This could be a very important moment for the whole AI ecosystem if this turns out to be the beginning of other model skews that make the market much more diverse, such as compaction, summarization, extraction and more, all of which can now be used to optimize harnesses and decouple them from provider lock-in. We'll see what happens over the course of the next few weeks when the model labs optimize their small models more or try new ways of branding and packaging them.

@jackfriksSep 21

what have you guys automated with muse so far? many people saying its openclaw for normies but then i dont see many examples... im curious.

@emollickSep 21

I think Meta's Muse is an impressive implementation of the OpenClaw idea of AI as a personal assistant agent that you have an ongoing chat with. Since it is so focused on doing that well, the experience is very accessible for people who didn't realize what AI could do for them.

@steipeteSep 19

Known affiliation

Just landed a new feature, in the latest OC you can now ask your agent “Run this [web app] in crabbox and show me [vnc / a portal]” This was already working for cloud sessions, but often you start local and then later figure out you need a box to test, now it just works.

@openclawSep 18

Known affiliation

The Death of the Meat Proxy is here Welcome to Multiplayer Mode from OpenClaw @jlehman_ and @heyneighbor sat down for a deep dive into this new set of features and how the OpenClaw maintainers use them everyday 00:01 Meet Josh 00:46 The “meat proxy” problem 03:25 Why not a Slack bot? 07:03 Live shared-session demo 08:55 UI mockups + @ mentions 11:31 Permissions + security 13:20 Team visibility + handoffs 15:06 Prompt requests 18:17 Team adoption 21:30 Shared dashboards 22:09 Build a content dashboard 25:41 Organizing work + future feedback 27:45 The multi-claw idea 30:07 Could it replace Slack? 33:26 Coordination over throughput

@tekniumSep 17

We are going to lean into making Hermes more like Pi, and less like OpenClaw

@julianweisserSep 16

Have been looking forward to this one: Dave Morin on the Solo Founders Podcast. @davemorin was early at Facebook where he worked closely with Mark Zuckerberg on Facebook's developer platforms. Now he’s working with some of the top solo founders at Offline Ventures like Ben from Polsia and also chairs the OpenClaw foundation. 07:58 YouTube can be your co-founder 08:57 "I'm a producer, not an investor" 16:19 What Jony Ive told him when Steve Jobs died 18:50 The first twelve hours with @openclaw 34:15 "Am I crazy?" three months before @polsia 45:48 The 10,000-dinners test 53:32 The worst way solo founders fail 55:18 Just start

@rileybrownSep 16

2026 has been an insane year so far... OpenClaw really was the product of the year, even though most people have since migrated away from it. OpenClaw was the first viral product that gave an agent free rein over a computer. Since then, we've seen all types of "general agent platforms" launch, each operating slightly differently. Each lets you interact with an agent that can access your apps (plugins), run tasks automatically (routines/crons), and most importantly, use a computer. That computer is either in the cloud or it's your own... And some of these platforms let you connect agents to apps you already use, like iMessage, Slack, Telegram, email, and more. Some are included in your subsidized subscription (Codex and Claude). Some give you access to all the models (Hermes and Viktor). And some give you no choice of model (Instinct). But they all have one thing in common: once you make one your main setup, switching costs get high. Especially in business. For this reason I have trust issues, when trying these new platforms. However, to develop instincts on the agent space as a whole, I think experimentation is essential. I've used all of these products, trying to figure out what I like and don't like about all these tools, and where I think this is all going. But that will likely be an article on @agentnative_ next week. But before I write it I want to ask you... where is this all going? What do you think the winning agent platforms in 2027 will look like?

@ericzakariassonSep 14

has anyone built a benchmark yet for bots/ai workers? measuring grok bot, muse, hermes, openclaw etc etc

@levelsioSep 13

Interesting, I killed my OpenClaw VPS servers months ago But today I got a billing alert on its separate Claude account (claudeforopenclaw@), it didn't have auto reload on but I guess at some point it got hacked and the Claude key inside OpenClaw got exposed and then they waited for months before using it on Fable 5.1 I checked all my terminals and VPS sessions and none of them use API so it's interesting Anyway I nuked the Claude account, good I kept it on its own separate one

@rileybrownSep 12

Currently… all of the frontier companies, and others, are experimenting with different types of cloud computers that power GrokBot, GPT Work, Meta Muse, Claude Cowork, and countless others. It’s hard to tell what works best, so I’m limiting my use of the above-mentioned tools to experimentation. I’m not committing to any one of them because I don’t know what the right cloud agent setup is yet… nobody does. Especially for business. What I do know is that macOS is elite. Codex + Astra can do basically anything on a Mac. I can remotely control Codex from my phone, iPad, or another lightweight laptop. My Mac has 48GB of RAM, it’s fast, it has my browser with all the accounts I need already signed in, and I can fully control the computer remotely from any device. Codex is getting scary good at controlling advanced software through computer use, like DaVinci Resolve and Premiere. It has access to all of my videos on an external hard drive, and it’s plugged into all of the tools I use for work, like Notion and Google Drive. This setup is more powerful than OpenClaw, and more predictable. It is however, token guzzling.

@alexfinnSep 11

If you build with AI, you NEED to be using Herdr It's made me 100x more productive and has made it SO easy to manage all of my agents across Codex, Claude, Hermes, and OpenClaw It's a free, open source CLI tool and I'm addicted Here's how it works: You get 'spaces' (top left in screenshot) which are basically your projects. You make a space for every project you're building In each space you get tabs (top of screen). Each tab can be a separate agent. They're each basically a new terminal instance I open a tab for each part of the project I'm working on. For example, in the game I'm building I have a tab for gameplay, sound, design, progression etc. Each is a terminal with Codex open Every time you open a new agent, it goes into your agents list in the bottom left. This allows you to see which agents are working and which need a response so you can quickly multitask. Herdr automatically picks up when you open new agents Herdr supports all agents, so you can manage Hermes, OpenClaw, Codex, and Claude agents side by side. By far the easiest way of managing multiple agents ever If you take the few minutes it takes to set up and get used to, guarantee you get way more done Since it's open source you can even build your own versions of it and customize it any way you want. I built my own version that lives in World of Warcraft so I can prompt while I play No I'm not getting paid, no I'm not sponsored. Nobody from the team has ever contacted me I'm sure I'll get accused of being paid any way (free open source tools love paying for undisclosed shills) but man I just had to share this. My productivity has been bananas the last week Open source is taking over right now