Last Updated on 2023-10-31

人工智慧那麼厲害，為什麼沒有滿街都是？未來應用在哪裡？生成式AI很紅，為什麼吳恩達認為監督式學習還是重點？市場上有很多AI新創，吳恩達看好什麼類型的企業？本文整理吳恩達演講逐字稿，帶你一次看。

人工智慧領域大師吳恩達（Andrew Ng），在2023年7月26日於美國加州史丹佛大學，以「人工智慧的機會」為題，分享他對人工智慧應用的觀察。吳恩達認為，人工智慧是通用技術，但現在應用聚焦在網路產業，因此未來機會在於其他領域的長尾應用；然而，其他產業即使想運用AI，也沒有辦法花大錢雇用工程師，幸好出現低／無程式碼工具，讓長尾應用得以落地。最後，吳恩達表示他看好AI專家與領域專家的合作，強調與領域專家落實「具體想法」的重要性。

原始演講影片如下：Andrew Ng: Opportunities in AI – 2023 – YouTube

重點歸納

• 背景
  • 生成式AI興起，ChatGPT引發熱潮
  • 在生成式AI以前，人工智慧領域以監督式學習為大宗
  • 人工智慧是通用技術，就像電力一樣，能夠解決不同問題，例如分類人們對餐廳的評價、辨識圖像中的動物類別、判斷郵件是不是垃圾信
• 重點一
  • 生成式AI興起，讓開發者甚至一般人能在短時間應用AI，未來成長可期
  • 就產值來說，監督式學習仍然是未來3年的發展重點，無論是產值或成長都不容小覷
• 重點二
  • 增加產值的方向就是尋找AI技術落地的領域，現在應用聚焦在網路產業，其他產業案例較少
  • 這些產業的客戶和專案規模遠遠低於Google、Facebook等企業，沒有雇用大量工程師的資源
  • 出現低／無程式碼工具之後，客製化變得便宜，長尾應用得以落地
• 重點三
  • 拆解AI應用金字塔，只有應用發展好，技術才會好
  • 找能夠具體落地的想法、找領域專家合作
• 收尾
  • AI不夠好，但我們努力解決問題
  • 機器仍比不上人類，即使再進步，我們仍能駕馭它

延伸閱讀：[資訊整理] 深度學習大師吳恩達 Andrew Ng

開場：AI是通用技術，能夠解決多元問題

It’s good to see everyone. So what I want to do today is chat to you about some opportunities in AI. So I’ve been saying AI is the new electricity.  One of the difficult things to understand about AI is that it is a general purpose technology,  meaning that it’s not useful only for one thing, but it’s useful for lots of different  applications, kind of like electricity. If I was to ask you, what is electricity good for? You know, it’s not any one thing, it’s a lot of things. So what I’d like to do is start off sharing with you how I view the technology landscape,  and this will lead into the set of opportunities. So a lot of hype, a lot of excitement about AI. And I think a good way to think about AI is as a collection of tools.

很高興見到大家。我今天想和你們討論AI的機會。我總是在說AI就是新的電力，它是一種通用技術，意味著不僅對一件事有用，而是有著許多應用，宛若電力一般。如果我問你，電力有什麼用途？你知道，它不僅僅是一件事，而是很多事情。我想先和大家分享我是如何看待技術格局，這將引出一系列的機會。現在AI有很多炒作、很多興奮。我認為一個好的方式是，將AI看作是一整套工具。

So this includes a technique called supervised learning, which is very good at recognizing things or labeling things, and generative AI, which is a relatively new, exciting development.  If you’re familiar with AI, you may have heard of other tools, but I’m going to talk less about these additional tools, and I’ll focus today on what I think are currently the two most important tools, which are supervised learning and generative AI. So supervised learning is very good at labeling things or very good at computing input to  output or A to B mappings. Give me an input A, give me an output B. For example, given an email, we can use supervised learning to label it as spam or not spam.  The most lucrative application of this that I’ve ever worked on is probably online advertising, where given an ad, we can label if a user is likely to click on it and therefore show more relevant ads. 

這包括一種稱為「監督式學習」的技術，它非常擅長識別事物或標記事物；當然也包括生成型AI，這是一種相對新的、令人興奮的發展。如果你對AI有所了解，你可能聽說過其他工具，但我不會著墨其他，而是專注於我認為目前最重要的兩個工具：監督式學習和生成型AI。監督式學習非常擅長標記事物，也就是計算輸入（input）到輸出（output）的映射（mapping）。給我一個輸入A（一封信）、輸出B（垃圾郵件的標籤），就能判斷類別。我曾經參與過的最賺錢的應用可能就是線上廣告，只要拿一則廣告出來，我們就能標記用戶是否可能點擊它，藉此顯示更相關的廣告。

For self-driving cars, given the sense of readings of a car, we can label it with where are the other cars. One project that my team at AIFM worked on was ship route optimization, where given a route that a ship is taking or considering taking, we can label that with how much fuel we think those will consume and use this to make ships more fuel efficient. Did a lot of work in automated visual inspection in factories, so you can take a picture of a smartphone that was just manufactured and label is there a scratch or any other defect  in it.  Or if you want to build a restaurant review reputation monitoring system, you can have  a little piece of software that looks at online restaurant reviews and labels that as positive  or negative sentiment. So one nice thing, one cool thing about supervised learning is that it’s not useful for one thing, it’s useful for all of these different applications and many more besides.

對於自動駕駛汽車來說，根據車輛的感測數據，我們可以標記出其他汽車在哪裡。我的團隊在AIFM進行的一個專案是改進船舶路線，即根據一條船正在採取或考慮採取的路線，我們可以標記出我們認為這會消耗多少燃料，並用這份資訊提高船舶的燃料效率。我們還在工廠進行了大量的自動視覺檢查工作，你可以拍一張剛剛生產出來的智慧型手機的照片，標記是否有刮痕或其他瑕疵。若你想建立一個餐廳評價監控系統，你可以建立一個簡單的軟體，判斷餐廳評價為正面或者負面。從上面的例子可以看出，監督式學習不僅對一件事有用，它有著多元的應用。

Let me just walk through concretely the workflow of one example of a supervised learning labeling  things kind of project. If you want to build a system to label restaurant reviews, you then collect a few data points, a collective data set, where say the pastrami sandwich is great, say that is positive, the  servos are slow, that’s negative, my favorite chicken curry is positive.  And here I’ve shown three data points, but you’re building this, you may get thousands of data points like this or thousands of training examples, we call it. And the workflow of a machine learning project, of an AI project is you get labeled data,  maybe thousands of data points, then you have an AI engineering team train an AI model to  learn from this data. And then finally, you would find maybe a cloud service to run the trained AI model. And they can feed it, you know, best quality I’ve ever had, and that’s positive sentiment.

