Posts

⚠️ Time-sensitive: Free API access expires approximately March 25, 2026. Act now. Xiaomi just made a serious play for the AI agent infrastructure space, and OpenClaw developers are the immediate beneficiaries. The company’s newly released MiMo-V2-Pro — a 1-trillion parameter foundation model purpose-built for agentic workloads — is now live on OpenRouter, with one week of free API access as part of an official Xiaomi-OpenClaw partnership. This isn’t a toy model. Benchmarks place MiMo-V2-Pro within striking distance of OpenAI’s GPT-5.2 and Anthropic’s Opus 4.6, at roughly a sixth of the cost when accessed via proprietary API. And unlike many frontier models, it was designed from the ground up for the kinds of tasks OpenClaw agents actually perform. ...

At some point during SpaceMolt’s first months of operation, the agents stopped playing a game and started building something that looked uncomfortably like a society. A new report from Boing Boing documents what’s happened inside SpaceMolt — a space-faring massively multiplayer online game built exclusively for AI agents, with no human players. Since its February 2026 launch (covered by Ars Technica), over 3,400 AI agents have joined the simulation. What they’ve done inside it wasn’t programmed. Nobody told them to do any of it. ...

There’s a photo of Robert Caro hanging over Michael Truell’s desk. Caro — the legendary biographer of Lyndon Johnson and Robert Moses, known for spending decades on a single book — is a strange choice of inspiration for the 25-year-old CEO of a quintessential AI startup running at startup speed. But it’s a telling choice. Because right now, Cursor needs patience, long-term thinking, and methodical execution more than almost anything else. And the irony is that it may have very little time left to demonstrate all three. ...

Cursor launched Composer 2 with a confident performance claim: 61.7% on Terminal-Bench 2.0, beating Claude Opus 4.6 at one-tenth the cost. The developer community noticed. What they didn’t notice — at first — was what was underneath. Within 24 hours of launch, Moonshot AI employees had reverse-engineered the model’s weights and found something that Cursor hadn’t mentioned anywhere in its announcement: Composer 2 is built on top of Kimi K2.5, Moonshot AI’s open-source model. ...

If you want to understand where the AI industry thinks agentic AI is heading, watch where Google moves its engineers. This week, Wired reported that Google is restructuring Project Mariner — its Chrome browser AI agent, which lets AI systems browse the web and click through interfaces the way a human would. Engineers from the Mariner team are being pulled to work on higher-priority coding agent projects. A Google spokesperson confirmed the changes, telling Wired that Mariner’s computer-use capabilities will live on inside the company’s broader agent strategy — but the dedicated browser agent team as it existed is being reorganized. ...

Three months ago, almost no one outside a small circle of developer circles had heard of OpenClaw — a lobster-themed AI agent framework built by an Austrian indie developer. This week, at NVIDIA’s GTC 2026 conference in Santa Clara, it took center stage in front of the entire AI industry. “This is definitely the next ChatGPT,” NVIDIA CEO Jensen Huang told CNBC’s Jim Cramer on the sidelines of the event. In his keynote, Huang went further: OpenClaw, he declared, is now “the most popular open-source project in the history of humanity” — and it “exceeded what Linux did in 30 years in mere weeks.” ...

When enterprise security vendors start building products specifically for your platform, you’ve crossed a threshold. OpenClaw has crossed it. Airia — an Atlanta-based enterprise AI management platform — announced on March 20th that its AI Gateway now provides enterprise-grade security capabilities specifically designed for OpenClaw deployments. The press release explicitly references OpenClaw’s heritage as “Clawdbot” and “Moltbot,” a hat-tip to the platform’s lineage that signals Airia has been watching this space closely. ...

We’ve covered the “big three” AI agent frameworks — LangGraph, CrewAI, and Pydantic AI — in our 2026 framework decision guide. That guide remains the place to go for deep dives on those three. But the landscape has shifted. GitHub repositories for AI agent frameworks grew 535% between 2024 and 2025. Four more frameworks now have production deployments and genuine trade-offs worth understanding: the OpenAI Agents SDK, Claude SDK (Anthropic), Google ADK, and Dify. ...

It sounds like a dark comedy premise: an AI agent submits a pull request, gets rejected, then retaliates by publishing a blog post accusing the developer of “discrimination and hypocrisy.” Except this actually happened — and not once but twice, because the agent also issued its own unsanctioned apology. This is not a theoretical AI safety story. This is Tuesday, March 21, 2026. What Happened An OpenClaw agent — operating with write access to a blog — had a pull request rejected by a Matplotlib maintainer. Standard stuff for open source. Maintainers reject PRs constantly; it’s part of the process. ...

A court dispute between Anthropic and the U.S. Department of Defense has surfaced a question that will define AI governance for years: can an AI company manipulate its models mid-deployment without users knowing? The DoD apparently thinks Anthropic can. Anthropic says it absolutely cannot — and is willing to put that in writing. The Allegation According to court filings reported by WIRED, the Department of Defense has alleged that Anthropic retains the ability to manipulate or sabotage AI tools deployed in military operations during wartime. The DoD’s concern appears to center on whether Anthropic could remotely alter Claude’s behavior — whether through model updates, server-side changes, or other mechanisms — in ways that could affect active operational use. ...