Good to be here.

很高興來到這裡。

Thanks for joining us.

感謝您的參與。

Thanks for having us.

感謝您的光臨。

The economists got them together to discuss what this technology means for our future, from employment and geopolitics to the survival of liberal democracy.

經濟學家們把他們聚在一起，討論這項技術對我們的未來意味著什麼，從就業、地緣政治到自由民主的生存。

If the economic system is fundamentally changed, will liberal democracy as we know it survive?

如果經濟體系發生了根本性的變化，我們所熟知的自由民主制度還能生存下去嗎？

Yuval Noah Harari, welcome.

尤瓦爾-諾亞-哈拉里，歡迎你。

You are a bestselling author, historian, I think a global public intellectual, if not the global public intellectual.

您是暢銷書作家、歷史學家，我認為您還是全球公共知識分子，甚至是全球公共知識分子之首。

Your books from Sapiens to 21 Lessons from the 21st Century have sold huge numbers of copies around the world.

從《智者》到《21 世紀的 21 堂課》，您的著作在全世界銷量巨大。

Thank you for joining us.

感謝您的參與。

It's good to be here.

很高興來到這裡。

Mustafa Suleiman, wonderful that you can join us too.

穆斯塔法-蘇萊曼，很高興你也能加入我們。

You're a friend of The Economist, a fellow director on The Economist board.

你是《經濟學人》的朋友，也是《經濟學人》董事會的董事之一。

You are a man at the cutting edge of creating the AI revolution.

您是創造人工智能革命的最前沿人物。

You are a co-founder of DeepMind.

您是 DeepMind 的聯合創始人之一。

You're now a co-founder and CEO of Inflection AI.

您現在是 Inflection AI 公司的聯合創始人兼首席執行官。

You are building this future, but you've also just published a book called The Coming Wave, which makes us a little concerned about this revolution that is being unleashed.

你正在建設這個未來，但你也剛剛出版了一本名為《即將到來的浪潮》的書，這讓我們對這場正在掀起的革命有些擔憂。

You're both coming from different backgrounds.

你們都來自不同的背景。

You are a historian, a commenter, a man who I believe doesn't use smartphones very much.

您是一位歷史學家，一位評論家，一位我相信不怎麼使用智能手機的人。

Not very much, no.

不是很多，沒有。

Mustafa, as I know from our board meetings, is right at the cutting edge of this, pushing everyone to go faster.

穆斯塔法，正如我在董事會會議上所瞭解的那樣，正站在這個領域的最前沿，推動著每個人走得更快。

So, two very different perspectives, but I thought it would be really interesting to bring the two of you together to have a conversation about what is happening, what is going to happen, what are the opportunities, but also what is at stake and what are the risks.

是以，你們的觀點截然不同，但我認為，把你們兩位召集在一起，就正在發生的事情、將要發生的事情、機遇以及利害關係和風險展開對話，會非常有趣。

So, let's start, Mustafa, with you.

穆斯塔法，我們從你開始吧。

And you are building this future.

而你們正在建設這個未來。

So, paint us a picture of what the future is going to be like.

那麼，請為我們描繪一下未來的景象吧。

And I'm going to give you a timeframe to keep it specific.

我會給你一個時間框架，讓它具體化。

So, let's say, I think you wrote in your book that within three to five years that you thought it was plausible that AIs could have human-level capability across a whole range of things.

所以，比方說，我想你在書中寫道，你認為在三到五年內，人工智能可以在一系列事情上具備人類水準的能力。

So, let's take five years, 2028.

那麼，讓我們用五年的時間，也就是 2028 年。

What does the world look like?

世界是什麼樣子的？

How will I interact with AIs?

我將如何與人工智能互動？

What will we all be doing and not doing?

我們都要做什麼，不做什麼？

Well, let's just look back over the last 10 years to get a sense of the trajectory that we're on and the incredible momentum that I think everybody can now see with the generative AI revolution.

好吧，讓我們回顧一下過去的 10 年，瞭解一下我們的發展軌跡，以及我認為每個人現在都能看到的生成式人工智能革命的驚人勢頭。

Over the last 10 years, we've become very, very good at classifying information.

在過去的 10 年裡，我們變得非常非常擅長對資訊進行分類。

We can understand it, we sort it, label it, organize it, and that classification has been critical to enabling this next wave because we can now read the content of images.

我們可以理解它、分類它、標記它、組織它，這種分類對實現下一波浪潮至關重要，因為我們現在可以讀取影像的內容。

We can understand text pretty well.

我們能很好地理解文字。

We can classify audio and transcribe it into text.