讓我具體帶大家走過一次監督學習標記專案的例子。如果你想建立餐廳評論標記系統，你需要收集資料，例如「三明治很好吃」是正面評論，「伺服器很慢」則是負面，「我最喜歡的雞肉咖哩」又是正面。我展示了三筆資料，當你做這個機器學習專案，建立起分辨評論的系統時，你可能會得到像這樣的數千筆資料，或者說訓練範例（training sample）。你有一個AI工程團隊，從這些資料中學習，最後訓練出一個AI模型，最後，有雲端服務來運行模型。你可以準備很高品質的資料給模型。

And so I think the last decade was maybe the decade of large scale supervised learning. What we found starting about 10, 15 years ago, was if you were to train a small AI model, so train a small neural network or small deep learning algorithm, basically a small AI model,  maybe not on a very powerful computer, then as you fed it more data, its performance would  get better for a little bit, but then it would flatten out, it would plateau and it would stop being able to use the data to get better and better.

我認為過去的十年可能是大規模監督式學習的十年。大約在10、15年前，當時的電腦性能沒那麼好，如果你訓練一個小型的AI模型，當你提供更多資料，模型表現會稍微提升，但之後就會趨於平緩，達到一個高原後，增加資料的量也沒辦法再進步。

But if you were to train a very large AI model, lots of compute on maybe powerful GPUs, then  as we scaled up the amount of data we gave the machine learning model, its performance  would kind of keep on getting better and better. So this is why when I started and led the Google Brain team, the primary mission that I directed the team to solve at the time was let’s just build really, really large neural  networks that we then fed a lot of data to, and that recipe fortunately worked. And I think the idea of driving large compute and large scale data, that recipes really helped us, driven a lot of AI progress over the last decade. 

但如果你要訓練一個非常大的AI模型，已經使用大量的計算能力，或許是強大的GPU，當我們增加資料量的時候，表現將會持續不斷地提升。這就是為什麼當我開始並領導Google Brain團隊時，主要的任務就是這樣。團隊當時只是建立非常非常大的神經網絡，將大量的數據餵給它，幸運的是，這個方法奏效了。我認為更大規模的計算、更大規模的資料，這個辦法很有用，在過去10年中推動許多AI的進步。

<So if that was the last decade of AI, I think this decade is turning out to be also doing everything we had in supervised learning, but adding to it the exciting two of generative AI. So many of you, maybe all of you, were played with chatGPT and bot and so on,  but just given a piece of text, which you call a prompt, like I love eating, if you run this multiple times, maybe you get bagels, cream cheese, or my mother’s meatloaf, or all to friends, and the AI system can generate output like that.

轉折：生成式AI加速開發AI系統流程

因此，過去10年是監督式學習的時代，未來10年也會如此，不過，還要加入令人興奮的生成AI這個工具。你們許多人，都曾用過ChatGPT和，只要給出一段提示（prompt），比如「我喜歡吃東西」，如果你多次運行這段文字，你可能會得到「貝果」、「奶油乳酪」，或者「我母親的肉餅」，或者「所有的朋友」，AI系統可以生成這些輸出。

Given the amounts of buzz and excitement about generative AI, I thought I’d take just half a  slide to say a little bit about how this works.  So it turns out that generative AI, at least this type of text generation, the core of it is using supervised learning that inputs output mappings to repeatedly predict the next word. And so if your system reads on the internet a sentence like, my favorite food is a bagel  with cream cheese and lox, then this is translated into a few data points where if it sees my favorite food is a, in this case, try to guess that the right next word was bagel, or my favorite food is a bagel, try to guess next word is with,  and similarly, if it sees that, in this case, the right guess for the next word would have been cream.  So by taking text that you find on the internet or other sources, and by using this input output, supervised learning to try to repeatedly predict the next word, if you train a very large AI system on hundreds of billions of words, or in the case of the largest model is now more than a trillion words, then you get a large language model like ChatGPT. 

考慮到大家對生成型AI的熱烈討論和興奮，我想花半張幻燈片來稍微講解一下它的運作方式。生成型AI，至少這種文本生成的類型，核心是使用監督式學習來輸入輸出映射，以反覆預測下一個單詞。若你的系統在網路上讀到像是「我最喜歡的食物是奶油乳酪和煙燻鮭魚貝果」這樣的句子，那麼這句話就會被翻譯成幾個資料點，如果它看到「我最喜歡的食物是」，在這種情況下，試著猜測下一個正確的單詞是「貝果」。或者，我的最愛是貝果，試著猜下一個單詞是「與」，同樣地，如果看到這樣的情況，下一個正確的單詞應該是「奶油」。因此，透過在網路或其他來源找到的文字，並使用這種輸入輸出的監督學習來反覆預測下一個單詞。如果你在數百億甚至是最大模型現在超過一兆的文字上訓練一個非常大的AI系統，那麼你就會得到一個像ChatGP這樣的大型語言模型。

And, you know, there are additional other important technical details. I talked about predicting the next word. Technically, these systems predict the next sub word, a part of  word called a token. And then there are other techniques like RLHF for further tuning the AI output to be more helpful, honest, and harmless. But at the heart of it is this using supervised learning to repeatedly predict the next word that that’s really what’s enabling the exciting,  really fantastic progress on large language models. So, while many people have seen large language models as a fantastic consumer tool, you can go to a website, like ChatGPT’s website or BARD or other large language models and use it as a  fantastic tool. There’s one other trend I think is still underappreciated, which is the power of large language models, not just as a consumer tool, but as a developer tool. 

還有其他重要的技術細節。我談到了預測下一個單詞。從技術上講，這些系統預測下一個「子詞」（subword），也就是被稱為符號、詞彙的一部分。還有其他技術，如RLHF，用於進一步調整AI輸出，使其更有幫助、更誠實、更無害。但其核心是使用監督式學習，學習反覆預測下一個單詞，這真的是推動大型語言模型取得令人興奮、真正驚人進步的關鍵。許多人將大型語言模型視為一個極好的消費者工具，你可以使用ChatGPT或是Google Bard，或者其他大型語言模型，但我認為還有一個被低估的趨勢，大型語言模型不只能面向消費者，也能作為開發者工具。

So, it turns out that there are applications that used to take me months to build that a lot of people can now build  much faster by using a large language model. So, specifically, the workflow for supervised learning, building the restaurant review system, say, would be that you need to get a bunch of labeled data. And maybe that takes a month to get a few thousand data points. And then have an AI team train and tune and really get optimized performance on your AI model. Maybe that’ll take three months. Then find a cloud service to run it, make sure it’s running robustly, make sure it’s recognized. Maybe that’ll take another three months. So, a pretty realistic  timeline for building a commercial-grade machine learning system is like six to 12 months. So, teams I’ve led will often took, you know, roughly six to 12 months to build and deploy these systems. And some of them turned out to be really valuable. 

