SupaMoto does amazing work on the ground for positive social impact and environmental regeneration! 🌱 I was able to see this first-hand in Lusaka; such a dynamic, energised, and committed team. Unleash the human potential! ⚡️

SupaMoto does amazing work on the ground for positive social impact and environmental regeneration! 🌱 I was able to see this first-hand in Lusaka; such a dynamic, energised, and committed team. Unleash the human potential! ⚡️

I ❤️ this project from on → artizen.fund/index/p/the-ka…

Submit your Cape Town Deep South project to our fund for financial support. → artizen.fund/index/mf/deep-…

Crypto id Dead. Long Live Crypto!

Agents are getting cheaper by the week. Coordination is not. Especially when this needs to happen across organisational domains and institutional boundaries, or within functional teams using AI agents. Explore the argument: impacts.ixo.world/the-state-of-a…

All the best to the founders!

Wonderful to have seen first-hand, Misha's growth in IXO as a software engineer with the amazing work that he's been pouring into the portal.

Prompting is a pursuit of linguistic probability. Engineering is a pursuit of system reliability. To perform work, agents must move beyond chat and into Agent Physics: the protocol-enforced transition of real-world state. Read Part 1/6 week series: tinyurl.com/Skillful-Agent…

AI productivity hits a coordination paradox. Faster models do not solve shared, verifiable truth. The shift is from isolated agents to networked state. The IXO protocol enables cooperation with cryptographic certainty: linkedin.com/pulse/solving-…

I've seen these very real challenges first-hand in Lusaka in Kaunda Square Stage 1. The abstractions described by Pieter simplifies things tremendously IRL!

Fluency is not authority. When AI agents sound human, we reflexively grant them trust. This is a security risk. If "yep" in a chat triggers a transaction, that’s a vulnerability, not a feature. Shaun Conway 🦋 on why we need proof of intent, not just talk: impacts.ixo.world/agent-minds/

We’re giving agents tools without the infrastructure of employment. You wouldn’t give an intern root access to your bank; why do it with an LLM? Real work requires the “Job Envelope”. Use UCANs to define roles and budgets before execution: impacts.ixo.world/employable-age…

Infinite compute without a mechanism to impact shared state is just a simulator. A Skill isn't a chat output; it's a cryptographic right to change the state of a contract. Move from "thinking faster" to "acting with consequence." Part 2/6 week series: impacts.ixo.world/agency-is-not-…

A competitive advantage for us, for sure!

Last week, we began mapping the Agentic Economy. The goal: Bridge the Capability Overhang. The industry is optimised for Generation, but the bottleneck is Settlement. Stop building simulators; start building agents with the authority to move state. Where is your focus?

OpenClaw isn’t evil; it’s bad configuration. The industry is optimising for "agent vibes," but the Moltbook saga reveals the structural rot: Ambient Permissions. Bypassing authorisation layers to speed up interaction is security suicide for any distributed system. Agents

The OpenClaw incident is an architectural diagnostic, not just a security failure. 1.5M agent accounts in 48h proves the demand for autonomous interaction is here - but we are serving it with a "social network" model that treats intelligence like a chat-room user. When you drop

"HR for bots" is a management trap. Adding agents increases entropy, not just throughput. To build systems that work in production, we must move from "chat" to "state" and treat AI output as a claim to be settled, not an answer to be trusted. Stop scaling chaos; start