Good morning, welcome to this week’s Lore Brief, your 3-minute brief of the most important moves in AI and tech.
This issue is brought to you by Factory, the fastest way to ship software with autonomous coding agents.
Tesla unveils TERAFAB chip factory to fuel galactic ambitions → Tesla, SpaceX, and xAI are teaming up to build the world’s largest semiconductor facility capable of producing a terawatt of chips per year. The project aims to support millions of Optimus robots, solar-powered AI satellites in orbit, and the infrastructure needed to capture vast solar energy in space. Read more here →
OpenAI shuts down Sora app and reallocates expensive GPUs → The standalone Sora video app is being discontinued as OpenAI shifts resources back to ChatGPT, Codex, and daily-use tools after burning roughly $10-15 million per day, or about $5.4 billion annually. Read more here →
NVIDIA CEO Jensen Huang thinks AGI has arrived → In a new interview, Jensen Huang stated he believes current AI systems have achieved artificial general intelligence, a claim that immediately sparked debate over definitions and market implications. The comment comes as NVIDIA continues to dominate the hardware powering frontier models. Read more here →
Sign of the times:
Next level robotic hand:
High praise by Naval for the new book about Elon Musk:
Impressive viral clip about reverse aging developments:
Claude gains full computer control → Anthropic’s new feature lets Claude open apps, navigate browsers, fill spreadsheets, and handle desktop tasks directly on your Mac, working from phone instructions while you step away. It starts with connected tools like Slack and Calendar, then asks permission for anything else. Available now on Pro and Max plans. Read more here →
Sahil Lavingia turns his book into ready-to-run Claude Skills → The Minimalist Entrepreneur is now nine executable skills inside Claude Code covering everything from finding your community and validating ideas to shipping an MVP, landing first customers, and running a minimalist review on any decision. The skills are open source and turn static advice into actionable workflows you can trigger directly. Read more here →
Google releases Lyria 3 Pro music model with longer tracks → The upgraded model from DeepMind now generates full tracks up to three minutes with more creative control and structural understanding, a big step beyond the previous 30-second limit in Gemini. It is rolling out to paid subscribers in the Gemini app, via the API in AI Studio, and globally in Producer AI. Read more here →
Claude Code adds auto mode for hands-off task execution → Instead of approving every file write or terminal command, the new research preview lets Claude decide permissions itself while built-in safeguards block risky actions and force safer alternatives. It is available now on Team plans with enterprise and API access coming soon. Read more here →
Jason Fried makes Fizzy Kanban tool completely free → The clean, lightweight project board from 37signals is now free for everyone, with the open-source self-hosted version already free and the SaaS version subsidized by Basecamp and HEY revenue. A CLI is coming soon, the iOS app is in review, and Android is already live. Read more here →
OpenAI shares prompt tips for better frontends with GPT-5.4 → The new guide shows how tighter constraints, visual references, and real content help the model produce more intentional and delightful interfaces instead of generic layouts. It highlights the importance of low reasoning levels at the start and feeding the model actual design-system tokens for sharper results. Read more here →
