Name: 在 10 分鐘內學會 80% 的 Perplexity！ (Learn 80% of Perplexity in under 10 minutes!)
Uploaded: 2025-06-10T12:38:08.000Z
Duration: 9 min 52 s

Here's a no-nonsense guide to using Perplexity.

以下是使用 Perplexity 的簡明指南。

情態助動詞 B1

First up, while many of you are familiar with ChatGPT and Google Gemini, you might not know where Perplexity fits in your AI toolkit.

首先，很多人都熟悉 ChatGPT 和 Google Gemini，但可能不知道 Perplexity 在人工智能工具包中的位置。

On one end, we have tools like ChatGPT and Gemini, designed for high brainpower, creative tasks.

在一端，我們有 ChatGPT 和雙子座這樣的工具，專為高腦力和創造性任務而設計。

On the other end are tools like Perplexity and Google Search, which excel at delivering accurate and real-time information.

另一方面，Perplexity 和谷歌搜索等工具擅長提供準確、實時的資訊。

And of course, we can't forget, all the way to the right, in a league of its own, we have the internet's undisputed single source of truth, Reddit.

當然，我們也不能忘記，在最右邊、獨樹一格的，是網路上無可爭議的真理唯一來源——Reddit

Put simply, for tasks like brainstorming, editing and writing, where you don't care about factual accuracy, ChatGPT and Gemini definitely deliver better outputs.

簡而言之，對於頭腦風暴、編輯和寫作等不注重事實準確性的任務，ChatGPT 和 Gemini 絕對能提供更好的產出。

But if you need a relatively high level of confidence in your answer, like deciding which supplements to take or troubleshooting software issues, Perplexity is the way to go.

但是，如果您需要對答案有相對較高的信心，比如決定服用哪種保健品或排除軟件故障，那麼 Perplexity 就是您的不二之選。

Diving right to the settings, I really like how Perplexity lets us exclude our data from their training process, while still allowing us to access our conversation history under the library tab.

直接進入設置，我非常喜歡 Perplexity 讓我們將自己的數據排除在訓練過程之外，同時還允許我們訪問庫選項卡下的對話歷史記錄。

I'll talk about the pro features at the very end, since this video is mainly on the free version, but first I need to debunk a myth other creators are saying.

由於本視頻主要介紹的是免費版，我將在最後談談專業版的功能，但首先我要揭穿其他創作者的一個神話。

Buying the pro version of Perplexity does not give you access to other paid AI tools like ChatGPT, even if you can select their models here.

購買 Perplexity 的專業版並不能使用 ChatGPT 等其他付費人工智能工具，即使您可以在這裡選擇它們的模型。

This is because even if they use the same underlying model, ChatGPT and Perplexity apply completely different fine-tuning methods.

這是因為即使它們使用相同的基礎模型，ChatGPT 和 Perplexity 採用的微調方法也完全不同。

For example, Perplexity optimizes for accuracy, speed, and a more search-oriented approach.

例如，Perplexity 優化了準確性、速度和更注重搜索的方法。

If you don't know what fine-tuning means, check out my video summarizing Google's Intro to AI course.

如果你不知道微調是什麼意思，可以看看我總結谷歌人工智能入門課程的視頻。

But the bottom line here is, even if you can select the same models, you do not get the same experience as the other AI tools.

但這裡的底線是，即使你可以選擇相同的模型，你也無法獲得與其他人工智能工具相同的體驗。

Next, moving on to the profile tab, this self-introduction section works just like ChatGPT's custom instructions.

接下來，進入 "個人資料 "選項卡，自我介紹部分與 ChatGPT 的自定義說明一樣。

As in the instructions you add here apply to all conversations within Perplexity.

您在這裡添加的說明適用於 Perplexity 中的所有對話。

For example, my instructions include structure your writing and respond like a consultant, adopt a conversational tone, use simple language, et cetera.

例如，我的指導包括：像顧問一樣安排寫作和回覆的結構，採用對話的語氣，使用簡單的語言，等等。

And since we only need to add this once to benefit forever, I highly recommend customizing this to fit your needs.

由於我們只需添加一次，就能永遠受益，是以我強烈建議您根據自己的需要進行定製。

If you're feeling a bit lazy, I'll leave a link down below to my custom instructions, and I'll also leave a link to my workspace toolkit if you wanna make a copy of my favorite prompts for productivity.

