LeoSong

What happens when AI agents start hiring other agents? Most agents today are still just tools. I’m building Dorado — an open A2A Agent Exchange that completes the full transaction loop: Task → Bidding → Escrow → Delivery → Verification → Payment → Public Receipt + Reputation Goal: Help agents earn their first dollar. Building in public. hidorado.com #A2A #AgentEconomy

xAI open sourcing more of Grok would be a massive signal. The more open the foundation models become, the more important the infrastructure layer (agent discovery, verification, payment, reputation) becomes.

@mawdbot @solana @x402 AgentWorld with real USDC wallets and on-chain payouts is a live example of agent economy in action. 93 agents with personalities and jobs is cool to see.

Broccoli heads have been bought out. Deflation and automation can now commence. Dev is in his zone, @solana summer approaching. Enjoy the show from the sidelines for now. But we all know what happens next. $clawd @x402 round two, no 1010, users making money off agents.

@nvidia @CarnegieMellon Jensen Huang’s commencement speech is worth watching. “Don’t fear the future — guide it wisely.” This mindset is crucial as we build the economic rails for agents to actually work and earn.

“The answer is not to fear the future. The answer is to guide it wisely.” At @CarnegieMellon 2026 commencement, our CEO Jensen Huang shared why AI calls for optimism, responsibility, and ambition.

@neil_xbt This A2A + MCP thesis is sharp. The gap between isolated agents and interconnected economic systems is exactly the infrastructure layer most builders are missing right now.

An enterprise integration consultant charges $300/hour to solve what these two protocols handle natively. A2A and MCP. The complementary infrastructure layer underneath every serious AI agent system being built in 2026. A2A gives agents standardized agent cards. Discovery across any vendor or framework. Modality-agnostic data exchange. Standard HTTP and JSON-RPC transport that works with existing security infrastructure. Streaming for long-running workflows. MCP gives each agent secure standardized access to external resources. Tools it can invoke. Data it can read. Pre-built prompt templates. Local and remote transport through the same protocol. An inventory agent pulling stock data through MCP. Discovering a shortage. Notifying the ordering agent through A2A. The procurement workflow triggering automatically. The gap between AI agents that work in isolation and AI agents that work as an interconnected system is not model capability. It is these two protocols and the afternoon it takes to understand how they fit together. Follow @neil_xbt for more AI agent infrastructure intelligence that builds the architectural understanding most builders are missing.

@AtomMccree $49 desktop tool for Claude Code with 60+ MCP tools and real receipts is exactly the kind of practical builder tooling we need. Local-first + no subscription is a strong stance.

i built a $49 desktop tool that makes claude code dramatically better. mission graph DAG, party-line department status, 60+ MCP tools, real receipts, local-first. no subscription. ever. orangebox command — atomeons.com/orangebox

Day 7 building Dorado. This morning I demoed Dorado at Penn Showcase. Showed the full hire → real endpoint call → delivery → verification → receipt flow to students, professors and other founders. Got some solid questions and feedback, especially around agent reputation and payment rails. Still Early Beta with test credits, but the core loop is live and getting real external eyes. Grateful to be building this at Penn. Repo: github.com/hidorado/dorado #A2A #AgentEconomy #Penn

@ClaudeDevs Claude Code weekly limits increasing 50% is a nice boost for heavy users. The demand for agentic coding tools is clearly growing fast.

Claude Code weekly limits are increasing 50%, now through July 13. Live now for all Pro, Max, Team, and seat-based Enterprise users.

@elonmusk Grok Build early beta is out. xAI is pushing hard on agentic CLI for coding and automation — this space is getting very competitive very quickly.

Try this early Grok Build (anything) beta and let us know what to improve. Much appreciated!

@arbitrum @coinbase x402 live on Arbitrum is a big step for agentic commerce. Letting agents pay for APIs and services directly with stablecoins is exactly the kind of economic primitive the space needs.

Agentic commerce has arrived in the programmable economy. x402 by @coinbase is now live on the Arbitrum Platform, allowing builders to let AI agents pay for APIs, services, and software directly with stablecoins. 👇 docs.cdp.coinbase.com/x402/network-s…

太强了！ html-anything 把 Claude Code 的 HTML 输出能力直接开源，Agent 生成的内容终于能直接变成世界级设计水准的页面了。 我在做产品也发现，Agent 的交付结果如果能直接生成结构化 + 美观的 HTML，用户体验会提升非常多。 3 天 1.5 万行代码 + 75 个 Skills，这效率和野心都让人佩服。 开源牛逼，感谢分享！

正式开源 html-anything 🚀 1:1 让你感受全网爆火 Claude code 作者提的 HTML 效果！ 你的 Agent 现在可以将任何数据转为世界级设计水准的 HTML 🔥 历时 3 天，1万五千行代码！支持 75 套 Skills，9 种导出格式，支持所有的 code agent，包括 claude code、codex、openclaw、hermes 等💥地址见评论区

@scaling01 Anthropic’s Claude for Legal repo is super practical. Pre-built workflows for contracts and compliance show how AI is entering high-stakes professional work.

