And in our century, there's a new and yet more powerful one, computation.

而在我們這個世紀，又出現了一種新的、更強大的計算。

For nearly 50 years, I've had the great privilege of building up an ever taller tower of science and technology that's based on that idea of computation.

近 50 年來，我有幸在計算理念的基礎上建立起一座越來越高的科技之塔。

And so today, I want to tell you a little bit about what that's led to.

今天，我想向大家介紹一下這一切的結果。

There's a lot to talk about, so I'm going to go quickly, and sometimes I'm going to summarize in a sentence what I've written a whole book about.

要談的東西很多，所以我會很快，有時我會用一句話概括我寫了一整本書的內容。

But, you know, I last gave a TED talk 13 years ago in February 2010, soon after Wolfram Alpha launched.

但是，你知道，我上一次發表 TED 演講還是 13 年前的 2010 年 2 月，當時 Wolfram Alpha 剛剛推出不久。

And I ended that talk with a question.

最後，我問了一個問題。

The question was, is computation ultimately what's underneath everything in our universe?

問題是，我們宇宙中的萬事萬物最終都是由計算支撐的嗎？

I gave myself a decade to find out.

我給了自己十年的時間去尋找答案。

And actually, it could have needed a century, but in April 2020, just after the decade mark, we were thrilled to be able to announce what seems to be the ultimate machine code of the universe.

實際上，這可能需要一個世紀的時間，但在 2020 年 4 月，也就是十年大關剛過之時，我們非常高興地宣佈，這似乎是宇宙的終極機器代碼。

And yes, it's computational.

是的，這是計算。

So computation isn't just a possible formalization, it's the ultimate one for our universe.

是以，計算不僅僅是一種可能的形式化，它還是我們宇宙的終極形式化。

It all starts from the idea that space, like matter, is made of discrete elements.

這一切都源於這樣一個想法：空間和物質一樣，是由離散的元素構成的。

And from that structure of space and everything in it, it's defined just by a network of relations between these elements that we might call atoms of space.

從空間結構和其中的一切來看，它只是由這些我們可以稱之為空間原子的元素之間的關係網絡所定義的。

So it's all very elegant, but deeply abstract.

是以，這一切都非常優雅，但又非常抽象。

But here's kind of a humanized representation, a version of the very beginning of the universe, and what we're seeing here is the emergence of space and everything in it by the successive application of very simple computational rules.

但這裡是一種人性化的呈現，是宇宙誕生之初的一個版本，我們在這裡看到的是通過連續應用非常簡單的計算規則而產生的空間和其中的一切。

And remember, these dots are not atoms in any existing space, they're atoms of space that get put together to make space.

請記住，這些點並不是任何現有空間中的原子，它們是空間的原子，被組合在一起形成空間。

And yes, if we kept going long enough, we could build our whole universe this way.

是的，如果我們堅持足夠長的時間，我們就能以這種方式構建我們的整個宇宙。

So eons later, here's a chunk of space with two little black holes that, if we wait a little while, will eventually merge, generating little ripples of gravitational radiation.

是以，數億年後，這裡有一大塊空間，其中有兩個小黑洞，如果我們再等一會兒，它們最終會合並，產生引力輻射的小漣漪。

And remember, all of this is built from pure computation.

請記住，這一切都建立在純粹的計算基礎之上。

But like fluid mechanics emerging from molecules, what emerges here is space-time and Einstein's equations for gravity, though there are deviations that we just might be able to detect, like that the dimensionality of space won't always be precisely three.

但是，就像從分子中產生的流體力學一樣，這裡出現的是時空和愛因斯坦的萬有引力方程，儘管我們可能會發現一些偏差，比如空間的維度並不總是精確的三維。

And there's something else.

還有一件事。

Our computational rules can inevitably be applied in many ways, each defining a different kind of thread of time, a different path of history that can branch and merge.

我們的計算規則不可避免地會以多種方式應用，每種規則都定義了不同的時間線，不同的歷史路徑，可以分支和合並。

But as observers embedded in this universe, we're branching and merging too, and it turns out that quantum mechanics emerges as the story of how branching minds perceive a branching universe.