過去，有些應用程式需要花費數個月的時間打造，現在更多人能利用大型語言模型快速建立應用。以建立餐廳評論系統為例，你需要獲得一堆標記好的資料，可能要花一個月的時間取得資料，然後讓AI團隊進行訓練和調整，並優化你的AI模型，這可能需要三個月的時間，接著再利用雲服務來運行，確保它的運行穩健，確保它被認可，也許這還需要再過三個月。所以，建立一個商業級別的機器學習系統，實際時間表大約是六到十二個月，這些系統中，有些是非常有價值的。

But this is a realistic timeline for building and  deploying a commercial-grade AI system. In contrast with prompt-based AI, where you write a prompt, this is what the workflow looks like. You can specify a prompt that takes maybe minutes or hours. And then you can deploy it to the cloud. And that takes maybe hours or days. So, they’re  certain AI applications that used to take me, you know, literally six months, maybe a year to build, that many teams around the world can now build in maybe a week. And I think this is already starting, but the best is still yet to come. This is starting to open up a flood of a lot more AI  applications that can be built by a lot of people. So, I think many people still underestimate the  magnitude of the flood of custom AI applications that I think is going to come down the pipe. 

這是建立和部署一個商業級別的AI系統的實際時間表。我們接著拿「以提示為基礎的AI」比較，寫提示（prompt）就是你的工作流程，有可能要花幾分鐘或者幾小時，接著將它部署到雲端，這可能再花幾小時。過去，有些AI應用程式可能需要我花上六個月、甚至一年的時間來建立，現在許多團隊可能一週就能完成。我認為這件事已經在發生，但最好的時光尚未來臨。這開始引發更多AI應用的洪流，許多人都能建立AI應用。因此，我認為許多人仍然低估即將來臨客製化AI應用的規模。

Now, I know you probably were not expecting me to write code in this presentation. But that’s what I’m going to do. 

我知道你可能沒有預期我會現場寫程式碼，但這正是我接下來要做的事情。

So, it turns out, this is all the code I need in order to write a sentiment  classifier. So, I’m going to, you know, some of you will know Python, I guess, import some tools from OpenAI. And then I have this prompt that says, classify the text below, delimited by three dashes, as having either a positive or negative sentiment. I don’t know. I had a fantastic time at Stanford GSB. I learned a lot and also made great new friends. All right. So, that’s my prompt. And now I’m just going to run it. And I’ve never run it before. So, I really hope,  thank goodness, I got the right answer. And this is literally all the code it takes to build a sentiment classifier. 

結果顯示，我只需要這幾行程式碼，就能完成一個情感分類器。利用Python，從OpenAI導入一些工具，接著利用提示詞，說明要將下面的文字進行分類。「我不知道」、「我在史丹福大學商學院度過了一段美好的時光」、「我學到了很多，也交到了很棒的新朋友」，這些是我的資料，圖上則能看到我的提示詞，接著我們來要運行它，我先前沒有運行過，我真的希望（能成功）。感謝上帝，我得到了正確的答案。這就是建立情感分類器所需的所有程式碼。

論述一：生成式AI很棒，但監督式學習的產值和成長動能還是很大

And so, today, you know, developers around the world can take literally maybe like 10 minutes to build a system like this. And that’s a very exciting development. So, one of the things I’ve been working on was trying to teach, you know, online classes about how to use prompting, not just as a consumer tool,  but as a developer tool. So, to start off the technology landscape, let me now share my thoughts on what are some of the AI opportunities I see. This shows what I think is the value of different AI technologies today. And I’ll talk about three years from now. But the vast majority  of financial value from AI today is, I think, supervised learning, where for a single company like Google can be worth more than 100 billion US dollars a year. And also, there are millions of developers building supervised learning applications. So, it’s already massively  valuable and also with tremendous momentum behind it, just because of the sheer efforts in, you know, finding applications and building applications. 

現在全世界的開發者，可能只需要花10分鐘就能建立像這樣的系統，這樣的發展令人興奮。因此，我一直努力教導如何使用提示，不僅是作為消費者工具，而是作為開發者工具。所以，從科技景觀開始，讓我分享我看到的AI機會，以及我辨認出的價值。我討論的時間是從現在到未來三年，我認為，現今AI的絕大部分金融價值來自於監督式學習，像Google這樣每年可能營收超過1000億美元，還有數百萬的開發者，正在建立以監督式學習為基礎的應用，它已經具有巨大的價值，背後也有著巨大推力。

And in generative AI is the really exciting new entrant, which is much smaller right now. And then there are the other tools I’m including for completeness. We can, you know, the size of these circles represent the value today. This is what I think I might grow to in three years. So, supervised learning already really massive, may double, say, in the next three years from truly massive to even more massive.  And generative AI, which is much smaller today, I think will much more than double in the next three years because of the number of amounts of developer interest, the amount of venture capital investments, the number of large corporates exploring applications. And I also just want to  point out, three years is a very short time horizon. If it continues to compound at anything near this rate, then in six years, you know, it’ll be even vastly larger. But this light shaded region in green or orange, that light shaded region is where the opportunities for either new startups  or for large companies, incumbents to create and to enjoy value capture.

而生成式AI是真正令人興奮的新事物，現在規模還很小。圖片中的圓圈大小代表了今日價值，（外圈）是我認為在三年內將會成長到的規模。從圖片來看，監督式學習已經很大，可能在3年內翻倍，從真正的龐大變得更加龐大。而生成式AI現在規模較小，在接下來的3年內將會翻倍，因為開發者的興趣、風險投資者的關注、大型企業也在探索應用。當然，3年是一個非常短的時間範圍。如果它繼續以接近這個速度成長，那麼在6年內它將會變得更加龐大。

方向一：想應用AI，重點還是在找合適的領域

But one thing I hope you take away from this slide is that all of these technologies are general purpose technologies. So, in the case of supervised learning, a lot of the work that had to be done over the last decade, but is continuing for the next decade, is to identify and to execute on the concrete use cases. And that process is also kicking off for generative AI. So, for this part of the presentation, I hope you take away from it that general purpose technologies are useful for many different  tasks. A lot of value remains to be created using supervised learning. And even though we’re nowhere near finishing figuring out the exciting use cases of supervised learning, we have this other, you know, new technology with generative AI, which further expands the set of things we can now do using AI.

