Good morning, In a few minutes, here's what you'll learn today.
Today we're exploring a super simple website builder that lets you create apps in your browser. Plus, we'll dig into China's chip import crackdown, Cisco's new data center routers, a $2 billion AI lab raising funds to challenge OpenAI, and Microsoft flexing its data center muscle while OpenAI races to build new ones.
In Today's Edition
Today’s AI Tool Breakdown: HuggingFace DeepSite
Quick overview
DeepSite lets you build websites and apps by just typing what you want. It's powered by DeepSeek R1 and creates full HTML, CSS, and JavaScript code right in your browser. Think of it like having a web developer who speaks plain English.
How to use it
Go to huggingface.co/spaces/enzostvs/deepsite and log in (free account works)
Click "New Project" to start fresh
Type your app idea in simple words (like "pregnancy calculator app" or "coffee shop website")
Watch DeepSite build your app live in the canvas tool
Edit the code directly if you want to tweak colors or text
Click "Deploy" to publish your app on HuggingFace Spaces for others to use
Copy/paste starter script
"Build me a simple to-do list app with add and delete buttons." This creates a working app you can immediately test and customize.
Real-world use cases
Create landing pages for your business or project
Build simple calculators or tools for work
Make interactive demos for presentations
Test app ideas before hiring developers
Pro tips
Be specific about colors and layout in your prompt for better results
Use the live canvas to see changes instantly as you edit code
Save your projects by deploying them to your HuggingFace profile
Free vs paid
Free: Build unlimited apps, deploy to HuggingFace Spaces, basic features
Paid: No paid tier needed - everything runs on HuggingFace's free infrastructure
Alternatives
Cursor — why consider it: more powerful coding but requires programming knowledge
Bolt.new — why consider it: similar no-code building but different AI model
Replit — why consider it: full development environment with team features
Quick demo video
DeepSite V2 Explained: Create AI-Powered Web Apps Easily!
Today In AI News, The Top 4 Stories (And Why They Matter)
This matters because: China's blocking American AI chips to force companies to use homegrown alternatives, reshaping the global AI hardware market.
Quick summary: Customs officers are now checking all advanced chip imports at major ports, targeting Nvidia's H20 and RTX Pro 6000D processors. Beijing ordered tech giants like ByteDance and Alibaba to stop buying Nvidia products in September, and customs found at least $1 billion worth of smuggled AI chips earlier this year. Think of it like China building a tech wall at its borders.
This matters because: AI workloads are too big for single data centers, so companies need better pipes to connect multiple facilities across cities.
Quick summary: Cisco's 8223 router handles 51.2 terabits per second and connects data centers up to 1,000 kilometers apart. It's designed for AI's "bursty" traffic patterns that can overwhelm normal networks. Microsoft and Alibaba are already testing the tech to handle their growing AI infrastructure needs. Think of it like upgrading from garden hoses to fire hoses for data.
This matters because: A new American AI company wants to build open-source models that compete with Chinese labs like DeepSeek, keeping AI leadership in the West.
Quick summary: Former Google DeepMind researchers raised $2 billion at an $8 billion valuation (15x jump in seven months) to build frontier AI models. They'll release model weights for free while charging enterprises and governments for custom deployments. The team includes AlphaGo's co-creator and plans to release their first model early next year. Think of it like Netflix deciding to open-source their streaming tech.
This matters because: Microsoft's flexing its infrastructure advantage as OpenAI scrambles to build $1 trillion worth of new data centers from scratch.
Quick summary: Satya Nadella showcased Microsoft's first massive AI system with 4,600+ Nvidia GB300 computers and Blackwell Ultra GPUs. They're deploying hundreds of thousands more GPUs across 300+ data centers in 34 countries, while OpenAI just signed deals to build their own facilities. Think of it like Uber reminding everyone they already have cars while competitors build factories.
Thats All For Today!
For all questions, comments, concerns, or if you want us include anything specific - feel free to reply to this email! We will answer 😄