但是，作為嵌入這個宇宙的觀察者，我們也在分支和合並，而事實證明，量子力學的出現就是分支思維如何感知分支宇宙的故事。

So the little pink lines you might be able to see here show the structure of what we call branchial space, the space of quantum branches.

是以，你可能會在這裡看到一些粉紅色的線條，它們顯示了我們所說的分支空間的結構，即量子分支空間。

And one of the stunningly beautiful things, at least for physicists like me, is that the same phenomenon that in physical space gives us gravity, in branchial space gives us quantum mechanics.

至少對我這樣的物理學家來說，其中一個令人驚歎的美妙之處在於，在物理空間中產生萬有引力的現象，在分支空間中也產生了量子力學。

So in the history of science so far, I think we can identify sort of four broad paradigms for making models of the world that can be distinguished kind of by how they deal with time.

是以，在迄今為止的科學史上，我認為我們可以找出四種建立世界模型的廣泛範式，這些範式可以通過如何處理時間來加以區分。

So in antiquity, and in plenty of areas of science even today, it's all about kind of what are things made of, and time doesn't really enter.

是以，在古代，甚至在今天的許多科學領域，都是關於事物是由什麼構成的，時間並沒有真正進入其中。

But in the 1600s came the idea of modeling things with mathematical formulas in which time enters, but basically just as a coordinate value.

但到了 1600 年代，人們開始想到用數學公式來模擬事物，在公式中加入時間，但基本上只是作為一個座標值。

Then in the 1980s, and this is something in which I was deeply involved, came the idea of making models by starting with simple computational rules and just letting them run.

20 世紀 80 年代，我深入參與了這一想法，即從簡單的計算規則入手，讓它們運行，從而製作模型。

So can one predict what will happen?

那麼，我們能預測會發生什麼嗎？

No.

不

There's what I call computational irreducibility, in which, in effect, the passage of time corresponds to an irreducible computation that we have to run in order to work out how it will turn out.

這就是我所說的計算的不可還原性，實際上，時間的流逝對應著一種不可還原的計算，我們必須運行這種計算才能知道結果如何。

But now there's kind of something even more.

但現在有了更多的東西。

In our physics project, there's things that become multi-computational with many threads of time that can only be knitted together by an observer.

在我們的物理項目中，有一些事情變得多計算，有許多時間線，這些時間線只能由觀察者編織在一起。

So it's kind of a new paradigm that actually seems to unlock things not only in fundamental physics, but also in the foundations of mathematics and computer science, and possibly in areas like biology and economics as well.

是以，它是一種新的範式，似乎不僅能解開基礎物理學的謎團，還能解開數學和計算機科學的謎團，甚至可能解開生物學和經濟學等領域的謎團。

So, you know, I talked about building up the universe by repeatedly applying a computational rule.

所以，你知道，我說過通過反覆應用計算規則來構建宇宙。

But how is that rule picked?

但這條規則是如何選出來的呢？

Well, actually, it isn't, because all possible rules are used, and we're building up what I call the Rulliard, the kind of deeply abstract but unique object that is the entangled limit of all possible computational processes.

實際上並非如此，因為所有可能的規則都被使用了，我們正在建立我所說的 "魯利亞爾"，一種深度抽象但又獨一無二的對象，它是所有可能的計算過程的糾纏極限。

Here's a tiny fragment of it, shown in terms of Turing machines.

下面是它的一個小片段，用圖靈機表示。

So this Rulliard is everything.

所以這個魯利亞德就是一切。

And we, as observers, are necessarily part of it.

而我們，作為觀察者，必然是其中的一部分。

In the Rulliard as a whole, in a sense, everything computationally possible can happen.

從某種意義上說，在整個魯利亞德，一切計算上可能發生的事情都可能發生。

But observers like us just sample specific slices of the Rulliard.

但像我們這樣的觀察者只是對魯利亞德的特定片段進行採樣。

And there are two crucial facts about us.

關於我們，有兩個至關重要的事實。

First, we're computationally bounded.

首先，我們的計算能力有限。

Our minds are limited.

