Elon Musk's
Your phone buzzed. You ignored it. Big mistake.
Tech moves fast. But today? Today it moved at light speed. From chips that will power your next phone to AI that can see through your screen. Here is what you missed.
Terafab. That is the name. The goal is annual production of 1 terawatt of computing power. The location is Austin, Texas. The operators are Tesla and SpaceX.
Musk said it plain: "Either we build Terafab, or we have no chips." The semiconductor industry cannot keep up with his demand. So he is building his own supply chain. The chips are for robots, AI, and — get this — space data centers.
What this means for you: When Musk builds, prices drop. When prices drop, technology spreads. In five years, your devices will run on chips designed in Austin and manufactured by a company that also builds rockets.
OpenRouter is the world's largest AI model API aggregator. Their latest data is staggering. Chinese AI models now handle 4.69 trillion Tokens per week. That is two weeks in a row they have beaten the US.
The top three most-used models globally? All Chinese.
Morgan Stanley predicts: China's AI inference Token consumption will grow from 10,000 trillion in 2025 to 3,900,000 trillion by 2030. That is a 389-fold increase.
Hidden fact: China has set a target to grow its AI-related industry to 10 trillion yuan by 2030. They are treating AI like the US treated the internet in the 90s — as an economic engine.
Apple wanted to turn the iPhone 18 Pro into a professional-grade camera. They considered buying Lux Optics, the company behind the Halide app. Then things got messy.
Ben Sandofsky, co-founder of Lux Optics, sued the other co-founder, Sebastiaan de With. The allegation? De With took confidential source code and documents when he joined Apple.
Takeaway: If you were waiting for the next iPhone's camera to replace your DSLR, wait longer.
Austin again. Amazon's Trainium chip lab is now open. The goal? Compete directly with Nvidia's GPUs. Trainium3 is designed for inference — the process AI uses to generate responses. Operating costs are estimated to be 50 percent lower than Nvidia.
AWS has already deployed over 1.4 million Trainium chips. Anthropic uses over 1 million Trainium2 chips for their Claude model. OpenAI and Apple are paying attention.
What you need to know: When AWS builds custom chips, prices drop across the entire cloud ecosystem. Your AWS bill? Might get cheaper.
Lace is the name. They are based in Norway. They just raised $40 million. Microsoft's venture arm M12 is in. Atomico led the round.
What do they do? Instead of using light to print circuits like ASML, they use a helium atom beam. The beam is 0.1 nanometers wide. That is the width of a single hydrogen atom.
Current ASML tools use 13.5 nanometer light. The difference is not small — it is atomic.
Implication: This could shrink transistors to "almost unimaginable" sizes. AI processors could become exponentially more powerful. Lace aims to have a test tool in a pilot fab by 2029.
The National AI Application Test Base for the energy sector opened in Guangzhou. Huawei, ZTE, and Baidu have already moved in. The plan? Bring in 30 companies this year. Test AI applications for power grids at scale.
China has 2.1 million EV charging stations nationwide. That is 48 percent growth in one year. The grid has to handle that load. AI is how.
For India: Watch this space. Our EV charging infrastructure is growing. AI-managed grids are coming.
Google is preparing to launch Gemini AI for macOS. The killer feature? Desktop Intelligence. It can mirror your screen, pull data from your calendar, documents, browsers — and understand what you are actually doing.
ChatGPT and Claude already have native macOS apps. Gemini is joining the fight. No release date yet, but external testing suggests it is soon.
Check your devices. If you use AWS, look for Trainium-based instances. If you use iPhone, hold off on the camera expectations. If you use Android, Gemini on Mac might be your bridge between ecosystems.
Tech news is not just about gadgets. It is about supply chains. It is about geopolitics. It is about who controls the chips, the AI, the data. Today, we saw shifts in all three.
Ask yourself: Where is your next device coming from? Who makes its brain? The answer is changing.