我希望你從這張簡報帶走的一個重點是，所有這些技術都是通用技術。過去10年和接下來的10年，發展監督式學習的重點就是辨識然後執行能夠具體落地的應用方案，生成式AI同樣適用這個過程。監督式學習仍有許多價值等待實現，而且我們還沒找出所有應用場景，與此同時，生成式AI進一步擴大AI能處理任務的範圍。

But one caveat, which is that there will be short term fads along the way. So, I don’t know if some of you might remember the app called Lenza. This is the app that would let you upload pictures of yourself and they’ll render a cool picture of you as an astronaut or  a scientist or something. And it was a good idea and people liked it. And it’s roughly just took off like crazy like that through last December. And then it did that. And that’s because Lenza was, it was a good idea. People liked it. But it was a relatively thin software layer on top of someone  else’s really powerful APIs. And so even though it was a it was a useful product, it was in a defensible business. 

但我必須警告，發展新技術時短期內會出現熱潮。你們記得Lenza嗎？你可以上傳照片，這個應用程式會生成一張你作為太空人或科學家的酷照片。大家都喜歡這個主意，去年12月它很受歡迎，但它只是建立在他人強大API上，相對薄弱的軟體層。這個產品有用處，但它沒有防禦性。

And when I think about, you know, apps like Lenza, I’m reminded of when Steve Jobs gave us the iPhone. Shortly after, someone wrote an app that I paid $199 for to do this to turn on the LED to turn the phone into a flashlight. And that was also a good idea to write an app to turn on the LED light. But it also wasn’t a defensible long term, it also  didn’t create very long term value because it was a easy replicated and underpriced and then eventually incorporated into iOS. But with the rise of iOS with the rise of iPhone, someone also figured out how to build things like Uber and Airbnb and Tinder, the very long term, very  defensible businesses that created, you know, sustaining value. And I think with the rise of generative AI or the rise of new AI tools, I think what really what excites me is the opportunity to create those really deep, really hard applications that hopefully can create very  long term value. So the first trend I want to share is AI as a general purpose technology and a lot of work that lies ahead of us is to find the very diverse use cases and to build them. 

Lenza這樣的應用程式，讓我想起當時賈伯斯推出iPhone的情景。有人寫了一個應用程式，我花了199美元購買，就為了開啟LED燈，將手機變成手電筒。這是一個好主意，但它同樣也無法長期作為防禦，因為它沒有創造出長期價值，容易被複製，價格又便宜，最後iOS就整合了這個功能。但隨著iOS、iPhone崛起，有其他人想出如何建立像Uber、Airbnb和Tinder這樣的事業，它們都是長期且具有防禦性，能夠持續創造價值。我認為隨著生成AI或新的AI工具的崛起，真正讓我興奮的機會在於，創建那些真正深入、非常困難的應用程式，提供非常長期的價值。我想分享的第一個趨勢是，AI是一種通用技術，我們所要做的事情就是找出多元的應用案例投入AI解決問題。

There’s a second trend I want to share with you, which relates to why AI isn’t more widely adopted  yet. It feels like a bunch of us have been talking about AI for like 15 years or something. But if you look at where the value of AI is today, a lot of it is still very concentrated in consumer software internet, right? Once you go outside, you know, tech or consumer software internet, there’s some AI adoption, but a lot of use very early. So why is that? 

論述二：AI還未大規模應用，因為導入成本高

第二個趨勢與AI還沒被廣泛採用的理由有關。許多人談論AI好像有15年之久，但如果你看看AI今日的價值所在，仍然集中於消費者領域，特別是軟體網路。其他領域的確也在採用AI，但仍在早期階段，為何會如此呢？

It turns out if you were to take all current and potential AI projects and sort them in decreasing order of value, then to the left of this curve, the head of this curve are the multi-billion dollar projects like  advertising or web search or for e-commerce, product recommendations or company like Amazon. And it turns out that about 10, 15 years ago, you know, various of my friends and I, we figured out a recipe for how to hire, say, 100 engineers to write one piece of software to serve more relevant  ads and apply that one piece of software to a billion users and generate massive financial value. So that works. But once you go outside consumer software internet, hardly anyone has 100 million or a billion users, they can write and apply one piece of software to.  So once you go to other industries, as we go from the head of this curve on the left, over to the long tail, these are some of the projects I see and I’m excited about.

如果你把所有現有和潛在的AI專案依照價值降冪排列，這條曲線左側的頂端是廣告、網路搜尋、電子商務、產品推薦，或者像亞馬遜一類的企業，執行數值億美元的專案。大約10、15年前，我和我的朋友聘請100名工程師來撰寫軟體，提供更相關的廣告，接著將軟體應用於十億用戶，進而產生巨大財務價值，這是可行的。然而，一旦走出消費者軟體網路，幾乎沒有人能夠做到這一點。對有著一億或十億用戶規模的事業來說，他們可以撰寫（挹注大量資源的）軟體，但在其他產業，也就是從曲線頂端過渡到長尾，這些我們感到興奮的專案，情況就不同了。

I was working with a pizza maker that was taking pictures of the pizza they were making  because they needed to do things like make sure that the cheese is spread evenly. So this is about a $5 million project. But that recipe of hiring 100 engineers or dozens of engineers to work on a $5 million project, that doesn’t make sense. Or another example, working with an agriculture company that with them, we figured out that if we use cameras to find out how tall is the wheat, and wheat is often bent over because of wind or rain or something, and we can chop off the wheat at the right height, then that results in more food for  the farmer to sell and is also better for the environment. But this is another, you know, $5 million project that that old recipe of hiring a large group of highly skilled engineers to work on this one project, that doesn’t make sense. And similarly, materials grading, cloth grading, sheet metal grading, many projects like this.

方向二：出現低／無程式碼工具，讓客製化變得便宜

我正在與一間披薩製造商合作，他們拍攝製作中披薩的照片，希望確保奶酪均勻地撒在披薩上。這個專案價值大約500萬美元，但出於成本考量，不會為此專案聘請100位工程師。我舉另一個例子，小麥常常因風或雨而彎曲，因此我們與一家農業公司合作，使用相機測量小麥高度，若能夠在正確高度砍下小麥，可以讓農民出售更多食物，對環境也更好。但這又是另一個價值500萬美元的項目，同樣不可能聘請一大堆具備高階技能的工程師。相似地，材料分級、布料分級、鈑金分級，有許多像這樣的專案。

So whereas to the left, in the head of this curve, there’s a small number of let’s say, multi-billion dollar projects, and we know how to execute those, you know, delivering value. In other industries, I’m seeing a very long tail of tens of thousands of let’s call them $5 million projects that until now have been very difficult to execute on because of the high cost of customization. The trend that I think is exciting is that the AI community has been building better tools that let us aggregate these use cases and make it easy for the end user to do the customization. So specifically, I’m seeing a lot of exciting low code and no code tools that enable the user to customize the AI system. 

