Tag: google

  • Google DeepMind Just Put Brains in Boston Dynamics Robots

    Google DeepMind Just Put Brains in Boston Dynamics Robots

    Boston Dynamics is now running Gemini Robotics integrated into its Spot robot. That partnership, announced this week through Google Cloud, means the robots that navigate physical space and make real decisions are now powered by multimodal AI that understands what it sees and what to do about it. The Spot robot with Gemini is being…

  • Google’s New Flash Model Breaks Its Own Price War

    Google’s New Flash Model Breaks Its Own Price War

    Google just released three new Gemini models on July 21, and the one that matters most is Gemini 3.6 Flash. It’s the workhorse model Google is betting the company will use for 80 percent of the work people actually do. Here’s the thing: it got 17 percent faster while getting cheaper. Fewer output tokens. Lower…

  • Google’s Chip Just Got Very Specific

    Google’s Chip Just Got Very Specific

    Google is building a server chip called Frozen v2 that hardwires Gemini’s architecture directly into silicon. Not weights. Architecture. The blueprint itself becomes hardware. The payoff: 6 to 10 times the token throughput per watt compared to general-purpose TPUs. This is a bet on a settling architecture. Once your model design is baked into silicon,…

  • Brussels Just Forced Google’s Hand on Android

    Brussels Just Forced Google’s Hand on Android

    The European Commission just ordered Google to do something the company has been avoiding for a decade: open Android to competing AI assistants. This is not a suggestion. These are binding decisions. Google must share its search data with rival AI developers and let them run on two billion Android phones without barriers. What makes…

  • Google Just Missed Deadline Number Three

    Google Just Missed Deadline Number Three

    Gemini 3.5 Pro was supposed to ship in June. Then the deadline slipped to July 17. This week, Google confirmed it is still not ready. They scrapped the original base model and restarted pretraining from scratch. That is not a normal engineering decision. That is a sign the project is in real trouble, that what…

  • Google Just Fired Back With 2M-Token Gemini

    Google Just Fired Back With 2M-Token Gemini

    Google launched Gemini 3.5 Pro today, and it came loaded. The model carries a 2-million-token context window, Deep Think reasoning that works step-by-step through problems, and stronger autonomous coding than anything Google has shipped before. This is six weeks late. Google scrapped the original base model after finding recursive tool-calling failures and rewrote the whole…

  • Google Just Admitted It Cannot Build Fast Enough. Meta Is Scrambling.

    Google told Meta it cannot supply as much Gemini compute capacity as Meta wants. Meta had been using Gemini for content moderation at massive scale. Now Meta is shifting that work to its own Muse Spark model to reduce dependence on Google. This is the sound of a constraint becoming visible. Google Cloud quarterly revenue…

  • Google Photos Just Got a Video Remix Button

    Google rolled out Video Remix in Google Photos this week, powered by Gemini Omni, and it works the way video editing should have worked five years ago. You pick a video you shot. You tell Gemini what you want. Want the scene bathed in sunset light? Done. Want the background to be the Eiffel Tower…

  • Google Gemini 3.5 Pro Gets a Full Rebuild

    Google Gemini 3.5 Pro Gets a Full Rebuild

    Google just announced it is scrapping the Gemini 3.5 Pro it had planned and building something new instead. The launch moves to July 17, and the reason is worth paying attention to. The original 3.5 Pro was not hitting the target on mathematical reasoning, image generation quality, or SVG scene-building. Rather than patch it, Google…

  • Google’s Gemini 3.5 Pro Just Hit a Wall

    Google’s Gemini 3.5 Pro Just Hit a Wall

    Gemini 3.5 Pro missed its June launch and now targets July 17, but Google is not confirming that either. The target slips. Enterprise testers flagged serious gaps: token efficiency is not where it needs to be, coding performance fell short, and long-form reasoning does not match what Google promised at I/O in May. So Google…