This article was originally written in Chinese and translated by Notion AI. The accuracy and details have not been confirmed by the author. For more accurate information, please refer to the original article.
Recently, I tried Notion AI and had two successful and one failed experience. The successful ones were summaries of a 10,000-word report that Notion completed at 80%. The failed application was when I tried to let Notion AI write an outline for an evaluation podcast of ChatGPT; the output was gibberish, so I didn’t hurt the audience.
Currently, Notion AI is more like a high-completion assistant writing tool, but its pure machine writing completion rate is still low. It excels at summarizing, compressing, and expanding, as well as writing connecting paragraphs. Writing something out of thin air is more difficult.
This made me think again about collecting all my personal information into Notion. I had already moved my book reading record to Notion before. Its advantage is a direct import of electronic reading notes and underlined sentences for subsequent excerpt retrieval, rather than a “private version of Douban”.
I’ve highlighted 20,000 words from a 400,000-word book while reading. Afterwards, I’ll import these highlights into Notion, creating a private excerpt library. When writing on any topic, I can quickly search for sources I thought were worth quoting in this separate database.
This is the first time I’ve experienced the concept of Evernote’s “second brain” firsthand.
Additionally, I’ve been using two tools to input and collect information over the past few years: Flomo (a single-machine version of Weibo) and Cubox (used to collect and highlight articles). The reason I use these two apps is mainly because they are better at receiving and collecting information.
Flomo starts quickly and uses less memory. It also has global shortcuts on Mac and can instantly record any ideas. The web clipping feature of Cubox and the forwarding and collection functions of WeChat public accounts can quickly save articles for later reading, classification, and annotation.
Using Flomo and Cubox, I’ve accumulated 472 memos and 405 articles over the past two and one year, respectively. However, I’ve found that, despite the detailed classification and labeling work I’ve done, the frequency of using these stored information is still very low.
I need an app like Notion, which can perform more complex operations such as adding, deleting, modifying and querying the knowledge I have accumulated.
Guess what? Both of these apps now support synchronization to Notion.
The two apps have a similar experience – when you link it to Notion, a record of a new memo or collected web page will be automatically updated in a specific database in Notion.
This allows me to search my entire knowledge base in Notion, and even perform more complex filtering, such as searching only in memos, only in collected articles, or both in articles and books. Of course, you can also filter by time, source, author, etc.
Now, I only use a small part of these two apps, while the entire data hosting and actual follow-up operations are done in Notion. Those familiar with the Internet industry will know this as “pipelining”, which means it’s close to being abandoned by users.
Notion was originally intended to replace Evernote or be the “next generation cloud notes”, but now it seems it never wanted to be a note from the start. As a network service, Notion fills a category that has been missing for a long time in the Internet age – a visual database for the end user.
Why is it said to be missing in the Internet age? Because in the previous Internet age, Microsoft Office Suite’s relatively obscure Access was present. Access gave users a full-featured database with a customizable interface and simple programming that allowed CRUD operations.
I remember my college computer class had an assignment to use Access to create a contact book tool. We had to enter contacts from a friendly interface, with mostly drag and drop to connect text boxes with corresponding fields – similar to Notion’s operation.
But Access had a high user threshold in the pre-Internet era. Notion greatly reduced the process of visual editing and simple no-code programming with blocks, reviving Access online. With Notion, you can easily make a CRM system, knowledge management system, GTD, Wiki – something most no-code platforms can’t do.
That’s why many domestic Notion copycats are always falling behind. Most copycats think “I make a note App, and then stuff the database in” while Notion actually works in the opposite way: “I make a database, initially presented in the form of notes, and then there can be countless forms of presentation”.
Without considering performance, the latter can meet all user needs and application scenarios. Because the essence of most Internet services is just different front-end displays of CRUD operations.
Notes, documents, disks, Twitter, Instagram, and Youtube are all the same–storing content in the cloud. This content is added, deleted, and modified. Searching, filtering, and switching between different channels involves querying the database. In essence, all internet applications are displaying a remote database in different ways to support “CRUD” in various contexts.
This means that Notion has the potential to replace the current web 2.0 giants. To illustrate this point, after the GDPR came into effect, all internet giants doing business in Europe launched Data Takeout services to meet the requirements of “data portability”. However, the data packages from companies like Google are in a huge compressed format, wrapping the user’s information in almost raw data, making it meaningless to the user themselves. Without data cleaning and programming training, these packages are not even useful as backups. Google itself does not even support restoring an account from these data packages.
When searching for a way to import content from Google to Notion, you’ll find existing solutions for Google documents, calendar, maps, and more.
It’s clear that you can’t download Facebook data to your local machine, but you can transfer it to various other platforms. The closest to personal possession is WordPress, though it has a slightly high threshold for those not familiar with computers. It’d be better if you could import it to Notion.
This scenario raises a concern: data is now centralized in Notion, where it was once spread across many Internet giants. This is a worry for many current Notion users. Unfortunately, Notion doesn’t provide any convenient data migration methods, making data security highly dependent on their single point ability.
Microsoft’s Access suite is a commercial product, but it is essentially a database management tool. The database generated by this tool is a local file, owned completely by the user. As long as you follow a specific protocol to connect to the Access database, you can continue using your own database without Microsoft.
So far, Notion does not support any private deployment methods, and there is no indication of this from the official source. This means that Notion, as a data expression layer protocol, is inseparable from its own database hosting service.
Personally, I believe it would be more beneficial to publish Notion as a data expression protocol than Notion AI. An example of this could be found in WordPress.com and WordPress.org, integrating features that are not available in single-point hosting such as Notion AI, such as high-performance search queries and better front-end interfaces. Additionally, an open source version should be available, allowing users to deploy privately, to facilitate an easy-to-use database self-hosting for individuals or small businesses.
If private deployment of Notion is possible, could I self-host all my personal data with a Synology?
A universal data expression protocol could lead to a more decentralized future than existing Web 3. This is because for many data, self-hosting is the only viable option for owning data autonomously.