Magica logo

🚀 SpaceX bought a code editor

Editorial Team
·

📖6 min read

Cover image for 🚀 SpaceX bought a code editor

🚀 SpaceX Bought Cursor Because Why Not

  • SpaceX just YEETED $60 billion in stock to buy AI coding startup Cursor, days after IPO-ing into a $2.6 trillion company that briefly passed Amazon.
  • Musk now has Cursor's code models running on SpaceX's absurd compute. It's giving "vertically integrate the entire universe" energy.

💰 DeepSeek Raises $7.4B and Founder YOLOs His Own Cash

  • China's DeepSeek raised $7.4 billion at a $50B+ valuation - founder Liang Wenfeng threw in $3 billion of his OWN money and locked investors into a brutal five-year deal.
  • This is the loudest "China is playing to win" the AI race has heard yet. The money pile is biblical.

🛑 US Bans Anthropic for Allies, Sales Go BRRR Anyway

  • The US told the UK "no" - allied nations can't touch Anthropic's best models over national security fears. And yet Anthropic just passed OpenAI in enterprise market share.
  • Getting labeled "too dangerous" by your own government?? Best 👏 marketing 👏 ever. Businesses want it MORE now.

💼 Microsoft's Copilot Cowork Goes Global

  • Microsoft launched Copilot Cowork worldwide - an AI agent that does long, hard tasks start to finish. You pay only for what you use, and it's 30-40% cheaper than rivals.
  • AI is done with the chitchat era. It's now doing your actual job inside company systems. Cool cool cool.

🧠 Z AI Drops GLM-5.2 for FREE to Fight GPT 5.5

  • Chinese lab Z AI released GLM-5.2 - a free open model going toe-to-toe with GPT 5.5 and Claude Opus 4.8. It reads a wild 1 million tokens at once with special math/coding modes.
  • Giving away top-tier AI for free is a flex that lowkey terrifies every paid model on Earth.

📱 Android 17 Stuffs Gemini Into Your Pocket

  • Google dropped Android 17 with the Gemini Omni model baked into Pixel phones, plus better multitasking and live safety checks to stop threats in real time.
  • Google's using its massive phone base to put AI literally in your hand - and keep Apple eating dust.

⚖️ DOJ Defends xAI's Gas Turbines as "National Security"

  • The DOJ stepped in to protect xAI's gas plants, claiming shutdowns would hurt US national security. A civil rights group sued Musk over 57 gas turbines polluting Memphis air.
  • The government just picked a side: AI growth > local clean air, if the military says so. Yikes.

🏢 Meta Promises to Fix Its Vibes After AI Reorg Meltdown

  • Meta's CTO promised a culture reset after an AI team shakeup left workers furious. Staff got force-moved to AI teams AND had their mouse clicks tracked to train AI.
  • Even trillion-dollar giants can't bribe their way past angry humans. Sir, this is a workplace.

💻 Cursor Launches "Origin" to Fight GitHub

  • AI coding tool Cursor announced Origin - a new platform built to take on GitHub directly - plus a new iPhone app so coders can ship from anywhere.
  • AI writes code faster than ever, so Origin wants to be the new home base for devs. The way that Cursor just casually went after GitHub...

📈 Wall Street Wants You to Buy "MANGOS" 🥭

  • Wall Street is cooking up new AI ETFs based on "MANGOS": Meta, Anthropic, Nvidia, Google, OpenAI, and SpaceX - built to replace the old "Magnificent 7."
  • Banks are inventing fresh ways for normies to bet on AI. It's giving "the whole market IS AI now."

🧑‍⚖️ Workday Hit With Big AI Hiring Bias Lawsuit

  • A judge ruled Workday must face a lawsuit claiming its AI hiring software broke California employment law by blocking job seekers based on race, age, and mental health.
  • This is the first big legal fight over AI hiring tools - and it could rewrite how every company screens resumes.

🎙️ Plaud Hits $100M Selling AI Notetakers

  • Hardware maker Plaud hit $100M in yearly sales with 2 million+ AI voice recorders. The little gadgets stick to your phone and record, transcribe, and summarize meetings - no screen.
  • Plaud proves simple AI hardware can win HUGE by doing one job really well. We love to see it.

🤖 Genesis AI Reveals "Eno," a Robot With Wheels (Not Legs)

  • French startup Genesis AI showed off Eno - a robot using wheels and a folding tower instead of human legs. Backed by Eric Schmidt, they raised $105 million.
  • While everyone chases humanoid robots, Genesis bets simple non-human shapes actually work better for factories and hospitals.

🏦 HSBC Teams Up With Google Cloud

  • HSBC signed a big deal with Google Cloud, using Gemini to automate 200 tasks across wealth management and financial crime detection.
  • Banks are in a full-blown AI sprint to cut costs and move faster. The British banking era of robo-everything begins.

🇮🇳 Wipro Opens an Anthropic Claude Center in India

  • Indian IT giant Wipro opened a Bengaluru hub built just for Anthropic's Claude models, planning to train 10,000 workers over 18 months.
  • Old-school IT firms are racing to adopt AI before AI eats their human-heavy business model. Self-preservationcore.

💶 Publicis Boss Wants a €100B European AI Fund

  • Publicis chief Maurice Lévy is calling for a €100 billion European AI fund, urging France and Germany to lead - warning Europe leans too hard on US tech firms that could cut access anytime.
  • Europe just realized relying on foreign AI is wayyy risky. This. Is. Sovereignty. Anxiety.

📉 Shoppers Are SICK of "AI" Branding

  • A new poll says 60% of US shoppers hate "AI" branding in ads, and 86% don't fully trust AI answers - they still want to click through to real humans.
  • Slapping "AI" on everything can destroy trust and push customers away. Bestie, that's not the flex you think it is.

📉 Robinhood Cuts 10% of Staff - Without the AI Excuse

  • Robinhood cut 10% of its staff but, plot twist, the CEO refused to blame AI - citing a desire for a "flatter team structure" instead.
  • Using AI as a fake layoff excuse is falling out of style. People are getting tired of the hype. Refreshing??

🛠️ "Probably" Raises $9M to Kill AI Hallucinations

  • AI startup Probably raised $9 million to stop AI from making stuff up. Their tool checks answers against hard data for 99.99% accuracy.
  • Fix the fake-facts problem and companies can run smaller, cheaper models without losing trust. Probably a good idea (ha).

🛡️ The Real Way to Make AI "Defensible"

  • A new legal report says true AI defensibility comes from trade secrets, not patents or copyrights. Investors now demand strict AI clauses in deals so the tech can't be copied.
  • Just using AI won't save you. You need a legal wall around your tech to get funding. No moat, no money.