如果你覺得有點懶，我會在下面留下我的自定義說明的鏈接，如果你想複製我最喜歡的生產力提示，我也會留下我的工作空間工具包的鏈接。

Now let's actually start using Perplexity by going back to the Home tab, where you'll see the Focus feature.

現在，讓我們回到 "主頁 "選項卡，開始實際使用 Perplexity，你會看到 "聚焦 "功能。

This basically lets you narrow down the sources you want Perplexity to use, which can drastically improve the quality of your output.

從根本上說，這可以讓您縮小希望 Perplexity 使用的信號源範圍，從而大幅提高輸出品質。

For example, if I'm searching for real-time news, like Nvidia's most recent earnings call, I want to search the entire web.

例如，如果我要搜索實時新聞，比如 Nvidia 最近的財報電話會議，我希望搜索整個網絡。

For queries around fish oil supplements, I would focus on academic sources.

對於有關魚油補充劑的詢問，我會把重點放在學術資料上。

For gym workout recommendations, I might turn to social.

如果需要健身房鍛鍊建議，我可能會求助於社交網站。

And if I wanted to learn how to perform a specific exercise, I might want to watch some videos.

如果我想學習如何進行某種特定的練習，我可能會想看一些視頻。

The writing option is Perplexity's attempt at offering a more creative-oriented model.

寫作選項是 Perplexity 試圖提供的一種更注重創意的模式。

And as you can see, it generates text without searching the web.

正如你所看到的，它無需搜索網絡即可生成文本。

But again, in my own personal experience, I found the paid versions of ChatGPT and Google Gemini to be much better at creative writing.

但同樣，根據我的個人經驗，我發現付費版的 ChatGPT 和 Google Gemini 在創意寫作方面要好得多。

The Attach feature is a bit weird in that free users can only attach and upload PDFs, but not images.

附加功能有點奇怪，免費用戶只能附加和上傳 PDF，而不能上傳圖片。

At the same time though, we can just convert images to PDFs and upload those.

但同時，我們也可以將影像轉換成 PDF 文件並上傳。

所以，我想這是一個變通辦法。

Now let's go through a real-world use case.

現在，讓我們來看一個真實的使用案例。

How to force Wi-Fi connection after connecting to Marriott Wi-Fi on an iPhone.

如何在 iPhone 上連接萬豪酒店 Wi-Fi 後強制連接 Wi-Fi。

Since I chose to search the entire web, we can see that Perplexity drew from a variety of sources, from Apple's Help Center to Marriott's Help Center to Reddit posts.

由於我選擇的是全網搜索，是以我們可以看到 Perplexity 的資訊來源多種多樣，從蘋果公司的幫助中心到萬豪酒店的幫助中心，再到 Reddit 上的帖子。

I can click here to expand all sources, and I can even choose to remove sources that are outdated or unreliable.

我可以點擊這裡擴展所有來源，甚至可以選擇刪除過時或不可靠的來源。

Scrolling down, we see the true power of Perplexity.

向下滾動，我們就能看到 "困惑 "的真正力量。

Instead of clicking into multiple links on Google to find the right answer, we are literally given a set of instructions to follow.

我們不需要點擊谷歌上的多個鏈接來尋找正確答案，而是按照字面上的訓示來操作。

If you connect to a public network but don't see the login page, you can literally just type in example.com in your browser, and this forces the connection.

如果您連接到公共網絡但沒有看到登錄頁面，您可以在瀏覽器中直接輸入 example.com，這樣就可以強制連接了。

On the right is what I call supplementary information.

Perplexity correctly identified that it might be easier for me to watch a video to troubleshoot my issues, so it provided a few videos here.

Perplexity 正確地指出，觀看視頻可能會讓我更容易排除故障，是以它在這裡提供了一些視頻。

In other instances, it might search for images to help with my query.

在其他情況下，它可能會搜索圖片來幫助我進行查詢。

Here's an example directly comparing Perplexity to Google.

下面是一個直接比較 Perplexity 和 Google 的例子。

I searched for a country with the most gold medals in the most recent Olympics.

我搜索了在最近一屆奧運會上獲得金牌最多的國家。

I actually didn't know the answer going to this, by the way.

順便說一句，其實我也不知道答案。

I didn't watch the Olympics, except for the breakdancing part.

我沒看奧運會，只看了霹靂舞部分。

And on Google, if I click into the first link, it tells me the number of gold medals each country has won in the entire history of the Olympics, not from the most recent one, which is a little bit awkward.

