A forensic post hit HackerNews #1 last week with 695 points: Chrome silently downloaded a 4GB Gemini Nano weight file in 14 minutes 28 seconds. No prompt, no notification, no opt-in. The file sits at OptGuideOnDeviceModel/weights.bin. Delete it and Chrome pulls it down again. Stopping it requires chrome://flags or enterprise policy — nothing in normal settings.
What actually got installed
Gemini Nano is Google’s on-device LLM, the model that powers Chrome’s Built-in AI APIs and tab summarization. Until now it shipped behind dev flags. The forensic shows it’s rolling out on stable Chrome too, quietly, to machines whose users never asked for a local LLM. Chrome’s global share sits around 64%, so we’re talking roughly 3.8 billion devices.
Why this is the bigger story
This is the largest on-device AI model deployment in history. Whatever Apple, Microsoft, and Meta do next, they’re looking at the same playbook: ship the weights, skip the dialog, take the heat afterward. The fight isn’t whether on-device AI is useful. It’s who gets to decide an LLM is running on your machine — and Google just made that call for a third of the planet.
You Might Also Like
- Google Lyria 3 Just Turned Gemini Into a Music Studio and im Weirdly Into it
- Google Nano Banana 2 Just Dropped and its Absurdly Fast
- Gemini Canvas in ai Mode Google Just Turned Search Into a Creative Workspace
- Google Releases Gemini Embedding 2 one Vector Space for Text Images Video and Audio
- Google is Deploying Gemini ai Agents to 3 Million Pentagon Employees Heres What That Means
Top AI Product 
Leave a comment