我們的思維是有限的。

And second, we believe we're persistent in time, even though we're made of different atoms of space at every moment.

其次，我們相信我們在時間上是永恆的，儘管我們每時每刻都由不同的空間原子組成。

So then, here's the big result.

那麼，最重要的結果來了。

What observers with those characteristics perceive in the Rulliard necessarily follows certain laws, and those laws turn out to be precisely the three key theories of 20th century physics.

具有這些特徵的觀察者在魯里亞中感知到的東西必然遵循某些規律，而這些規律恰恰就是 20 世紀物理學的三大關鍵理論。

General relativity, quantum mechanics, and statistical mechanics and the second law.

廣義相對論、量子力學、統計力學和第二定律。

So it's because we're observers like us that we perceive the laws of physics we do.

所以，正是因為我們是像我們這樣的觀察者，我們才會感知到我們所感知到的物理定律。

We can think of sort of different minds as being at different places in Rullial space.

我們可以把不同的思維看作是在魯里亞空間的不同位置。

Human minds who think alike are nearby, animals further away, and further out we get to kind of alien minds where it's hard to make a translation.

思維相似的人類思維在附近，動物思維在更遠的地方，再遠一點我們就會遇到外星思維，很難進行翻譯。

So how can we get intuition for all of this?

那麼，我們怎樣才能獲得這一切的直覺呢？

Well, one thing we can do is use generative AI to take what amounts to an incredibly tiny slice of the Rulliard aligned with images we humans have produced.

那麼，我們可以做的一件事就是利用生成式人工智能，從魯利亞德的影像中提取出與我們人類製作的影像相一致的極小部分。

We can think of this as sort of a place in the Rulliard described by using the concept of a cat in a party hat.

我們可以用 "戴著派對帽子的貓 "這個概念，把這裡看成是魯利亞爾描述的一個地方。

So, zooming out, we have, we saw there, what we might call Cat Island.

是以，我們把視線拉遠，就能看到我們可以稱之為 "貓島 "的地方。

Pretty soon we're kind of in interconcept space.

很快，我們就進入了概念間空間。

Occasionally things will look familiar, but mostly what we'll see is things we humans don't have words for.

偶爾我們會看到熟悉的事物，但大多數情況下，我們看到的都是人類無法用語言描述的事物。

In physical space, we explore the universe by sending out spacecraft.

在物理空間，我們通過發射航天器探索宇宙。

In Rullial space, we explore more by expanding our concepts and our paradigms.

在 Rullial 空間，我們通過擴展我們的概念和範式來進行更多探索。

We can kind of get a sense of what's out there by sampling possible rules, doing what I call Rulliology.

我們可以通過對可能的規則進行抽樣調查，來了解現有的規則，我稱之為 "規則學"（Rulliology）。

So, even with incredibly simple rules, there's incredible richness.

是以，即使規則簡單得令人難以置信，但內容卻豐富得令人難以置信。

But the issue is that most of it doesn't yet connect with things we humans understand or care about.

但問題是，其中大部分內容還與我們人類理解或關心的事物無關。

It's like when we look at the natural world and only gradually realize that we can use features of it for technology.

這就像我們在觀察自然界時，才逐漸意識到我們可以利用自然界的特徵來發展科技。

So, even after everything our civilization has achieved, we're just at the very, very beginning of exploring Rullial space.

是以，即使我們的文明已經取得了如此巨大的成就，我們也只是剛剛開始探索魯里亞空間。

What about AIs?

人工智能呢？

Well, just like we can do Rulliology, AIs can, in principle, go out and explore Rullial space.

好吧，就像我們可以做 Rulliology 一樣，人工智能原則上也可以出去探索 Rullial 空間。

Left to their own devices, though, they'll mostly just be doing things we humans don't connect with or care about.

不過，如果任由它們自生自滅，它們大多隻會做一些我們人類不會聯繫或關心的事情。

So, the big achievements of AI in recent times have been about making systems that are closely aligned with us humans.

是以，近代人工智能的重大成就就是製造出與我們人類緊密結合的系統。

We train LLMs on billions of web pages so they can produce text that's typical of what we humans write.

