Many of my friends have also started using Notion AI, but don't quite know how to use it. I plan to use this thread to record my actual experience and cases of using Notion AI to improve efficiency in work and life, and update it from time to time 👉 #notionAI #chapGPT
1 / Reword emails or Slack messages. Written communication is crucial in the workplace. Beyond the specific message, it's also important to consider the wording and tone. I currently use Notion AI to help me refine my wording. For example, when asking for help from colleagues or adding options to a project, I might use "Friendly" or "Confident" or "Casual." To do this: Write a rough draft, then call the AI and select "Change Tone."
2 / Summarize a video or podcast into a timestamp-equipped shownote. To do this, you first need to convert the video or podcast audio into subtitles with timestamps (there are many tools available, I used iFlytek Tingting Subtitles), paste it into Notion, and let the AI summarize it into notes. Command: Summarize an outline of the video transcript, including timestamps (I tested it with a 10-minute video, and it only recognized 7 minutes).
3 / Continuing from the previous point, translating subtitles with a timeline is also very simple. Simply select the subtitles you want to translate, call out the AI, and choose to translate into Chinese or other languages. The translation quality is much better than some subtitle software's built-in translation. Of course, the same goes for translating subtitles from Chinese videos into English. (Although my Bear Academy translation turned into "熊猫学院", the other translations were basically fine and very good.)
4 / Streamline the text in my voice memos. I use voice recording software to capture my inspiration and thoughts, then use AI to organize them into coherent text and break them into paragraphs. Tip: Streamline the text into a coherent text, preserving as much of the original content as possible and breaking it into paragraphs for easier reading. Note: You'll still need to make some final edits, but it's already much easier.
5 / Output the text in the format of a Twitter thread. This is a usage I just thought of. It is particularly suitable for long articles that I have already written. I can directly cut it into a 140-character format suitable for Twitter. I just tried it with the previous example and it works very well. Command: Output in the format of a Twitter thread
6 / Translate content. Content creators can use Notion AI to translate Chinese posts into English, and vice versa. This allows them to maintain content in two languages and share it across different channels. To do this: Select the text, call out the AI, and select the language for translation. Note: I compared Notion AI's translations with ChatGPT's, and the difference isn't significant. For most people, either is fine. Both are significantly better than DeepL.
7 / I'm beginning to discover the power of Notion AI: it allows you to call upon your saved information and work on it. This is different from ChatGPT, where you still have to input, and is more valuable for creators. Next, I'll share a few of my test cases: Summarizing the content of my own video (podcast or article). This assumes that you have already converted the speech into text and saved it in Notion. I asked it to summarize my English video in Chinese.
8 / After I finish a book, I save all my excerpts and notes to a Notion document. I just tried it out. For a literary book, I asked Notion to select the five most compelling passages from my excerpts in "The Lost Satellite." It came up with a pretty good selection! This can also be used to apply to other situations.
9 / The real kicker is using AI for nonfiction books. For example, I read the English audiobook "Digital Minimalism" and took a few unorganized notes after finishing it. I selected the text and asked the AI to write a YouTube video script based on the selected content. Note that the notes are in English, but the script is in Chinese!
10 / Here comes something even more impressive: My original excerpt included an example from the original book about the Amish. I found it quite interesting and worthy of further discussion, so I asked the AI to help write a Chinese script for a video on the Amish theme. The result is this. It's quite valuable. However, it's important to note that human verification and proofreading are required, as AI's current nature can lead to fabricated content.
11 / Continuing with the previous theme: let the AI compare the lives and work styles of Amish people and programmers (or designers, or whatever your target audience is), and generate a table. This is incredibly useful for tech blogs comparing products. (All of these features are available in ChatGPT.)
12 / Write a review for a colleague. This falls under the category of expanded content, and it's a prime area for leveraging AI's language capabilities. You can write a few basic lines of description, then let Notion AI embellish it with some polish to create a coherent and accessible text. In the example above, I first casually wrote down some specific events related to this colleague, then let the AI organize it into coherent content, and then rewrite it into bullet points for easy review.