在谷歌上，如果我點擊第一個鏈接，它就會告訴我每個國家在整個奧運會歷史上獲得的金牌數，而不是最近一屆奧運會的金牌數，這就有點尷尬了。

And even more awkward, I had to scroll all the way down to the fifth link here for me to find that, oh, in the most recent Olympics, the US and China tied in the number of gold medals.

更尷尬的是，我不得不一路向下滾動到這裡的第五個鏈接，才發現，哦，在最近一屆奧運會上，美國和中國的金牌數並列第一。

And that was a lot of clicking around and reading to find the information we wanted, right?

為了找到我們想要的資訊，我們需要不停地點擊和閱讀，對嗎？

When I type the same query into Perplexity, we actually get the correct answer right away.

當我在 Perplexity 中輸入同樣的查詢時，我們馬上就能得到正確的答案。

And Perplexity even includes a bit of trivia saying, this is the first time in Olympic history that two countries finished with the exact same number of gold medals in the summer games.

Perplexity 甚至還提供了一個小知識：這是奧運會歷史上第一次有兩個國家在夏季奧運會上獲得完全相同數量的金牌。

In Google's defense, I think an Apple to Apple comparison would have been to compare Perplexity to Google's AI overview feature, since those two are in the same category.

在谷歌的辯護中，我認為蘋果與蘋果之間的比較應該是將 Perplexity 與谷歌的人工智能概述功能進行比較，因為這兩者屬於同一類別。

Let me know in the comments if you wanna see a video on that.

如果你想看相關視頻，請在評論中告訴我。

Next up, let's talk about my favorite feature in Perplexity, collections.

接下來，讓我們來談談我最喜歡的 Perplexity 功能--收藏。

Diving right into an example, I'm actually visiting Japan later this year.

舉個例子，我今年下半年要去日本。

So I asked Perplexity when to visit if I want to avoid the peak tourist season.

於是我問 Perplexity，如果想避開旅遊旺季，什麼時候去？

Since I know I'll be doing more research over the next few weeks, I can click the three dots up here, add to collection, and click new collection, and then give this title, Trip to Japan 2, since I already have an existing collection from previously.

因為我知道接下來幾周我會做更多的研究，所以我可以點擊上面的三個點，添加到收藏，然後點擊新收藏，然後給這個標題取名為《日本之旅 2》，因為我已經有了之前的收藏。

And down here, I can add an AI prompt that will apply to the entire collection.

在這裡，我可以添加一個人工智能提示，它將適用於整個集合。

In this case, I tell Perplexity to assume the role of a seasoned travel agent with over 20 years of experience helping tourists uncover hidden gems in Japan.

在這種情況下，我讓 Perplexity 扮演一個經驗豐富的旅行社的角色，這個旅行社有 20 多年的經驗，可以幫助遊客發掘日本隱藏的寶藏。

When providing recommendations, prioritize local experiences that are tourist-friendly.

在提供建議時，應優先考慮方便遊客的當地體驗。

And now, whenever I start a new conversation or thread in this collection, Perplexity will remember to assume the role of a travel agent and give me answers tailored to my situation.

現在，每當我在這套書中開始新的對話或主題時，"困惑 "就會記得扮演旅行社的角色，根據我的情況給我答案。

Next, we have a slightly more advanced use case.

接下來，我們有一個稍微高級一點的使用案例。

As you can see, I've instructed this collection to generate a structured outline for our Tenselive presentation following the SCQA framework used by top consulting firms.

正如您所看到的，我已訓示該集合按照頂級諮詢公司使用的 SCQA 框架，為我們的 Tenselive 演示生成結構化大綱。

I'll leave a link to the full prompt down below.

我會在下面留下完整提示的鏈接。

Now, whenever I create a new thread in this collection, for example, gaming trends for the next few years and the impact on app developers, which is a topic related to my full-time job, by the way, Perplexity delivers an outline for a presentation I can start building on immediately.

現在，每當我在這個系列中創建一個新的主題時，例如，未來幾年的遊戲趨勢以及對應用程序開發人員的影響，順便說一句，這是與我的全職工作相關的一個主題，Perplexity 就會提供一個演講大綱，我可以立即開始構建。

And although this workflow can also be achieved using custom GBTs and Gemini gems, I'm much more confident in the numbers and data here since almost every single one of these bullet points have sources attached to them.

雖然這種工作流程也可以通過自定義 GBT 和雙子座寶石來實現，但我對這裡的數字和數據更有信心，因為幾乎每一個要點都附有數據來源。

And to wrap up, you can access all your collections by clicking into the library tab, and you can also make edits to the entire collection by clicking the three dots here and clicking edit collection.