我們通過數十億個網頁來訓練 LLM，這樣它們就能寫出與我們人類所寫內容相同的文本。

And yes, the fact that this works is undoubtedly telling us some deep scientific things about the semantic grammar of language and generalizations of things like logic that perhaps we should have known centuries ago.

是的，這一事實無疑告訴了我們一些關於語言語義文法和邏輯等事物概括的深奧科學知識，也許我們早在幾個世紀前就應該知道了。

You know, for much of human history, we were kind of like the LLMs, figuring things out by kind of matching patterns in our minds.

要知道，在人類歷史的大部分時間裡，我們就像法學碩士一樣，通過腦中的模式匹配來解決問題。

But then came more systematic formalization and eventually computation.

但隨後出現了更系統的形式化，並最終實現了計算。

And with that, we got a whole other level of power to truly create new things and to, in effect, go wherever we want in the Rulliad.

有了它，我們就有了另一種力量，可以真正創造出新的東西，實際上，在《魯利亞德》中，我們想去哪裡就去哪裡。

But the challenge is to do that in a way that connects with what we humans and our AIs understand.

但我們面臨的挑戰是，如何將我們人類和人工智能的理解聯繫起來。

In fact, I've devoted a large part of my life to kind of trying to build that bridge.

事實上，我一生中很大一部分時間都在努力搭建這座橋樑。

It's all been about creating a language for expressing ourselves computationally, a language for computational thinking.

這一切都是為了創造一種計算表達語言，一種計算思維語言。

The goal is to formalize what we know about the world in computational terms, to have computational ways to represent cities and chemicals and movies and formulas and our knowledge about them.

我們的目標是將我們對世界的瞭解用計算術語形式化，用計算的方式來表示城市、化學物質、電影和公式以及我們對它們的瞭解。

It's been a vast undertaking that's spanned more than four decades of my life, but it's something very unique and different.

這是一項跨越我四十多年人生的浩大工程，但它又是非常獨特和與眾不同的。

But I'm happy to report that in what has been Mathematica and is now the Wolfram Language, I think we've firmly succeeded in creating a truly full-scale computational language.

但我很高興地告訴大家，在 Mathematica 和現在的沃爾夫拉姆語言中，我認為我們已經成功地創造了一種真正全面的計算語言。

In effect, every one of these functions here can be thought of as formalizing and encapsulating in computational terms some facet of the intellectual achievements of our civilization.

實際上，這裡的每一個功能都可以被視為用計算術語形式化和概括了我們文明智力成就的某些方面。

It's sort of the most concentrated form of intellectual expression that I know, sort of finding the essence of everything and coherently expressing it in the design of our computational language.

這是我所知道的最集中的智力表達方式，它找到了萬事萬物的本質，並在我們的計算語言設計中將其連貫地表達出來。

For me personally, it's been an amazing journey, kind of year after year, building the sort of tower of ideas and technology that's needed and nowadays sharing that process with the world in things like open live streams and so on.

對我個人來說，這是一段奇妙的旅程，年復一年，我建立了所需的思想和技術之塔，如今又通過公開直播等方式與世界分享這一過程。

A few centuries ago, the development of mathematical notation and what amounts to the language of mathematics gave a systematic way to express math and made possible algebra and calculus and eventually all of modern mathematical science.

幾個世紀前，數學符號和數學語言的發展為數學提供了一種系統的表達方式，使代數和微積分成為可能，並最終使所有現代數學科學成為可能。

And computational language now provides a similar path, letting us ultimately create a computational X for all imaginable fields, X.

現在，計算語言也提供了類似的途徑，讓我們最終為所有可以想象的領域 X 創造出一個計算 X。

I mean, we've seen the growth of computer science, CS, but computational language opens up something ultimately much bigger and broader, CX.

我的意思是，我們已經看到了計算機科學、CS 的發展，但計算語言開創的東西最終要大得多、廣得多，CX。

I mean, for 70 years we've had programming languages which are about telling computers in their terms what to do, but computational language is about something intellectually much bigger.

