Original video in Chinese.
Key Takeaway
- Gemini in Chrome transformation: native browser integration of Gemini turns the browser from a content display tool into a content understanding tool, with multimodal processing of web text and images, and Ctrl+G opens the sidebar.
- Add tabs as temporary RAG: add multi-tab context with one click (for example, 5 articles on the AI memory crisis), and Gemini analyzes and summarizes them together, efficiently handling webpages on the same topic.
- Help Me Write and the agentic future: right-click in input boxes for writing assistance, a context-aware thinking plug-in; the next step is the browser automatically carrying out actions (such as booking tickets), with practical productivity far beyond toy-level agents.
The Chrome you’ve used for more than a decade is finally going to get smart!
Google has officially launched Gemini in Chrome, fully integrating Gemini into the world’s most widely used browser.
What does that mean?
Chrome is changing from a “content display tool” into a “content understanding tool.”
Let me give you a typical example, and also a real situation from my past two days.
I opened 5 webpages, and these 5 articles were all about the AI memory crisis. I used the shortcut Ctrl plus G to open the Gemini sidebar. You’ll see that it has already loaded the current article as context.
But that’s still not enough. I want Gemini to handle all 5 articles together, because they’re on the same topic.
So, in the lower-left corner, I clicked the Add tabs button, added the other 4 articles as well, and asked Gemini what they were all about.
As I said in Knowledge Planet, this Add tabs feature is my favorite new feature at the moment. It’s basically a small, temporary RAG, and it’s extremely convenient. And it’s not just text; images on webpages can be handled too, because Gemini follows a native multimodal approach and has very strong image recognition.
Beyond the core multi-tab processing, there are two little details that make me feel Google really wants you to start using Gemini.
One is the always-available menu bar. When Chrome is not in the foreground, pressing Ctrl plus G will automatically pop up the Gemini dialog at the top. That’s so convenient.
I had also been wondering why Google still hadn’t released a Gemini desktop app. Now I see that it was integrated into Chrome instead, so there’s no need for a separate app.
The other is Help Me Write. In any input box in Chrome, right-clicking will bring up this feature popup. You can ask Gemini to generate text content for you.
These features may look plain and unremarkable, but put together, they really can improve day-to-day efficiency.
And for the industry as a whole, I think this is a key step in the evolution of browser form.
In the past, the core of a browser was the Rendering Engine — turning code into a visual display. Now, the browser is being implanted with something new: the Inference Engine — letting it think. This shift completely changes the logic of how people interact with the internet:
In the past, the browser only had to faithfully “translate” code into pixels. It didn’t care whether what was shown on the screen was a paper or a cat picture — it was just a pipeline.
Now, the browser can “understand” the semantics of the page. It takes over the first round of information filtering for your brain. It is helping you “actively digest” things.
In the past, the browser provided an input box, and you had to figure out every word yourself and type it in.
Now, through “Help Me Write,” the browser understands your context. It begins to participate in the creative process and becomes a thinking plug-in for you.
You see, that’s what I mean by “Chrome changing from a ‘content display tool’ into a ‘content understanding tool.’” And this still isn’t the end point.
Chrome can already “understand” webpage content, so the next step is for it to operate on your behalf. That is exactly the Agentic Web that Google is exploring. The logic is very simple:
Since the browser already understands that “this is a ticket-booking website,” and also knows that “you want to book tickets for tomorrow,” then naturally it should just help you complete the whole process, shouldn’t it?
I know everyone has recently been drawn to the big buzz around Clawdbot. But honestly, before the privacy and security issues are solved, Clawdbot is just a toy. AI version Chrome, on the other hand, is something that can genuinely boost your productivity. Remember to try it. I posted the method in Knowledge Planet, so you can go take a look.
OK, that’s all for this episode. If you want to learn about AI, want to become a super individual, and want to find like-minded people, come join our newtype community. See you next time!