13 / Here's the big deal: How to use Notion AI to analyze YouTube videos, save the video knowledge as your own ddownsub.comedownsub.comtions (almost without any prompts!). 1. Find the video you want to save and copy the link. Here's a fitness video as an example. 2. Paste the link at https://t.co/Gx9nZc1ysE to download the video subtitles as a TXT file. Paste it into Notion in code format (mainly for ease of viewing, but you can also paste it directly as text). 3. Call up the AI and use the default options of Summarize and Find actions to summarize and identify actionable suggestions. 4. Use the default AI options again and click to translate directly into Chinese. This method works for almost any YouTube video. If it's in another language, download it as a txt file and use the AI to translate it first—it's just a click anyway. It's completely free (of course, Notion AI does cost money)!
14 / Generate a competitive analysis list. I use this to list product designer portfolios for easy review. Sentence: "I'm working on my portfolio as a product designer in New Sealand, working with large corps. Find me 5-10 portfolio examples to do competitor research." Honestly, this isn't very user-friendly; some of the websites in the generated list don't even open. I find traditional Google searches more efficient.
15 / Another example of text organization. This is the unorganized text (top) and the text organized using Notion AI (bottom). Because I've been using voice recording for years, the recognition rate is decent, but using AI to organize it is very convenient, requiring little adjustment. See the previous recommendation for the sentence.
16 / Writing rejection emails. This feature is incredibly useful. I usually struggle with the tone and wording when writing emails to reject requests, which can be quite time-consuming. Using AI to respond is incredibly helpful. For example, in this example, a book author contacted me about being a podcast guest. I wrote a short sentence and explained to the AI that I was considering my audience as a poor fit for their book's topic. The rejection letter AIx wrote was spot-on.
17 / Multiple windows solve the problem of processing long articles. Sometimes we need to process long articles, such as translating a long English article into Chinese. However, Notion AI, like other AI, has token restrictions and requires selecting paragraphs one by one, which is inefficient. My "old-fashioned" method is to open multiple browsers, open the same Notion article, and process a paragraph in each window. This can improve efficiency a bit :)
18 / The lightning-fast development of AI has caused anxiety for many, including me. This weekend, I decided to seriously reflect on the relationship between AI and my life, and how I can harness it rather than let it interfere with me. I started with something simple: I wrote down a rough outline of my daily schedule, organized my time, and listed which activities could potentially be influenced by AI. This analysis revealed that 30% of my time is potentially influenced by AI. Since AI can generally be involved in both my work and personal hobbies, 30% is a significant percentage. If I were a chef or construction worker, it would probably be only 10% at most. Next, I'll sort out my daily activities (such as recording podcasts and producing videos) and identify those that could be optimized by AI. Next, I'll sort out the tasks I do at work. Incidentally, I also used Notion AI to help me with this process. I wrote down the diary in the upper left corner, then used a prompt: "Organize this text into a table. The first column is the time, starting at 6 am and ending at 11 pm. The second column is the specific project, one line per sentence." After getting the table, I just adjusted the time slightly and let the AI calculate the time for each project. Super convenient. #NotionAI
19 / Continuing from the previous post, I started to organize my daily tasks, and then see which ones can use AI to help improve efficiency. The same operation as before. I first wrote down what I did outside of work, which is divided into six categories: learning, writing, drawing, video, podcast, and courses. Then list the specific tasks in each area. Let AI organize it into a table, and then add the current tools and tasks that AI is expected to replace. Two processes were used: 1. Organize the selected text into a table. The first column is the category, such as "writing" and "drawing", and the second column is the specific tasks. Write a separate line for each sentence before a period. 2. Add a column titled "AI" and list how AI can help complete each task in each task. Then get this table. It can be used as a reference for me in the future. I don’t know what specific AI tools can complete some tasks (such as automatic video editing), but with this table, I can add the corresponding details later. #NotionAI
20 / I completed the same process for my own work. The idea and method were the same, but this time I used English and made a slight adjustment: I had the AI add two columns to the generated table. One column automatically added a method for using AI to help me, and the second column added an example. And this time, it was amazing: the examples it provided included many tools I didn't know about or had heard of but hadn't used yet, such as OtterAI. This made my subsequent research much easier. This was my first time using AI extensively in Notion to create tables and organize categorized information. It was very practical, far more practical than ChatGPT. The lesson: Use different AI-based tools for different scenarios.