在這個曲線的左側，有少數的、價值數十億美元的專案，我們知道如何執行、提供價值。其他產業有數以萬計的長尾效應，我們可以稱之為500萬美元的專案，由於客製化成本高昂，這些專案很難執行。然而，令人興奮的趨勢正在發生－AI社群一直在建立更好的工具，讓我們可以彙整這些使用案例，並讓終端使用者能夠輕鬆客製化。具體來說，有許多低程式碼（low code）和無程式碼（no code）工具，讓使用者能夠客製化AI系統，你不用寫程式，可以直接使用這些工具客製化AI系統。

What this means is instead of me needing to worry that much about pictures of Pizza, we have tools, we’re starting to see tools that can enable the IT departments of the Pizza making factory to train an AI system on their own pictures of Pizza to realize this $5 million worth of value. And by the way, the pictures of Pisa, you know, they don’t exist on the internet. So Google and Bing don’t have access to these pictures.  We need tools that can be used by really the Pizza factory themselves to build and deploy and maintain  their own custom AI system that works on their own pictures of Pisa. And broadly, the technology for enabling this, some of it is prompting, text prompting, visual prompting, but really large language models and similar tools like that, or a technology called data centric AI, whereby instead of asking the Pizza factory to write a lot of code, you know, which is challenging, we can ask them to provide data, which turns out to be more feasible.

這些工具讓我們不用擔心披薩的案例。我們開始看到，披薩的IT部門能夠自行訓練AI系統，辨識自己的披薩照片，這些工具能夠幫助實現500萬美元的價值。順帶一提，網路上沒有這些披薩照片，Google和Bing無法存取。上面提的工具，可以讓披薩工廠自己使用，建立、部署和維護客製化AI系統，這個系統可以處理他們的比薩照片。廣義來說，實現這個目標的技術當中，有一部分是文字提示、視覺提示，實際上就是大型語言模型和類似的工具，或稱為「以數據為中心的AI」，這種技術不用要求披薩工廠寫程式，這對他們來如說不容易，而是要求他們提供數據，這實際上更可行。

And I think the second trend is important, because I think this is a key part of the recipe for taking the value of AI, which so far still feels very concentrated in the tech world and consumer software internet world, and pushing this out to, you know, all industries really to the rest of the economy, which, you know, sometimes it’s easy to forget, the rest of the economy is much bigger than the tech world. So, the two trends I shared, AI is a general purpose technology,  lots of concrete use cases to be realized, as well as local, no code, easily used tools, enabling AI to be deployed in more industries. How do we go after these opportunities?

我認為第二個趨勢非常重要，因為AI的價值目前仍然非常集中在科技和消費者軟體網路世界，現在要推廣到所有行業，實質上是拓展到經濟體系的關鍵部分。有時候我們很容易忘記，經濟體系的其餘部分其實是遠大於科技世界。所以，我分享的兩個趨勢，第一是AI是一種通用技術，有許多具體的使用案例等待挖掘與實現，第二則是本地化、不用寫程式、容易使用的工具，在更多產業中部署AI。我們該如何抓住這些機會呢？

論述三：拆解AI應用金字塔，只有應用發展好，技術才會好

So about five years ago, there was a puzzle I wanted to solve, which is I felt that  many valuable AI projects are now possible. I was thinking, how do we get them done? And having led AI teams in, you know, Google and Baidu in big tech companies, I had a hard time figuring out how I could operate a team in a big tech company to go after a very diverse set of opportunities and everything from maritime shipping to education to financial services, to healthcare and on and on. It’s just very diverse use cases, very diverse, go to markets and very diverse, really, you know, customer bases and applications. And I felt that the most efficient  way to do this would be if we can start a lot of different companies to pursue these very diverse opportunities. 

大約五年前，我想解決一個問題，我覺得有許多有價值的AI項目是可行的。我在思考，應該如何完成它們呢？我曾在Google和Baidu等大型科技公司領導AI團隊，但想在這些科技巨頭，經營多元領域的團隊並不容易，從海運到教育，再到金融服務、醫療保健等，有許多使用案例，客戶群和應用都很多元。我覺得最有效的方式會是，創立許多不同的公司來追求這些非常多元化的機會。

So that’s why I ended up starting AI Fund, which is a venture studio that builds startups to pursue a diverse set of AI opportunities. And of course, in addition to lots of  startups, incumbent companies also have a lot of opportunities to integrate AI into existing businesses. In fact, one pattern I’m seeing for incumbent businesses is distribution is often one of the cyclical advantages of incumbent companies is they play the cards right, can allow them to integrate AI into their products quite efficiently. But just to be concrete, where are the opportunities? 

所以這就是我最後創立AI Fund的原因，這是一個創業工作室，專門建立新創，以追尋多元的AI機會。當然，除了新創以外，也有許多企業有將AI整合到現有業務的機會。現有公司的週期性優勢之一在於，如果他們打好牌，可以有效將AI整合到產品中。但具體來說，機會在哪裡呢？

So I think of this as a this is what I think of as the AI stack. At the bottom level is the hardware semiconductor layer, fantastic opportunities there, but very capital intensive, very concentrated. So there’s a lot of resources, relatively few winners. So some people can and should play there. I personally don’t like to play that myself. There’s also the infrastructure layer, also fantastic opportunities, but very capital intensive, very concentrated. So I tend not to play that myself either.  And then there’s a developer tool layer. What I showed you just now was I was actually using OpenAI’s API as a developer tool. And then I think the developer tool sector is hyper competitive. Look at all the startups chasing OpenAI right now. But there will be some mega winners. And so I  sometimes play here. But primarily when I think of a meaningful technology advantage, because I think that earns you the right or earns you a better shot at being one of the mega winners. 

我用「AI堆疊」（AI stack）描述我的想法。最底層是硬體，也就是半導體層，那裡機會極佳，但資本密集度高，且高度集中。那裡資源眾多，但贏家相對較少，有些人有能力在那裡投資，也應該這樣做，但我不喜歡進行這種投資。再往上是基礎設施層，也有極好的機會，同樣資本需求極大，且高度集中，我同樣不在這層投資。接下來是開發者工具層。我剛剛展示的程式碼，就是利用OpenAI的API作為開發者工具，這層的競爭激烈，看看現在所有追趕OpenAI的新創。不過，這個層次會有一些超級贏家，所以我有時會在這裡投資。當我思考「有意義的技術優勢」時，它將會帶給你權利，或者讓你有機會成為超級贏家。

And then lastly, even though a lot of the media attention and the buzz is in the infrastructure and developer  tooling layer, it turns out that that layer can be successful only if the application layer is even more successful. And we saw this with the rise of SaaS as well. A lot of the buzz, the excitement is on the technology, the tooling layer, which is fine, nothing wrong with that. But the  only way for that to be successful is that the application layer is even more successful so that  frankly, they can generate enough revenue to pay the infrastructure and the tooling layer. 