我的意思是，70 年來，我們一直在使用編程語言，用它們的術語告訴計算機該做什麼，但計算語言涉及的是智力上更大的東西。

It's about taking everything we can think about and operationalizing it in computational terms.

這就是把我們所能想到的一切，用計算的方式操作出來。

You know, I built the Wolfram language first and foremost because I wanted to use it myself, and now when I use it, I feel like it's kind of giving me some kind of superpower.

你知道，我創建 Wolfram 語言首先是因為我想自己使用它，現在當我使用它時，我感覺它給了我某種超能力。

I just have to imagine something in computational terms and then the language sort of almost magically lets me bring it into reality, see its consequences and build on them.

我只需要用計算術語來想象一些東西，然後這種語言就會幾乎神奇地讓我把它帶入現實，看到它的後果，並在此基礎上繼續發展。

And yes, that's the sort of superpower that's let me do things like our physics project.

沒錯，就是這種超能力讓我完成了我們的物理項目。

And over the past 35 years, it's been my great privilege to share this superpower with many other people, and by doing so, to have enabled an incredible number of advances across many fields.

在過去的 35 年裡，我非常榮幸地與許多其他人分享了這一超能力，並是以在許多領域取得了令人難以置信的進步。

It's sort of a wonderful thing to see people, researchers, CEOs, kids, using our language to fluently think in computational terms, kind of crisping up their own thinking and then in effect automatically calling in computational superpowers.

看到人們、研究人員、首席執行官、孩子們使用我們的語言流暢地用計算術語進行思考，這真是一件美妙的事情。

And now it's not just people who can do that.

現在不只是人可以這樣做了。

AIs can use our computational language as a tool too.

人工智能也可以使用我們的計算語言作為工具。

Yes, to get their facts straight, but even more importantly, to compute new facts.

是的，是為了弄清事實，但更重要的是，是為了計算新的事實。

There are already some integrations of our technology into LLMs.

我們的技術已經集成到了 LLM 中。

There's a lot more you'll be seeing soon.

你很快就會看到更多。

And you know, when it comes to building new things in a very powerful emerging workflow, it's basically to start by telling the LLM roughly what you want, then to have it try to express that in precise Wolfram language, then, and this is a critical feature of our computational language compared to, for example, a programming language, you as a human can read the code.

要知道，在一個非常強大的新興工作流程中構建新事物時，基本上是先告訴 LLM 你大致想要什麼，然後讓它嘗試用精確的 Wolfram 語言表達出來，然後，與編程語言相比，這是我們計算語言的一個關鍵特徵，即人類可以讀取代碼。

And if it does what you want, you can use it as kind of a dependable component to build on.

如果它能滿足你的要求，你就可以把它作為一個可靠的組件來使用。

Okay, but let's say we use more and more AI, more and more computation.

好吧，但假設我們使用越來越多的人工智能，越來越多的計算。

What's the world going to be like?

世界會變成什麼樣？

From the Industrial Revolution on, we've been used to doing engineering where we can effect, see how the gears mesh to understand how things work.

從工業革命開始，我們就習慣於做工程，在工程中我們可以看到齒輪是如何齧合的，從而瞭解事物是如何工作的。

But computational irreducibility now shows us that that won't always be possible.

但計算的不可還原性現在告訴我們，這並不總是可能的。

We won't always be able to make a kind of simple human or say mathematical narrative to explain or predict what a system will do.

我們並不總能用一種簡單的人類或數學敘述來解釋或預測一個系統會做什麼。

And yes, this is science in effect eating itself from the inside.

是的，這實際上是科學在從內部吞噬自己。

From all the successes of mathematical science, we've come to believe that somehow, if we only could find them, there'd be formulas to kind of predict everything.

從數學科學的所有成功中，我們開始相信，只要我們能找到它們，就會有公式來預測一切。

But now computational irreducibility shows us that that isn't true.

但現在，計算的不可還原性告訴我們，事實並非如此。

And that in effect, to find out what a system will do, we have to go through the same irreducible computational steps as the system itself.

實際上，要想知道一個系統會做什麼，我們必須經歷與系統本身相同的不可還原的計算步驟。

Yes, it's a weakness of science.