我們可以對音頻進行分類，並將其轉錄為文本。

The machines can now have a pretty good sense of the conceptual representations in those ideas.

現在，機器可以很好地感知這些想法中的概念表徵。

The next phase of that is what we're seeing now with the generative AI revolution.

下一階段就是我們現在看到的生成式人工智能革命。

We can now produce new images, new videos, new audio, and, of course, new language.

我們現在可以製作新的影像、新的視頻、新的音頻，當然還有新的語言。

And in the last year or so, with the rise of chat GPT and other AI models, it's pretty incredible to see how plausible and accurate and very finessed to these new language models are.

而在過去一年左右的時間裡，隨著哈拉 GPT 和其他人工智能模型的興起，我們看到這些新的語言模型是多麼可信、準確和精細，這真是令人難以置信。

In the next five years, the frontier model companies, those of us at the very cutting edge who are training the very largest AI models, are going to train models that are over 1,000 times larger than what you currently see today in GPT-4.

未來五年，前沿模型公司，也就是我們這些正在訓練最大人工智能模型的最前沿公司，將訓練出比目前在 GPT-4 中看到的模型大 1000 多倍的模型。

And with each new order of magnitude and compute, that is 10x more compute used, we tend to see really new capabilities emerge.

每增加一個數量級和計算能力，也就是增加 10 倍的計算能力，我們就會看到真正的新功能出現。

And we predict that the new capabilities that will come this time, over the next five years, will be the ability to plan over multiple time horizons.

我們預測，在未來五年內，將出現的新能力將是在多個時間範圍內進行規劃的能力。

Instead of just generate new text in a one shot, the model will be able to generate a sequence of actions over time.

該模型不再是一次性生成新文本，而是能夠隨著時間的推移生成一系列動作。

And I think that that's really the character of AI that we'll see in the next five years.

我認為，這才是未來五年我們將看到的人工智能的真正特徵。

Artificial capable AIs, AIs that can't just say things, they can also do things.

人工智能不僅能說話，還能做事。

What does that actually mean in practice?

這實際上意味著什麼？

Just use your imagination.

發揮你的想象力吧。

Tell me what my life will be like in 2028.

告訴我 2028 年我的生活會是什麼樣子。

How will I interact with them?

我將如何與他們互動？

What will I do?

我該怎麼辦？

What will be different?

有什麼不同？

I've actually proposed a modern Turing test, which tries to evaluate for exactly this point.

實際上，我已經提出了一個現代圖靈測試，它試圖評估的正是這一點。

The last Turing test simply evaluated for what a machine could say, assuming that what it could say represented its intelligence.

最後一次圖靈測試只是評估機器能說什麼，假定它能說什麼就代表它的智能。

Now that we're kind of approaching that moment where these AI models are pretty good, arguably they've passed the Turing test, or maybe they will in the next few years.

現在，我們正在接近這樣一個時刻：這些人工智能模型已經相當不錯了，可以說它們已經通過了圖靈測試，或者說它們在未來幾年內就會通過圖靈測試。

The real question is, how can we measure what they can do?

真正的問題是，我們如何衡量它們的能力？

So I've proposed a test which involves them going off and taking a $100,000 investment, and over the course of three months, trying to set about creating a new product, researching the market, seeing what consumers might like, generating some new images, some blueprints of how to manufacture that product, contacting a manufacturer, getting it made, negotiating the price, drop shipping it, and then ultimately collecting the revenue.

是以，我提出了一個測試方案，讓他們拿出 10 萬美元投資，在三個月的時間裡，嘗試著手創造一種新產品，研究市場，看看消費者可能會喜歡什麼，生成一些新圖片，繪製一些如何生產該產品的藍圖，聯繫製造商，生產產品，談好價格，下單發貨，然後最終收取收益。

And I think that over a five-year period, it's quite likely that we will have an ACI, an artificial capable intelligence, that can do the majority of that task autonomously.

我認為，在五年的時間裡，我們很有可能會擁有一個 ACI（人工智能），一個能夠自主完成大部分任務的人工智能。

It won't be able to do the whole thing, there are many tricky steps along the way, but significant portions of that.

它無法完成整個過程，途中會有許多棘手的步驟，但其中相當一部分是可以完成的。

It will be able to make phone calls to other humans to negotiate.

它還能打電話給其他人進行談判。

It will be able to call other AIs in order to establish the right sequence in a supply chain, for example.

例如，它可以呼叫其他人工智能，以便在供應鏈中建立正確的順序。

And of course, it will learn to use APIs, Application Programming Interfaces, so other websites or other knowledge bases or other information stores.