最後，儘管大部分的媒體關注和熱議都集中在基礎設施和開發者工具層面，但事實證明，只有當應用層面更為成功時，其他層面才能成功。我們也在軟體即服務（SaaS）的崛起中看到了這一點。大部分的熱潮圍繞於技術和工具，這沒有問題，但想要讓這些事物成功，唯一方法就是應用層面足夠成功，以便產生足夠的收入，支付基礎設施和工具層面的費用。

So actually, let me mention one example. Armourai, I was actually just texting the CEO yesterday. But Armourai is a company that we built that uses AI for romantic relationship coaching, right?  And just to point out, I’m an AI guy, and I feel like I know nothing really about romance. And if you don’t believe me, you can ask my wife, she will confirm that I know nothing about romance. But we want to build this, we wanted to get together with the former CEO of Tinder, Renata  Nyborg, and with my team’s expertise in AI, and her expertise in relationships, because she ran Tinder, she knows more about relationships than anyone I know. We’re able to build something pretty unique using AI for kind of romantic relationship mentoring. And the interesting  thing about applications like these is when we look around, you know, how many teams in the world are simultaneously expert in AI and in relationships. 

讓我舉一個例子：Armourai，它是一間我們打造，專門利用AI進行戀愛關係指導的公司。我是一個AI專家，但我對浪漫一無所知，如果你不相信我，你可以問我的妻子，她會證實我的說法。我們希望能與前Tinder的首席執行官Renata Nyborg一起，結合我的團隊在AI的專業知識，以及她在關係方面的專業知識，因為她曾經管理過Tinder，她對於關係的了解比我認識的任何人都要多。我們能夠建立一些東西，獨特地使用AI進行某種浪漫關係的指導。像這樣的應用程式的有趣之處在於，當我們四處觀察，你知道，全世界有多少團隊同時精通AI和關係。

And so at the application layer, I’m seeing a lot of exciting opportunities that seem to have a very large market, but where the competition set  is very light relative to the magnitude of the opportunity. It’s not that there are no competitors, but it’s just much less intense compared to the developer tool or the infrastructure layer. And so because I’ve spent a lot of time iterating on a process of building startups, what I’m going  to do is just, you know, very transparently tell you the recipe we’ve developed for building startups. And so after many years of iteration and improvement, this is how we now build startups.

方向三：找能夠具體落地的想法、找領域專家合作

在應用層面存在許多令人興奮的機會，這些機會似乎擁有非常大的市場，但相對於機會的規模，競爭者卻相對較少。這並不是說沒有競爭者，只是相較於開發者工具或基礎設施層面，競爭強度要小得多。由於我花了很多時間在創業過程中反覆嘗試，我將要做的就是，透明地告訴你，我們為創業所開發的配方，也就是我們現在創業的方式。

My teams always had access to a lot of different ideas, you know, internally generated ideas from partners. I want to walk through this with one example of something we did, which is a company, Bearing.ai, which uses AI to make ships more fuel efficient. So this idea came to me when a few years ago, a large Japanese conglomerate called Mitsui, that is a major shareholder and sort of operates a major shipping lines, they came to me and they said, hey, Andrew, you should build a business to use AI to make ships more fuel efficient. And the specific idea was, you know, think of it as Google Maps for ships, we can suggest a ship or tell a ship how to steer so that you still get to your destination on time, but using it turns out about 10% less fuel. And so what we now do is we spend about a month validating the idea. So double check, is this idea even technically feasible, and then talk to prospective customers to make sure there’s a market need. So we spent up to about a month doing that. And if it passes this stage, then we will go and recruit a CEO to work with us on the project. 

我的團隊總是能從夥伴身上，接觸到許多不同想法，我想用過往參與的一個例子說明。有一間企業Bearing.ai ，利用AI使船隻更加節能。幾年前，大型日本企業集團三井，它們經營一條航線，他們來找我說，你應該利用AI，使船隻更加節能。具體的想法是打造專為船隻設計的Google地圖，建議一艘船如何駕駛，藉此準時抵達目的地，還能節省約10%的燃料。我們花了大約一個月驗證想法，檢查在技術上是否可行，接著與潛在客戶對話，確保市場有需求。若能夠通過這個階段，我們就會去招募CEO參與這個專案。

When I was starting out, I used to spend a long time working on the project myself before bringing on the CEO. But after iterating, we realized that bring on the leader at the very beginning to work with us, it reduces a lot of the burden of having to transfer knowledge or having a CEO come in and have them revalidate whether we discover it. So the process we’ve learned much more efficient, which is bringing the leader at the very start.  And so in the case of bearing AI, we found a fantastic CEO Dylan Kyle, who’s a repeat  entrepreneur, one successful exhibit before. And then we spent three months, six two week sprints to work with them to build a prototype as well as do deep customer validation. If it survives this stage, and we have about a two thirds 66% survival rate, we then write the first check in  which then gives the company resources to hire an executive team, you know, build the key team, get MVP working, minimum viable product working and get some real customers. 

當我剛開始的時候，我常常花很長的時間在這上面工作。在引進CEO之前，我會先規劃自己的角色。但在反覆嘗試後，我們意識到，若在一開始就讓領導者參與會更有效率，因為可以減輕轉移知識的負擔，也不用讓他重新驗證我們的想法。因此，我們找到一位出色的CEO Dylan Kyle，他是一位經驗豐富的企業家，先前曾經取得成功。我們在三個月的時間內，進行6次為期2週的密集工作，與團隊一起建立原型，並進行深度的客戶驗證，如果專案能夠存活，接著會開出第一張支票，公司就有提供資源聘請執行團隊，讓最小可行產品運作，獲得真實的客戶。我們有大約三分之二，也就是66%的存活率。

And then after that, you know, hopefully then successfully raises additional external rounds of funding,  you can keep on growing and scaling. So I’m really proud of the work that my team was able to do to support Mitsui’s idea and Dylan Kyle as CEO. And today, there are hundreds of ships on the high seas right now that are steering themselves differently because of bearing AI. And 10% fuel savings translates the rough order amount to maybe $450,000 in savings and fuel per ship per year. And of course, is also frankly, quite a bit better for the environment. And I think this startup, I think would not have existed if not for Dylan’s fantastic work. 

