InsightInBytes – June 26, 2025

From billion-dollar research pledges to courtroom wins and robotic breakthroughs — here’s what moved the AI world today.

Author

InsightInBytes Team
June 24, 2025

Good day, boundary-pushers.
From AI funding that fuels the next generation of researchers to court decisions redefining fair use, today's briefing is a sharp take on where AI meets power, law, and ambition.

Let’s decode today’s top 5.

🔹 BYTE 1: $100M Bet on AI for Good
Databricks & Perplexity co-founder Andy Konwinski just pledged $100M to fund non-commercial AI research via his new Laude Institute.

  • 🚀 $3M/year will anchor a new AI Systems Lab at UC Berkeley, led by Ion Stoica.

  • 🎯 Focus: Long-term, human-centric AI challenges — civic discourse, healthcare, and re-skilling.

  • 🏛️ Structured as both a nonprofit and a venture fund, with investments labeled “Slingshots & Moonshots.”

🧩 Finally, an AI research fund that doesn’t scream “exit strategy.”

🔹 BYTE 2: Grok Wants Your Spreadsheets
A leak reveals that xAI's Grok is building an AI-powered file editor — and it can soon help you edit spreadsheets via chat.

  • 🧪 Reverse-engineered code suggests Grok will be embedded in productivity tools.

  • ✍️ Earlier, Grok Studio launched split-screen workspaces for doc/code collaboration.

  • 🔄 Could this make X (formerly Twitter) your new Google Workspace?

👀 Elon Musk’s “everything app” vision gets a serious productivity twist.

🔹 BYTE 3: Fair Use Goes to Court — and AI Wins
A federal judge sided with Anthropic, ruling that training AI models on books without permission can fall under fair use.

  • 📚 The court acknowledged downloading pirated books was illegal… but training on them might be legal.

  • ⚖️ A separate trial will determine liability and damages related to pirated data.

🚨 This could set a massive precedent for future AI-vs-creators lawsuits.

🔹 BYTE 4: Europe’s Tech Sovereignty Alarm
European leaders are sounding the alarm after Microsoft disabled an ICC prosecutor’s email following US sanctions.

  • 🔄 Some EU officials are now migrating to ProtonMail and pushing for tech independence.

  • 💬 “The smoking gun,” one diplomat said, for why Europe needs digital sovereignty.

💡 Tech policy is no longer just local. It’s geopolitical.

🔹 BYTE 5: Robots Get Gemini-Powered Brains (Locally)
Google DeepMind launched Gemini Robotics On-Device, an LLM that runs offline and can control real robots.

  • 🤖 Used on robots like Franka FR3 and Apptronik’s Apollo to do real tasks: unzip bags, fold clothes.

  • 🧠 Supports natural language control, local fine-tuning, and generalization across different bots.

  • 🧪 Includes an SDK with simulator support for developers.

🔋 Robots with brains — no internet required.

Insight Drop of the Day
The future of AI isn’t just smarter models — it’s smarter humans deciding how (and why) we use them.

References