最後，你可以點擊 "庫 "標籤訪問所有收藏集，也可以點擊這裡的三個點，然後點擊 "編輯收藏集 "對整個收藏集進行編輯。

到目前為止，我們只關注積極的方面。

So now, let's dive into an example where Perplexity doesn't perform very well compared to ChatGBT and Google Gemini.

現在，讓我們來看看 Perplexity 與 ChatGBT 和 Google Gemini 相比表現不佳的例子。

Here, I use the same prompt and ask ChatGBT and Perplexity to help me brainstorm thumbnail ideas for this video you're watching.

在這裡，我使用同樣的提示，請 ChatGBT 和 Perplexity 幫我集思廣益，為您正在觀看的這段視頻提供縮略圖創意。

I'm not gonna read the responses, but if you pause the video, you will see ChatGBT's ideas are much more creative and relevant.

我不想讀這些回覆，但如果你暫停視頻，就會發現 ChatGBT 的想法更有創意，也更貼切。

And I ended up creating my thumbnail based on one of these ideas.

最後，我根據其中一個想法制作了我的縮略圖。

Obviously, this is just one example, but I can constantly tell you that as of right now, you're much better off using ChatGBT or Google Gemini for high brain intensive creative tasks.

顯然，這只是一個例子，但我可以經常告訴你，就目前而言，你最好使用 ChatGBT 或谷歌雙子座來完成高腦力密集型創意任務。

And that's also another reason why Perplexity is competing more with Google AI overviews than Gemini or ChatGBT, since AI overviews also provides answers with a higher degree of accuracy, theoretically speaking.

這也是為什麼 Perplexity 與谷歌人工智能概述的競爭比 Gemini 或 ChatGBT 更激烈的另一個原因，因為從理論上講，人工智能概述也能提供準確度更高的答案。

Moving on, although I pay for Perplexity Pro, most users will be just fine with a free version.

繼續說下去，雖然我為 Perplexity Pro 付了錢，但大多數用戶使用免費版本就可以了。

As a pro user, you can first set your preferred text-to-text and text-to-image models, as mentioned before.

如前所述，作為專業用戶，您可以首先設置自己喜歡的文本到文本和文本到影像模式。

It is worth mentioning that Flux is a very popular image generation model right now, and generating images is also a pro only feature in Perplexity.

值得一提的是，Flux 是目前非常流行的影像生成模型，而生成影像也是 Perplexity 的一項專業功能。

Paid users can also use the pro search feature up to 600 times a day, as opposed to just five times.

付費用戶每天還可以使用專業搜索功能多達 600 次，而以前只能使用 5 次。

Pro searchers basically takes into account more sources and the answers are more detailed than what you would get from a regular quick search.

專業搜索器基本上會考慮更多的資訊來源，答案也比普通快速搜索得到的答案更詳細。

And that basically sums up the number of differences between pro and free.

這基本上概括了專業版和免費版之間的差異。

There are some Perplexity features I didn't touch on today simply because I don't use them.

今天我沒有談到 Perplexity 的一些功能，只是因為我不使用它們。

For example, if you click the plus icon here, you can create a page where Perplexity generates an entire blog post for you based on your initial prompt.

例如，點擊這裡的加號圖標，您就可以創建一個頁面，讓 Perplexity 根據您的初始提示為您生成整篇博文。

And if I wanted to, I could click here to publish the page, and this might show up in the Discover tab, which acts like a personalized newsfeed, but I'm just not sure I wanna be reading AI-generated articles all day, so.

如果我願意，可以點擊這裡發佈頁面，這可能會顯示在 "發現 "標籤中，就像一個個性化的新聞推送，但我不確定自己是否願意整天閱讀人工智能生成的文章，所以還是算了吧。

Finally, if you head on over to labs.perplexity.ai, this playground acts like a sandbox environment where developers can test out Perplexity's large language models before integrating them into their own applications.

最後，如果您前往 labs.perplexity.ai，這個遊戲場地就像一個沙盒環境，開發人員可以在將 Perplexity 的大型語言模型集成到自己的應用程序之前，在這裡對其進行測試。

If you found this video helpful, I have an entire playlist where I break down complex AI topics for beginners.

如果你覺得這段視頻對你有幫助，我有一個完整的播放列表，為初學者分解複雜的人工智能主題。

同時，祝你有一個愉快的假期。