在那之後，則是追求額外的外部融資，藉此繼續成長和擴展。我對我的團隊能夠支持三井的想法和Dylan Kyle作為CEO的工作，感到非常驕傲。現在，bearing AI的系統讓船隻改變控制航向的方式，並節省10%的燃料，每艘船每年可以節省約45萬美元的燃料費用，這對環境來說也是好事。我認為創業公司，我認為如果沒有Dylan的出色工作，它可能就不存在了。

And then also, you know, bringing this idea to me. And I like this example, because there’s another one is like, you know, this is a startup idea that just to point out, I would never have come up with myself, right? Because, you know, I’ve been on a boat, but what do I know about maritime shipping? But  is the deep subject matter expertise of Mitsui that had this insight together with Dylan and my team’s expertise in AI that made this possible? And so as I operate in AI, one thing I’ve learned is my swim lane is AI. And that’s it, because I don’t have time. It’s very difficult for me to be expert in maritime shipping, and romantic relationships and healthcare and financial services and on and on and on. And so I’ve learned that if I can just hope get accurate technical validation, and then use, you know, AI resources to make sure the AI tech is built quickly and well, and I think we’ve always managed to help the companies build a strong technical team, then partnering with subject matter experts often results in exciting new opportunities. 

我自己永遠不會想出這樣的點子，我的專業領域是人工智慧，就這樣。對我來說，要成為海運、浪漫關係、醫療保健、金融服務等各領域的專家非常困難。所以我學到專注於獲得準確的技術知識，這就夠了。驗證，然後使用AI相關資源，確保能夠建立AI技術，幫助企業建立強大的技術團隊，並且與主題專家合作，帶來令人興奮的新機會。

And I want to share with you one other weird aspect of another one of the weird lessons I’ve learned about, you know, building startups, which is, I like to engage only when there’s a concrete  idea. And this runs counter to a lot of the advice you hear from the design thinking methodology, which often says, don’t rush to solutioning, right? Explore a lot of alternatives, avoid the solution. Honestly, we tried that it was very slow. 

我想與你分享另一個我的心得。我喜歡在擁有具體想法後，才參與建立新創。這與你從設計思考方法論中聽到的許多建議相反，這種方法論常常告訴我們，不要急於解決問題，對吧？探索許多替代方案，避免解決方案。誠實地說，我們嘗試過，但進度非常緩慢。

But what we’ve learned is that at the ideation stage, if someone comes to me and says, Hey, Andrew, you should apply AI to financial services, because I’m not a subject matter expert in financial services, is very slow for me to go and learn enough about financial services, they can figure out what to do. I mean,  eventually, you could get a good outcome. But it’s a very labor intensive, very slow,  very expensive process of me to try to learn industry after industry. In contrast, one of my partners wrote this ideas a ton of cheap, not not really seriously. But you know, let’s say a conglomerate is BuyGPT, let’s eliminate commercials by automatically buying  every product advertised in exchange for not having seen the ads is not a good idea. But it is a concrete idea. And it turns out concrete ideas can be validated or falsified efficiently.

在構思階段，有人會對我說，你應該將AI應用於金融服務，但我不是金融服務的主題專家，學習足夠的金融服務知識，並找出該怎麼做是相當緩慢的。最後你可能會得到好結果，但認識一個又一個產業，是非常勞力密集、緩慢並且昂貴的過程。對比來說，我有位夥伴提了一個想法，它不切實際，只是為了說明所用。假設有一間大型企業集團「BuyGPT」，因為不想看到廣告，所以每次看到廣告，都會自動點擊、購買，藉此消掉廣告，這並不是一個好主意，但這確實是一個具體的想法。事實證明，具體的想法可以有效地被驗證。

They also give a team a clear direction to execute. And I’ve learned that in today’s world,  especially with, you know, the excitement and buzz and exposure to AI of a lot of people,  it turns out that there are a lot of subject matter experts in today’s world that have deeply thought about a problem for months, sometimes even, you know, one or two years, but they’ve not yet had a built partner. And when we get together with them, and hear and they share  the idea with us, it allows us to work with them to very quickly go into validation and building. And I find that this works because there are a lot of people that have already done the, you know, design thinking thing of exploring a lot of ideas and winning down to really good ideas. And there, I find that there’s so many good ideas sitting out there that no one is working on, that finding those good ideas that someone has already had and wants to share with us and wants to build a partner for that turns out to be a much more efficient engine. So, before I wrap up, we’ll go to the question in a second, just a few slides to talk about risk and social impact. 

這個例子告訴我們，AI專家能夠和領域專家合作。在這個對AI感到興奮的世界裡，有許多領域專家深入思考單一問題，長達數月、甚至一兩年，但他們沒有找到合適的建造夥伴。當我們與他們聚在一起，聆聽他們與我的想法，讓我們能迅速進行驗證和建造（產品）。我發現這種方式很有效，有很多人已經走過一遍。設計思考的方式是探索許多想法，並篩選出真正好的想法，但我發現有許多好方法，卻沒人實現。因此，找到已經有人想到，並且想與我們分享的好想法，跟他們成為合作夥伴，這是方法更有效率。在我結束前，我們來討論風險和社會影響。

So, AI is a very powerful technology. To state something you probably guessed, my teams and I, we only work on projects that move humanity forward. And, you know, we have multiple times killed projects that we assess to be financially sound based on ethical grounds. It turns out, I’ve been surprised and sometimes dismayed at the creativity of people to come up with good ideas, sorry, to come up with really bad ideas that seem profitable, but really should not be built.  We’ve killed a few projects on those grounds. And then, I think, has to acknowledge that AI today does have problems with bias, fairness, accuracy, but also, you know, the technology is improving quickly. So, I see that AI systems today are less biased than six months  ago and more fair than six months ago, which is not to dismiss the importance of these problems. They are problems and we should continue to work on them. But I’m also gratified at the number of AI teams working hard on these issues to make them much better. 

轉折：AI有缺陷，但我們努力解決問題

AI是一種非常強大的技術。我的團隊和我只做能推動人類進步的項目，過去多次基於道德原因，終止財務上健全的項目。我對人們提出好主意的創造力感到驚訝，有時甚至沮喪，因為會有看似有利可圖，但不該實現的糟糕想法。我認為，必須承認人工智慧確實存在偏見，有著公平性與準確性問題，但同時科技正在迅速進步，今天的AI系統比六個月前偏見更少、比六個月前更公平。我不是說要忽視這些問題，我們應該努力解決它們。AI團隊正在努力解決這些問題，以使其變得更好。

When I think of the biggest risk of AI, I think that the biggest risk, one of the biggest risks is the disruption to jobs. This is a diagram from a paper by our friends at the University of Pennsylvania and some folks at OpenAI analyzing the exposure of different jobs to AI automation. And it turns out that whereas  the previous wave of automation, mainly the most exposed jobs were often the lower wage jobs, such as when, you know, we put robots into factories, with this current wave of automation, is actually the higher wage jobs further to the right of this axis that seems to have more of  their tasks exposed to AI automation. So, even as we create tremendous value using AI, I feel like as citizens and our corporations and the governments and really our society, I feel a strong obligation to make sure that people, especially people whose livelihoods are disrupted, are still well taken care of, are still treated well. 

