Original video in Chinese.
Key Takeaway
- AI system for ordinary people: no need for a complicated Claude Code framework; the Google suite (the $20 Google One subscription) is the best solution; low barrier to entry (direct payment with a China Merchants Bank card, no account bans), high value for money (Gemini 3 Pro/Flash + Nano Banana Pro + Veo 3 + NotebookLM).
- Two usage layers: learning layer (NotebookLM: upload documents/YouTube/webpages/photos of blackboard notes/your own notes, generate podcasts/summaries/flashcards/PPTs, output forces input); creation layer (Ask Gemini embedded in Docs/Gmail/Slides, direct contextual conversation/insert content).
- Core advantages: seamless integration with the Google ecosystem (Docs/Gmail/Slides), work without leaving the tools; strongest models (text/image/video/learning), repeatedly recommended by the author for a year, the first choice for maximizing value for ordinary people.
For ordinary people, there’s really no need to go to the trouble of building such a complicated AI system. For example, like me, building it on the Claude Code framework.
Let’s be honest: I myself think the cost is a bit high. And it’s limited to desktop use. I only open VS Code and start Claude Code when I’ve sat down and want to do deep thinking or create something.
Most of the time, I just use Gemini. Or more accurately, I use the various tools within the Google ecosystem.
I’ve been recommending Gemini for a year. Now, I’m officially recommending the Google suite. I believe that for most ordinary people, putting together this Google setup is the most suitable AI system for you.
First, the barrier to entry is the lowest.
Among all foreign AI tools, Google is the most friendly to Chinese users like us. I’ve had four accounts banned on Anthropic’s side. On Google’s side, there’s absolutely no problem.
Look, I just use a China Merchants Bank MasterCard to pay $20 a month. There’s no need to mess around with overseas cards or find a proxy payment service.
Second, it offers the best value for money.
For the same $20, paying Google is absolutely the most cost-effective. Let me break it down for you:
Gemini 3 Pro and Flash, Nano Banana Pro, and Veo 3—these models can meet your needs for generating text, images, and video. They’re all the best right now.
If you want to learn, you can use NotebookLM.
If you want to make PPTs, you can use Google Slides.
If you want to build a knowledge base, just put the documents into Google Docs.
If you want to program, use Antigravity.
And Google still has more things coming soon, such as Chrome with upgraded AI features.
You can compare across the board. For just $20, which company can give you so many good things?
So the question is: Google’s suite of tools really has AI added to everything, and everything is connected. But how do you combine them and use them together to maximize their value? In this video, I’ll share my logic and experience.
Hello everyone, welcome back to my channel. Modestly speaking, I’m one of the few bloggers in China who can explain the Why and How of AI clearly. What I offer is far more valuable than tutorials. Remember to hit follow. If you want to connect with me, come join our newtype community. This community has been running for 600 days, and more than 1,900 friends have paid to join.
If you’re a user in China, you can join through Knowledge Planet. If you’re overseas, you can join through Substack. My first course, daily newsletter, and exclusive videos can all be seen inside the community.
Back to today’s topic: how to use the Google suite.
I divide the Google ecosystem, which I connect through Gemini, into two layers.
The first layer is the data layer. Its role is a “long-term memory repository.” The carrier is Google Drive.
Right now, there are three products in the Google ecosystem that support document uploads: Gemini, NotebookLM, and Google Drive. So why choose Google Drive as the long-term memory repository?
I’ve talked about this in the community.
Gemini only allows up to 10 documents to be uploaded.
NotebookLM allows more—the premium version supports up to 300 per notebook, while the free version supports up to 50.
Google Drive has the highest limit. You can upload hundreds of documents, and you can also create multi-level folders for organization.
Besides the difference in quantity, the Gemini model settings in these three products are not quite the same.
NotebookLM is the strictest. Because its primary task is to “not talk nonsense,” it must and can only find answers from the documents. If the answer isn’t written in the documents, even if it “knows” the answer, it will probably choose to reply, “I can’t find relevant information in the documents.”
Gemini in Drive is more lenient. It has Source-Grounding, but the restrictions are not as strict. After reading your documents, it will answer by synthesizing them.
I think the model in the Gemini app is the most permissive of all. You ask it a question, and it will combine its existing knowledge, information found online, and the input you provide to give the best answer.
So, taking the model presets and document limitations together, I recommend using Google Docs as the long-term memory repository.
The second layer is the interaction layer.
If you need AI to interact with you based on global data, choose the Gemini app.
Type @ in Gemini’s chat box, and you’ll see that many Google products are already connected to Gemini.
For example, I can @Gmail and have Gemini help me find past email exchanges with a certain person. I can also @Google Drive and have it help me look up documents. Or after typing @, I don’t even have to ask it to look things up—I can just ask a specific question directly. Gemini will automatically find the relevant background information inside and complete the answer.
That’s what I mean by “global interaction.” In the Gemini app, you can interact with your private-domain data—just like I demonstrated a moment ago—or interact with public-domain data, and Gemini will search the web.
If your need is for deep reading and knowledge internalization, then use NotebookLM.
NotebookLM is a closed and rigorous sandbox. Its feature is that it “doesn’t hallucinate” (Grounded). So it’s very suitable for handling complex, domain-specific knowledge bases, such as financial reports, papers, and so on.
Also, the mobile version of NotebookLM now supports photo uploads. That means when you’re reading a book, you can photograph a few pages and upload them; when you’re in class, you can photograph the teacher’s blackboard notes and upload them; after you finish writing notes, you can also just take photos and upload them directly. All of these materials will be used as references for NotebookLM’s answers.
If you need to communicate with AI while writing something, I recommend using Ask Gemini.
In Docs or Gmail, click Ask Gemini in the upper right corner, and a sidebar will open, showing the familiar chat box.
In this mode, Gemini is embedded in the productivity tool. You don’t need to leave the document or email to chat; instead, you let it “work” directly at the cursor position. By default, Gemini uses your document or email content as context, so the interaction is very convenient.
In addition, there is usually an Insert button below Gemini’s response. You don’t need to copy and paste; you can directly add the AI-generated paragraph into the document.
If you fully understand the logic behind these two layers, then you’ll have a fairly deep understanding of the AI version of the Google suite. Not only will you know how to maximize the value of the $20 subscription fee, you’ll also know why I repeatedly recommend Google and Gemini over the course of a year. It is definitely the best solution for most ordinary people.
OK, that’s it for this episode. If you want to understand AI, want to become a super individual, and want to find like-minded people, come join our newtype community. See you next time!