當然，它還將學會使用 API（應用程序編程接口），以便使用其他網站或其他知識庫或其他資訊存儲。

And so, you know, the world is your oyster.

所以，你知道，世界是你的牡蠣。

You can imagine that being applied to many, many different parts of our economy.

你可以想象，這適用於我們經濟的許多不同部分。

So Yuval, a man who doesn't use a smartphone very much, you listen to this, does it fill you with horror, and do you agree with it?

那麼尤瓦爾，一個不怎麼使用智能手機的人，你聽了這番話，是否感到恐懼，你同意嗎？

Do you think that's the kind of thing that is likely to happen in the next five years?

你認為未來五年會出現這種情況嗎？

I would take it very seriously.

我會非常認真地對待它。

I don't know.

我不知道。

I'm not coming from within the industry, so I cannot comment on how likely it is to happen.

我不是業內人士，所以無法評論這種可能性有多大。

But when I hear this as a historian, for me, what we just heard, this is the end of human history.

但是，當我作為一個歷史學家聽到這句話時，對我來說，我們剛才聽到的，是人類歷史的終結。

Not the end of history, the end of human-dominated history.

不是歷史的終結，而是人類主導的歷史的終結。

History will continue with somebody else in control.

歷史將在別人的掌控下繼續。

Because what we just heard is basically Mustafa telling us that in five years, there'll be a technology that can make decisions independently and that can create new ideas independently.

因為我們剛才聽到的基本上是穆斯塔法在告訴我們，五年後將會出現一種能夠獨立做出決策、能夠獨立創造新想法的技術。

This is the first time in history we confronted something like this.

這是我們有史以來第一次遇到這樣的事情。

Every previous technology in history, from a stone knife to nuclear bombs, it could not make decisions, like the decision to drop the bomb on Hiroshima was not made by the atom bomb.

歷史上的每一種技術，從石刀到核彈，都無法做出決定，就像在廣島投擲原子彈的決定不是由原子彈做出的。

It was made by President Truman.

這是杜魯門總統提出的。

And similarly, every previous technology in history, it could only replicate our ideas.

同樣，歷史上的每一種技術都只能複製我們的想法。

Like radio or the printing press, it could make copies and disseminate the music or the poems or the novels that some human wrote.

就像收音機或印刷機一樣，它可以複製和傳播某些人創作的音樂、詩歌或小說。

Now we have a technology that can create completely new ideas.

現在，我們有了一種可以創造全新創意的技術。

And it can do it at a scale far beyond what humans are capable of.

而且其規模遠遠超出人類的能力。

So it can create new ideas and in important areas within five years, we'll be able to enact them.

是以，它可以創造新的想法，在重要領域，我們將在五年內將其付諸實施。

And that is a profound shift.

這是一個深刻的轉變。

Before we go on to the many ways in which this could be the end of human history, as you put it, and the potential downsides and risks of this, can we just for a second, just indulge me?

在我們繼續討論你所說的這可能是人類歷史終結的多種方式以及潛在的弊端和風險之前，我們能不能先聽我說幾句？

I'm an optimist at heart.

我骨子裡是個樂觀主義者。

Can we talk about the possibilities?

我們能談談可能性嗎？

What are the potential upsides of this?

這樣做的潛在好處是什麼？

Because there are many, and they are really substantial.

因為有很多，而且真的很有分量。

I think you wrote that there is the potential that this technology can help us deal with incredibly difficult problems and create tremendously positive outcomes.

我認為你在信中寫道，這項技術有可能幫助我們解決令人難以置信的難題，並創造出巨大的積極成果。

So can we just briefly start with that before we go down the road of the terrible things?

那麼，我們能不能在討論那些可怕的事情之前，先簡單說說這個問題？

When I say the end of human history, again, I'm not talking necessarily about the destruction of humankind or anything like that.

當我說人類歷史的終結時，我並不是在說人類一定會毀滅或類似的事情。

There are many positive potential.

有許多積極的潛力。

It's just that control, power, is shifting away from human beings to an alien intelligence, to a non-human intelligence.

只是，控制權、權力正在從人類手中轉移到外星智慧生物，轉移到非人類智慧生物手中。

We'll also get to that because there's a question of how much power, but let's stick with the potential upsides first.

我們還將討論這個問題，因為這涉及到功率大小的問題，但我們還是先說說潛在的好處吧。

The opportunities must suffer.

機會必須受苦。

Everything that we have created in human history is a product of our intelligence.

人類歷史上所創造的一切，都是我們智慧的產物。

Our ability to make predictions and then intervene on those predictions to change the course of the world is, in a very abstract way, the way we have produced our companies and our products and all the value that has changed our century.

我們能夠做出預測，然後對這些預測進行干預，從而改變世界的進程，從一個非常抽象的角度來說，這就是我們生產公司和產品的方式，以及改變我們這個世紀的所有價值。

If you think about it, just a century ago, a kilo of grain would have taken 50 times more labor to produce than it does today.

想想看，就在一個世紀前，生產一公斤穀物所需的勞動力是現在的 50 倍。

That efficiency, which is the trajectory you've seen in agriculture, is likely to be the same trajectory that we will see in intelligence.

這種效率，就是你們在農業領域看到的軌跡，也很可能是我們在智能領域看到的軌跡。

Everything around us is a product of intelligence, and so everything that we touch with these new tools is likely to produce far more value than we've ever seen before.

我們周圍的一切都是智能的產物，是以，我們用這些新工具接觸到的一切都可能產生比我們以前所看到的更多的價值。

I think it's important to say these are not autonomous tools by default.

我認為有必要說明的是，這些並非默認的自主工具。

These capabilities don't just naturally emerge from the models.

這些能力並不是從模型中自然產生的。

We attempt to engineer capabilities, and the challenge for us is to be very deliberate and precise and careful about those capabilities that we want to emerge from the model, that we want to build into the model, and the constraints that we build around it.

我們試圖設計能力，而我們面臨的挑戰是要非常謹慎、準確和小心地對待我們希望從模型中產生的能力、我們希望在模型中建立的能力以及我們圍繞模型建立的約束條件。

It's super important not to anthropomorphically project ideas and potential intentions or potential agency or potential autonomy into these models.

不要把想法、潛在意圖、潛在代理權或潛在自主權擬人化地投射到這些模型中，這一點非常重要。

The governance challenge for us over the next couple of decades to ensure that we contain this wave is to ensure that we always get to impose our constraints on the trajectory of this development.

在未來幾十年裡，為確保遏制這一浪潮，我們在治理方面面臨的挑戰是確保我們始終能夠對這一發展的軌跡施加限制。

But the capabilities that will arise will mean, for example, potentially transformative improvements in human health, speeding up the process of innovation, dramatic changes in the way scientific discovery is done, tough problems, whether it's climate change.

但是，所產生的能力將意味著，例如，人類健康的潛在變革性改善、創新進程的加速、科學發現方式的巨大變化、棘手問題（無論是氣候變化還是其他問題）的解決。

A lot of the big challenges that we face could be much more easily addressed with this capability.

有了這種能力，我們面臨的許多重大挑戰都可以更容易地得到解決。

Everybody is going to have a personal intelligence in their pocket, a smart and capable aide, a chief of staff, a research assistant, constantly prioritizing information for you, putting together the right synthesized nugget of knowledge that you need to take action on at any given moment.

每個人的口袋裡都會有一個個人情報員、一個精明能幹的助手、一個參謀長、一個研究助理，不斷地為你安排資訊的優先順序，把你在任何時候都需要採取行動的正確的綜合知識彙總起來。

And that for sure is going to make us all much, much smarter and more capable.

這肯定會讓我們變得更聰明、更有能力。

Does that part of it sound appealing to you, Yuval?

這部分內容聽起來吸引你嗎，尤瓦爾？

Yes, absolutely.

是的，當然。

I mean, again, if there was no positive potential, we wouldn't be sitting here.

我的意思是，再說一遍，如果沒有積極的潛力，我們就不會坐在這裡了。

Nobody would develop it.

沒有人會開發它。

Nobody would invest in it.

沒有人願意投資。

Again, it's so appealing.

同樣，它是如此吸引人。

The positive potential is so enormous in everything, again, from much better health care, higher living standards, solving things like climate change.

同樣，從更好的醫療保健、更高的生活水準到解決氣候變化等問題，一切方面的積極潛力都是如此巨大。

This is why it's so tempting.

這就是它如此誘人的原因。

This is why we are willing to take the enormous risks involved.

這就是我們願意承擔巨大風險的原因。

I'm just worried that in the end the deal will not be worth it.

我只是擔心到頭來這筆交易不值得。

And I would comment especially on, again, the notion of intelligence.

我想再次特別談談智力的概念。

I think it's overrated.

我認為它被高估了。

I mean, Homo sapiens at present is the most intelligent entity on the planet.

我的意思是，目前智人是地球上最聰明的實體。

It's simultaneously also the most destructive entity on the planet.

它同時也是地球上最具破壞性的實體。

And in some ways, also the most stupid entity on the planet.