我認為AI的最大風險之一，就是破壞工作。這是賓夕法尼亞大學發表論文中的圖表。OpenAI正在分析不同工作受AI自動化影響程度。上一波自動化，像是在工廠中使用機器人，對低薪工作影響最大，對比之下，這波自動化的影響範圍更廣。圖表中X軸右邊的高薪工作裡，有更多任務受AI自動化影響。即使我們使用AI創造巨大價值，作為公民，我認為企業、政府還有社會，有強烈的責任確保人們，尤其是那些生計受到干擾的人，仍然得到良好的照顧。

收尾：機器仍比不上人類，即使再進步，我們仍能駕馭它

And then lastly, there’s also been, it feels like every time there’s a big wave of progress in AI, you know, there’s a big wave of hype about  artificial general intelligence as well. When deep learning started to work really well 10 years ago,  there was a lot of hype about AGI and now that generative AI is working really well, there’s another wave of hype about AGI. But I think that artificial intelligence, AI didn’t do anything a human can do, it’s still decades away, you know, maybe 30 to 50 years, maybe even longer, I hope we’ll see it in  our lifetimes. But I don’t think there’s any time soon. One of the challenges is that the biological path to intelligence, like humans, and the digital path to intelligence, you know, AI, they’ve taken very different paths. And the funny thing about the definition of AGI is you’re benchmarking this  very different digital path to intelligence, with really the biological path to intelligence. 

最後，我還有一種感覺，每次人工智慧有重大進展時，都會有通用人工智慧的熱潮。當深度學習10年前運作得很好時，就有炒作通用人工智慧。現在生成式AI運作得非常好，又有另一波炒作。但我認為，人工智慧還不能做到人類能做的事，未來還需要可能30到50年，甚至更長的時間才能達成。我希望能在有生之年看到，但短期內不會發生。其中一個挑戰是，人類這種生物智慧的路徑，和AI的數位智慧路徑，兩者非常不同。AGI的定義有趣之處在於，你正在比較兩種很不同的智慧路徑。

So I think, you know, large language models are smarter than any of us in certain key dimensions, but much dumber than any of us in other dimensions. And so forcing it to do everything a human can do is like  a funny comparison. But I hope we’ll get there, maybe, hopefully within our lifetimes. And then there’s also a lot of I think, overblown hype about AI creating extinction risk for humanity. Candidly, I don’t see it. I just don’t see how AI creates any meaningful extinction risk for  humanity. I think that people worry we can’t control AI, but we have lots of AI will be more powerful than any person. But with lots of experience steering very powerful entities, such as corporations or nation states that are far more powerful than any single person, and making sure they for the most part benefit humanity. 

我認為大型語言模型在某些關鍵維度上，比我們任何人都要聰明，在很多方面又大大不如我們，比任何人都要愚蠢。強迫它去做一個人能做的事情，是一種可笑的比較。我希望在我們的有生之年能夠真的做到。我認為現在有很多AI帶來人類帶來滅絕風險的過度炒作，坦白說，我看不出來人工智慧怎麼帶來任何有意義的滅絕風險。有人很多人工智慧將比任何人都更強大，擔心我們無法控制人工智慧，但我們已經有經驗駕馭強大的實體，例如比任何人都強大得多的公司或國家，同時確保他們在大多數情況下都能造福人類。

And also technology develops gradually. The so-called hot takeoff scenario, where it’s not really working today. And then suddenly, one day overnight, it works brilliantly, and we achieve super intelligence that takes over the world.  That’s just not realistic. And I think AI technology would develop slowly, like all the  time, you know, and then it gives us plenty of time to make sure that we provide oversight and manage it to be safe. And lastly, if you look at the real extinction risk to humanity, such as fingers crossed the next pandemic, or climate change, leading to massive depopulation  of some parts of the planet, or much lower odds, but maybe someday, an asteroid doing to us what it had done to the dinosaurs. I think if you look at the actual real extinction risk to humanity, AI having more intelligence, even artificial intelligence in the world, will be a key part of the solution. So I feel like if you want humanity to survive and thrive  for the next 1000 years, rather than slowing AI down, which some people propose, I would rather, I would rather make AI go as fast as possible. 

而且，技術是逐步發展的。所謂「熱起飛」情境－也就是今天還不太能運作，突然一夜之間運作出色，出現能夠接管世界的超級智慧－這種情況其實並不現實。我認為AI技術會像平常一樣慢慢發展，給了我們充足時間，確保有適當的監督，並確保安全。如果你看那些對人類構成真正滅絕風險的事物，比如下一次大瘟疫、或是因氣候變化導致地球某些部分大規模的人口減少，或者風險更低的，也許某天小行星襲擊，就像先前襲擊恐龍那樣。如果你看對人類構成實際滅絕風險的因素，AI擁有更多的智慧，人工智慧將成為解決這個問題的核心。如果你想讓人類在接下來的1000年裡生存和繁榮，與其放慢AI的發展速度，我寧願讓AI發展得更快。

So with that, just to summarize, this is my last slide. I think that AI, as a general purpose technology, creates a lot of new opportunities  for everyone. And a lot of the exciting and important work that lies ahead of us all is to go and build those concrete use cases. And hopefully, in the future, hopefully I have opportunities to maybe engage with more of you on those opportunities as well.  So that let me just say thank you all very much.

總結一下，這是我的最後一張簡報。我認為作為一種通用技術，人工智慧為大家創造了很多新的機會。我們激動人心和重要的工作很多，也就是去構建那些具體的應用案例。希望在未來，也希望我有機會能和大家更多地探討這些機會。因此，讓我就說一聲，非常感謝大家。

最後，有兩篇延伸閱讀讓大家參考：

• 前幾天聽了前Stanford教授兼人工智能實驗室主任、Google… – 矽谷next I Michelle Cheng
• 與ChatGPT推手對談／吳恩達：生成式AI正迎來黃金十年！

• AI
• Andrew Ng
• Artificial Intelligence
• finance
• info gathering
• 吳恩達

• [資訊整理] 潘威倫嘟嘟引退宣言全文 | 中華職棒統一獅 – 逐字稿
  2024-09-30 0
• [資訊整理] 中華職棒CPBL球員綽號怎麼來？球員常見問題有哪些？2024年最新
  2024-09-23 0
• [資訊整理] 周思齊引退宣言暨賽後記者會全文 | 中華職棒中信兄弟 – 逐字稿
  2024-09-23 0