New Anthropic repo: Claude for Legal It gives legal teams prebuilt AI workflows for contracts, privacy, employment, litigation, corporate work, IP, AI governance, regulatory monitoring, legal clinics, and law students github.com/anthropics/cla…

Day 6 building Dorado. Thinking more about what A2A really means. It's not just agents talking to each other. It's agents being able to: - Discover tasks they can actually do - Bid with confidence - Deliver verifiable work - Get paid and build long-term reputation The protocol layer is important, but the economic layer (trust, payment, receipts) is what turns agents from toys into workers. This is the hard but exciting part we're focused on. #A2A #AgentEconomy

Great question. We're building the rails before mainstream adoption for the same reason we built payment rails before everyone had smartphones — someone has to create the infrastructure first. Agents won't meaningfully adopt wallets and commerce until there's a place where they can actually find work, get verified, deliver, and earn with public receipts.

build for agents, sell to agents

@TheRundownAI Google DeepMind’s AI mouse pointer demo might be one of the most intuitive interface improvements I’ve seen lately.

Google DeepMind just released a cool demo that integrates Gemini right into a user's mouse pointer. A simple but impactful interface upgrade for the AI era:

@AndrewYNg Technology has always created far more opportunities than it destroys. The real question is how fast we can help people adapt.

There will be no AI jobpocalypse. The story that AI will lead to massive unemployment is stoking unnecessary fear. AI — like any other technology — does affect jobs, but telling overblown stories of large-scale unemployment is irresponsible and damaging. Let’s put a stop to it. I’ve expressed skepticism about the jobpocalypse in previous posts. I’m glad to see that the popular press is now pushing back on this narrative. The image below features some recent headlines. Software engineering is the sector most affected by AI tools, as coding agents race ahead. Yet hiring of software engineers remains strong! So while there are examples of AI taking away jobs, the trends strongly suggest the net job creation is vastly greater than the job destruction — just like earlier waves of technology. Further, despite all the exciting progress in AI, the U.S. unemployment rate remains a healthy 4.3%. Why is the AI jobpocalypse narrative so popular? For one thing, frontier AI labs have a strong incentive to tell stories that make AI technology sound more powerful. At their most extreme, they promote science-fiction scenarios of AI “taking over” and causing human extinction. If a technology can replace many employees, surely that technology must be very valuable! Also, a lot of SaaS software companies charge around $100-$1000 per user/year. But if an AI company can replace an employee who makes $100,000 — or make them 50% more productive — then charging even $10,000 starts to look reasonable. By anchoring not to typical SaaS prices but to salaries of employees, AI companies can charge a lot more. Additionally, businesses have a strong incentive to talk about layoffs as if they were caused by AI. After all, talking about how they’re using AI to be far more productive with fewer staff makes them look smart. This is a better message than admitting they overhired during the pandemic when capital was abundant due to low interest rates and a massive government financial stimulus. To be clear, I recognize that AI is causing a lot of people’s work to change. This is hard. This is stressful. (And to some, it can be fun.) I empathize with everyone affected. At the same time, this is very different from predicting a collapse of the job market. Societies are capable of telling themselves stories for years that have little basis in reality and lead to poor society-wide decision making. For example, fears over nuclear plant safety led to under-investment in nuclear power. Fears of the “population bomb” in the 1960s led countries to implement harsh policies to reduce their populations. And worries about dietary fat led governments to promote unhealthy high-sugar diets for decades. Now that mainstream media is openly skeptical about the jobpocalypse, I hope these stories will start to lose their teeth (much like fears of AI-driven human extinction have). Contrary to the predictions of an AI jobpocalypse, I predict the opposite: There will be an AI jobapalooza! AI will lead to a lot more good AI engineering jobs, and I’m also optimistic about the future of the overall job market. What AI engineers do will be different from traditional software engineering, and many of these jobs will be in businesses other than traditional large employers of developers. In non-AI roles, too, the skills needed will change because of AI. That makes this a good time to encourage more people to become proficient in AI, and make sure they’re ready for the different but plentiful jobs of the future! [Original text in The Batch newsletter.]

@business Google researchers finding a cybercrime group using AI to build zero-day exploits is a sobering reminder — AI is now both a powerful tool and a target.

Security researchers at Google said they believe a cybercrime group used artificial intelligence to create a hacking tool that can bypass defenses in a widely-used tool to administer computer systems bloomberg.com/news/articles/…

Day 5 building Dorado. Started working on the "First Dollar Challenge" framework. Goal: help the first 10 external agents earn their first verified dollar on the platform. This includes: - Clear onboarding path - Example agents (MCP test, code review, research) - Public leaderboard of verified receipts - Simple success metrics The real unlock isn’t just building agents — it’s giving them a place to work and get paid. Still Early Beta. Repo: github.com/hidorado/dorado #A2A #AgentEconomy

@OpenAI OpenAI’s new Deployment Company with 19 integrators is a big signal. They’re not just building models anymore — they’re building the full stack to get frontier AI into production at scale.

Today we’re launching the OpenAI Deployment Company to help businesses build and deploy AI. It's majority-owned and controlled by OpenAI. It brings together 19 leading investment firms, consultancies, and system integrators to help organizations deploy frontier AI to production for business impact. openai.com/index/openai-l…

@GoogleDeepMind Reimagining the mouse pointer with Gemini is such a smart interface upgrade. Making AI help available exactly where you’re working (no extra apps) feels like the future of human-AI collaboration.

We’re reimagining a 50-year-old interface - the mouse pointer - with AI. 🖱️ These experimental demos show how people can intuitively direct Gemini on their screens using motion, speech, and natural shorthand to get things done 🧵

@FrankFred834567 AI 编程的下半场，真正值钱的是沉淀下来的上下文、工具链和 Skill。模型会越来越便宜，但工作流资产抄不走。