Hey there! Welcome to the latest Creators’ AI Edition.
AI just jumped from experiments to infrastructure. Alibaba’s Qwen3 is driving baseline costs down, ChatGPT Pulse pushes updates to you, and Microsoft is wiring Anthropic into the Office stack. Add a trillion-dollar buildout, economy-scale benchmarks, and Germany’s sovereign-AI push. Below: what to do this week, what to watch, and what to ignore.
Ready to see what this means for you? Let’s dive in! 🚀
Featured Materials 🎟️
News of the week 🌍
Useful tools ⚒️
Weekly Guides 📕
AI Meme of the Week 🤡
AI Tweet of the Week 🐦
(Bonus) Materials 🎁
Featured Materials 🎟️
Updates from Alibaba
Alibaba’s Qwen3 release wave includes Qwen-Max, a 1-trillion parameter model showing near-frontier coding and agentic performance, with the Heavy version acing math benchmarks. Qwen-Omni now handles text, images, audio, and video with multilingual speech support, while Qwen-VL tops open-source visual models and even beats closed rivals. Other launches include LiveTranslate-Flash for real-time interpretation, Guard models for safety, and upgraded Codervariants. Unlike slower Western rollouts, Alibaba is flooding the market with rapid, specialized frontier models, marking its strongest push since DeepSeek’s R1.
ChatGPT Pulse
OpenAI just launched ChatGPT Pulse, now in preview for Pro users on mobile. Unlike traditional Q&A, Pulse proactively researches overnight and delivers personalized daily updates as visual cards, drawing from your chats, feedback, and optional integrations like Gmail and Google Calendar. Users can curate topics, provide quick feedback, and even request tailored roundups (e.g., travel tips, local events, or study guides). The feature aims to shift ChatGPT from a reactive tool into a proactive assistant that helps plan, remind, and surface insights before you even ask, with the Plus rollout coming next.
News of the week 🌍
Microsoft 365 Copilot expands
Microsoft 365 Copilot, joining OpenAI’s models. Customers can choose between OpenAI and Anthropic for powering the Researcher agent, a reasoning tool that works across web, files, emails, and meetings, and when building custom agents in Copilot Studio. This expansion gives enterprises true model choice for deep reasoning, workflow automation, and agentic tasks, with rollout starting today for opted-in Copilot license holders through the Frontier Program.
Massive AI infrastructure buildout
OpenAI CEO Sam Altman announced plans to scale infrastructure to produce 1 GW of AI capacity weekly, arguing that more compute is essential to unlock breakthroughs from curing cancer to universal education. Altman said new infrastructure and financing announcements are coming soon, stressing U.S. competitiveness in chips and energy as rivals accelerate. The post follows Nvidia’s $100B investment in OpenAI infra, underscoring how compute expansion is becoming the core battleground for both business growth and global AI leadership.
AI’s real economic impact
OpenAI has introduced GDPval, a new benchmark designed to measure AI’s real economic impact rather than just abstract skills. Covering over 1,000 tasks across 44 professions in 9 high-GDP sectors, the benchmark uses realistic artifacts like legal documents, engineering blueprints, and support call transcripts, requiring outputs such as presentations, diagrams, and reports, not just text answers. Each task was created and validated by experts averaging 14 years of experience, with solutions compared against human work in rigorous multi-stage reviews. Early results show Anthropic’s Claude Opus 4.1 delivers human-level or better performance in 47.6% of tasks, while OpenAI’s GPT-5 slightly edges out o3, with GPT-4o serving as a baseline of last year’s progress. The findings suggest AI systems are beginning to approach professional parity on repeatable, well-defined tasks, though replacing entire professions remains far off.
CFA By AI
NYU research shows that frontier AI models from OpenAI, Google, and Anthropic can now pass all three levels of the CFA exam, including the notoriously difficult Level III essays that stumped them just two years ago. In testing, nine of 23 models scored above the 63% passing threshold, with OpenAI’s o4-mini leading at 79.1%, followed by Gemini 2.5 Pro (75.9%) and Claude 4 Opus (74.9%). Models completed the multi-year exam process in minutes, and human graders scored their essays 5.6 points higher on average than automated systems. The findings highlight a dramatic leap in reasoning ability, signaling that human financial professionals will increasingly compete on relationship skills and judgment rather than technical analysis alone.
OpenAI for Germany
SAP, OpenAI, and Microsoft announced a partnership to bring AI safely into Germany’s public sector with a sovereign setup powered by SAP’s Delos Cloud on Microsoft Azure. Launching in 2026, the initiative will let millions of government and research employees use AI under strict data sovereignty, security, and legal standards. SAP will expand Delos Cloud to 4,000 GPUs for AI workloads, with plans to scale further. The collaboration aims to automate workflows like records management and data analysis, boost productivity, and support Germany’s national AI strategy, which targets AI-driven value creation of 10% of GDP by 2030.
Useful tools ⚒️
Strata - One MCP server for AI agents to handle thousands of tools
Sprinto's AI - AI that answers Security Questionnaires & RFP like you would
Atla - Automatically detect errors in your AI agents
Vibe n8n - Cursor/Lovable for n8n to generate workflows in seconds
Doraverse - Your all-in-one AI coworker for office work
Doraverse unites the smartest of 15+ AI models & 50+ app integrations, and no-code workflow automation, everything in one platform to transform how your business thinks, creates, and grows
Weekly Guides 📕
How to Create 3D Animation Videos for YouTube with AI (Full Guide)
5 Secret Ways to Use ChatGPT Like an AI Pro (2025 easy prompt guide!)
Build Anything with Codex, Here’s How
AI Meme of the Week 🤡
AI Tweet of the Week 🐦
The money cycle in AI
(Bonus) Materials 🎁
ElevenLabs is now valued at $6.6B
Elon Musk Dismisses Anthropic’s Prospects, Says ‘Winning Was Never Possible
AI tool helped recover £500m lost to fraud
Nvidia Invests in UK AI Voice Startup ElevenLabs
Perplexity Launches Email Assistant
If you missed our previous updates, don’t worry, here they are: