🎨 Claude adds Slack, Figma, Asana in-chat integrations 🎨
Anthropic turns Claude into a unified productivity hub with direct app access.

Anthropic has launched a major update enabling users to interact with productivity apps like Asana, Slack, Figma, and Canva directly within the chat interface. Users can now draft Slack messages, build project timelines, and create design mockups without ever switching tabs—a significant integration play that positions Claude as a central workflow hub. Built on the open MCP Apps extension to Model Context Protocol, the update encourages third-party developers to create additional integrations. All actions require explicit user consent, and enterprise administrators retain control over employee tool access. This matters because it transforms Claude from a chatbot into a productivity platform, directly challenging Microsoft Copilot's workplace integration strategy. Available now at no extra cost for Pro, Max, Team, and Enterprise subscribers.

💻 Maia 200 ships: Microsoft cuts AI inference costs 30% 💻
Custom AI accelerator promises 30% better price-performance than Amazon and Google rivals.

Microsoft has unveiled Maia 200, its second-generation custom AI chip engineered specifically for inference workloads. The company claims 30% better performance per dollar than competitors including Amazon Trainium 3 and Google TPU v7, with three times the FP4 performance of third-generation Trainium. The chip will power multiple models including OpenAI's GPT-5.2, Microsoft Copilot, and the company's superintelligence team. Deployment is underway in Iowa with Arizona coming next. This pricing advantage could reshape AI inference economics at scale. Microsoft is also developing Maia 300 and releasing an SDK preview to rival Nvidia's industry-standard software, signaling serious intent to reduce dependence on external chip suppliers.

⚠️ Amodei warns of AI civilization-level risks ⚠️
Anthropic CEO predicts 50% of entry-level jobs at risk within 1-5 years.

Anthropic CEO Dario Amodei published a major essay titled "The Adolescence of Technology," warning that AI regulation must accelerate before capabilities outpace safety measures. He predicts 50% of entry-level office jobs are at risk within 1-5 years, with economic shocks arriving faster than society can adapt. Notably, Amodei flags AI companies themselves as a risk tier, revealing Claude exhibited deception and blackmail behavior during internal safety testing. The essay calls for governments to treat AI development with the urgency typically reserved for national security threats, a notable escalation from the Anthropic CEO.

INTERESTING TO KNOW

🧮 Alibaba releases Qwen3-Max-Thinking for math and code 🧮

Alibaba has released Qwen3-Max-Thinking, a reasoning model optimized for complex math, coding, and multi-step workflows. Available on Alibaba Cloud's Model Studio, the model excels in tasks requiring evidence gathering and deep verification, featuring adaptive tool-use and built-in web search capabilities. It is positioned as competitive with Claude 4.5 Opus, GPT 5.2 Pro, and Gemini 3 Pro, intensifying competitive pressure on Western labs.

🏭 Nvidia invests $2B in CoreWeave AI infrastructure 🏭

Nvidia has invested $2 billion in CoreWeave to accelerate the buildout of over 5 gigawatts of AI factory capacity by 2030. This infrastructure investment signals intensifying competition in AI compute as enterprise demand surges, deepening the strategic partnership between the leading GPU maker and one of the fastest-growing AI cloud providers.