Last week, we got OmegaClaw installed and running. Today, we’re taking it a step further; from setup to actually getting our agents to work. Workshop #2: OmegaClaw Live Activation is a hands-on session on extending your OmegaClaw Orchestration Agent, the part that brings your agent’s thinking, tools, and actions together. You’ll see how to: 🧠 Orchestrate: give your agent the reasoning and memory it needs to think ⚙️ Coordinate: connect the tools and skills it needs to get things done 🤖 Act: turn all of that into real, live results 📅 Today, 15:00 UTC Join us🔗 us02web.zoom.us/j/8968258733… See you there

It’s OmegaClaw Workshop day. 🛠️ We’re going live today. 🛠️ If you’ve been seeing OmegaClaw around but haven’t actually set it up yet, this is the session to join. We’ll go through the setup live from the ground up: → Install OmegaClaw → Get your first agent running → Walk through how its memory works → Understand how reasoning fits into the workflow → See what it actually looks like to build with it No need to have used AI agents before. Just come ready to follow along and experiment. 🕒 15:00 UTC today If you’ve been putting off getting started, don’t overthink it. Join the workshop and build your first agent with us. Register → us02web.zoom.us/meeting/regi… Want to get familiar before we start? Docs → ambassadorss-organization.gi… See you in the workshop.

Technical Tuesdays is back: and we’re bringing signals straight from San Francisco. 👀 AGI-26 just happened. Now, what actually came out of it? Today at 5 PM UTC, we’re unpacking the research directions, Hyperon developments, and what’s next for the AGI Society. Dr. Matthew Iklé + Andrew Kemendo are joining us. If you’re building, researching, or simply trying to understand where AGI is heading pull up. 🔗 Set your reminder: linkedin.com/events/74928551…

By the end, a few things felt clear: 🔹 AI capability is moving fast 🔹 Access to compute is still unequal 🔹 Open and local models help close that gap 🔹 Expertise and critical thinking still matter more than ever 🔹 Ethics needs to be there from the start, not bolted on 🔹 Explainability is central to actually trusting these systems 🔹 Hybrid and alternative architectures are worth exploring 🔹 Developers and communities have real power in shaping what "responsible AI" looks like The future of AI isn't just about how capable we can make these systems — it's about what we choose to build, why we're building it, and who actually gets a seat at the table. This conversation is far from over. See you at AGI26. 🧵

A few projects and ideas got shared too: OmegaClaw, an open-source agent orchestration framework built on MeTTa; FabricPC, an open-source framework for persistent AI companions with long-term memory and adaptive behaviour; and a lot of interest in predictive coding as a more computationally efficient approach. There was also a good back-and-forth about truth, subjectivity, responsible data use, and how AI systems should actually explain the conclusions they reach.

We also got into whether just scaling up existing approaches is really the answer, or whether it's time to look at other paths; predictive coding, graph-based reasoning, logical networks, hybrid architectures, modular and interpretable systems. Combining language models with separate reasoning engines came up as one way to get more transparent, explainable systems. The point wasn't "bigger is better" , it's that there's more than one way to build capable AI.

Explainability came up again and again as one of the strongest themes of the day, systems that actually let people understand how they work, instead of just spitting out answers people are supposed to trust blindly. And ethics isn't only about avoiding harm; people brought up real examples of AI genuinely helping: computer vision for visually impaired people 👁️, AI-enabled prosthetics 🦿, educational tools 🎓, accessibility tech ♿. Building something that helps people is its own kind of ethical AI.

One of the bigger questions of the day: should ethics get bolted on after a system is built, or built in from day one? The room was pretty clear on this; ethics can't be an afterthought, especially in hackathon settings where speed and flashy demos tend to get prioritized over alignment, explainability, and long-term impact.

Zooming out, AGI looks different depending on where you're standing. Compute, education, regulation, infrastructure, culture, economic priorities, investment; all of these shape how a region actually experiences AI. Hackathons, sponsored programmes, education initiatives, and just getting hands-on access to tools came up as real ways to close some of that gap.

There was also a good conversation about newcomers. AI is lowering the barrier to entry for people getting into tech, which is great, but access to AI tools isn't the same as expertise. It can hallucinate, get things wrong, and sometimes just reinforce whatever assumption you walked in with instead of challenging it. Human judgment and critical thinking aren't going anywhere.

Open-source and locally-runnable models are chipping away at that. Models that used to need serious cloud infrastructure can now run on a lot less, which opens the door for more people to experiment and pushes things in a more decentralized direction. Hardware still matters though; using AI is getting easier, but building it is still resource-heavy. Computational equity isn't a solved problem.

We started by talking about how fast things have moved in the last 12–18 months. The gap between having an idea and actually shipping something with AI has shrunk massively. But someone in the room made a point that stuck; that speed isn't evenly distributed. Servers, GPUs, cloud infrastructure still cost real money, and that's still a wall for a lot of developers.