在某些方面，也是這個星球上最愚蠢的實體。

The only entity that puts the very survival of the ecosystem in danger.

唯一會危及生態系統生存的實體。

So you think we are trading off more intelligence with more destructive risk?

所以，你認為我們是在用更多的破壞性風險來換取更多的情報？

Yes.

是的。

Again, it's not deterministic.

同樣，這也不是決定性的。

I don't think that we are doomed.

我不認為我們註定要失敗。

I mean, if I thought that, what's the point of talking about it if we can't prevent the worst case scenario?

我的意思是，如果我這麼想，如果我們不能防止最壞的情況發生，談論它還有什麼意義呢？

Well, I was hoping you thought you'd have some agency in actually effecting...

好吧，我希望你認為你在實際影響......

We still have agency.

我們仍然有代理權。

There are a few more years, I don't know how many, 5, 10, 30.

還有幾年，我不知道有多少年，5 年、10 年、30 年。

We still have agency.

我們仍然有代理權。

We are still the ones in the driver's seat shaping the direction this is taking.

我們仍然是主導方向的人。

No technology is deterministic.

任何技術都不是決定性的。

This is something, again, we learned from history.

這也是我們從歷史中學到的。

You can use the same technology in different ways.

你可以用不同的方式使用相同的技術。

You can decide which way to develop it.

您可以決定以哪種方式進行開發。

So we still have agency.

所以，我們仍然有代理權。

This is why you have to think very, very carefully about what we are developing.

這就是為什麼你必須非常、非常仔細地考慮我們正在開發的東西。

Well, thinking very carefully about it is something that Mustafa has been doing in this book.

穆斯塔法在這本書中一直在認真思考這個問題。

And I want to now go through some of the most commonly discussed risks.

現在我想介紹一些最常討論的風險。

I was trying to work out how I would go in sort of order of badness.

我在試著按照糟糕程度排序。

So I'm starting with one that is discussed a lot, but relative to human extinction is perhaps less bad, which is the question of jobs.

是以，我先從一個討論得很多，但相對於人類滅絕來說可能沒那麼糟糕的問題說起，那就是就業問題。

And will, you know, artificial intelligence essentially destroy all jobs because A.I.s will be better than humans at everything?

那麼，人工智能是否會因為在任何事情上都比人類做得好而從根本上毀掉所有工作呢？

You know, I'm an economist by training.

要知道，我可是受過專業訓練的經濟學家。

I, you know, history suggests to me that that has never happened before, that the lump of labor fallacy indeed is a fallacy.

我，你知道，歷史告訴我，這種情況以前從未發生過，勞動組合謬誤確實是一種謬誤。

But tell me what you think about that.

不過，請告訴我你是怎麼想的。

Do you think there is a risk to jobs?

您認為就業會有風險嗎？

It depends on the time frame.

這取決於時間框架。

So over a 10 to 20 year period, my intuition, and you're right that so far the evidence doesn't support this, is that there isn't really going to be a significant threat to jobs.

是以，在 10 到 20 年的時間裡，我的直覺是，就業不會受到重大威脅，你說得沒錯，目前的證據並不支持這一點。

There's plenty of demands.

要求很多。

There will be plenty of work.

工作會很多。

Right.

對

Over a 30 to 50 year time horizon, it's very difficult to speculate.

在 30 至 50 年的時間跨度內，很難進行推測。

I mean, at the very least, we can say that two years ago, we thought that these models could never do empathy.

我的意思是，至少我們可以說，兩年前，我們認為這些模型不可能做到移情。

We said that humans were always going to preserve kindness and understanding and care for one another as a special skill that humans have.

我們說過，人類總是會把善良、理解和相互關愛作為一種特殊的技能保留下來。

Four years ago, we said, well, AIs will never be creative.

四年前，我們說，人工智能永遠不會有創造力。

You know, humans will always be the creative ones inventing new things, making these amazing leaps between new ideas.

要知道，人類總是創造性地發明新事物，在新想法之間實現驚人的飛躍。

It's self-evident now that both of those two capabilities are things that these models do incredibly well.

現在不言而喻，這兩項功能都是這些機型做得非常出色的地方。

And so I think for a period of time, AIs augment our skills.

是以，我認為在一段時間內，人工智能會增強我們的技能。

They make us faster, more efficient, more accurate, more creative, more empathetic and so on and so forth.

它們讓我們更快、更高效、更準確、更有創造力、更有同理心，等等等等。

Over a many decade period, it's much harder to say what are the set of skills that are the permanent preserve of the human species, given that these models are clearly very, very capable.