OpenClaw is now reportedly at 366,000 GitHub stars — the most-starred software project in history — with creator Peter Steinberger's TED talk on the East/West AI agent divergence going viral. The AI-writes-code revolution is accelerating: Anthropic's own engineers reportedly stopped writing code months ago, and Snap's Evan Spiegel disclosed that two-thirds of Snap's code is now AI-built.
Two major regulatory developments broke this morning: the EU AI Act deal collapsed with member states and Parliament unable to agree on sector exemptions, while New York's RAISE Act (transparency and safety requirements for frontier AI models) is now in force. Meanwhile, a sobering ICLR 2026 paper warns that more capable reasoning models actually hallucinate more tool calls — a paradox at the heart of the agentic boom.
A new managed cloud platform makes OpenClaw self-hosting trivial — WhatsApp, Telegram, Discord, Slack, Signal, and iMessage as frontends with zero CLI. Designed for non-developers who want to chat with their AI agent like a colleague. — Breaking AC, Apr 28
A new ICLR 2026 paper delivers uncomfortable news: train a model to reason harder and it will hallucinate more of the tools it should call. This is the central reliability paradox for agentic systems — more intelligence ≠ more reliable action. — Asanify, Apr 29
Amazon introduced agentic hiring software that eliminates the face-to-face interview for its massive holiday hiring cycle. Goal: "humanize AI" while removing humans from the process — a striking signal of how deeply agents are penetrating enterprise workflows. — Reuters, Apr 28
A last-ditch effort to salvage a watered-down EU AI Act failed this morning. Member states and Parliament couldn't agree on sector exemptions for industries already subject to product safety rules. Europe's AI governance remains in limbo. — Reuters, Apr 29
New York's RAISE Act — imposing transparency, compliance, safety, and reporting requirements on frontier AI model developers — took effect March 19, with amendments signed March 27. The most significant U.S. AI regulatory action to date for enterprise builders. — Alston & Bird, Apr 2026
Selected posts from the edition, with links to the originals.
AI/ML · VIP "we also need inference chips for MCP servers and agentic terminal use. WAIT A MINUTE…"
View post on X ↗"I JUST READ SOMETHING INTERESTING! ANTHROPIC ENGINEERS STOPPED WRITING CODE MONTHS AGO. Everyone still coding by hand just fell behind. And they do not even know it yet. This is not a headline. This is a new hire at Anthropic describing what his team actually does every day."
View post on X ↗"Evan Spiegel said two-thirds of his company's code is now built by AI. That's changed how he thinks companies will use their resources."
View post on X ↗"claude code users when they realize this.md file will 10x their outputs"
View post on X ↗"HOLY MOLY running a 35B model locally on a MacBook shouldn't be THIS FAST 🤯 Spent my weekend in @atomic_chat_hq testing Qwen…"
View post on X ↗"I pulled 3,182 tweets analyzing @NousResearch Hermes Agent versus Claude Code to understand exactly why developers are switching…"
View post on X ↗"things are moving so fast in this industry that there will be MANY important clients and that OpenAi's importance to the 'movement' will be minimized more"
View post on X ↗"Congratulations to @NVIDIAAI on launching Nemotron 3 Nano Omni — day-0 MLX support 🚀 Highlights: video+speech comprehension, GUI automation, OCR, and speech transcription."
View post on X ↗⚠️ AI Governance @Reuters "Elon Musk says OpenAI was his idea, before executives looted it"
View post on X ↗OpenClaw community "Hello @karpathy — it would be insanely valuable if you made a 'Deep dive into AI agents like OpenClaw' or whatever agent tools you're using now"
View post on X ↗