是的，這是科學的弱點。

But it's also why the passage of time is significant and meaningful.

但這也是為什麼時間的流逝意義重大。

And why we can't just sort of jump ahead to get the answer.

為什麼我們不能跳到前面去找答案？

We have to live the steps.

我們必須按部就班。

It's actually going to be, I think, a great societal dilemma of the future.

我認為，這將是未來社會的一大難題。

If we let our AIs achieve their kind of full computational potential, they'll have lots of computational irreducibility, and we won't be able to predict what they'll do.

如果我們讓人工智能充分發揮其計算潛能，它們就會有很多計算上的不可還原性，而我們將無法預測它們會做什麼。

But if we put constraints on them to make them more predictable, we'll limit what they can do for us.

但是，如果我們對它們施加限制，讓它們變得更可預測，我們就會限制它們為我們所做的事情。

So what will it feel like if our world is full of computational irreducibility?

那麼，如果我們的世界充滿了計算的不可還原性，會是什麼感覺呢？

Well, it's really nothing new, because that's the story with much of nature.

其實，這並不新鮮，因為大自然中的很多東西都是這樣。

And what's happened there is that we've found ways to operate within nature, even though nature can sometimes still surprise us.

儘管大自然有時仍會給我們帶來驚喜，但我們已經找到了在大自然中運作的方法。

And so it will be with the AIs.

人工智能也將如此。

We might give them a constitution, but there will always be consequences we can't predict.

我們可以給他們一部憲法，但總會有我們無法預料的後果。

Of course, even figuring out societally what we want from the AIs is hard.

當然，即使從社會角度來考慮我們對人工智能的需求也很困難。

Maybe we need, you know, a promptocracy where people write prompts instead of just voting.

也許我們需要一個提示民主制，讓人們寫提示，而不僅僅是投票。

But basically, every control-the-outcome scheme seems full of both political philosophy and computational irreducibility gotchas.

但從根本上說，每一種 "控制結果 "方案似乎都充滿了政治哲學和計算不可還原性的缺陷。

You know, if we look at the whole arc of human history, the one thing that's systematically changed is that more and more gets automated.

要知道，縱觀整個人類歷史，有一件事一直在發生變化，那就是自動化程度越來越高。

And LLMs just gave us a dramatic and unexpected example of that.

而法學碩士剛剛給了我們一個戲劇性的、意想不到的例子。

So what does that mean?

這意味著什麼？

Does that mean that in the end, us humans will have nothing to do?

這是否意味著最終我們人類將無所事事？

Well, if we look at history, what seems to happen is that when one thing gets automated away, it opens up lots of new things to do.

如果我們回顧一下歷史，就會發現，當一件事情被自動化之後，就會有很多新的事情可以做。

And as economies develop, the pie chart of occupations seems to get more and more fragmented.

而隨著經濟的發展，職業的餅狀圖似乎越來越分散。

And now we're back to the Rouillade, because at a foundational level, what's happening is that automation is opening up more directions to go in the Rouillade.

現在我們又回到了 "魯瓦德"，因為在基礎層面上，自動化正在為 "魯瓦德 "開闢更多的發展方向。

But there's no abstract way to choose between these.

但這兩者之間並沒有抽象的選擇方式。

It's a question of what we humans want, and it requires kind of humans doing work to define that.

這是一個我們人類想要什麼的問題，需要人類通過工作來確定。

So a society of AIs sort of untethered by human input would effectively go off and explore the whole Rouillade, but most of what they do would seem to us random and pointless, much like most of nature doesn't seem to us right now like it's achieving a purpose.

是以，一個不受人類影響的人工智能社會將有效地去探索整個魯瓦德，但在我們看來，它們所做的大部分事情都是隨機的、毫無意義的，就像我們現在看來，大部分自然界都沒有達到目的一樣。

I mean, one used to imagine that to build things that are useful to us, we'd have to do it kind of step by step.

我的意思是，人們曾經認為，要製造出對我們有用的東西，我們必須一步一步地去做。

But AI and the whole phenomenon of computation tell us that really what we need is more just to define what we want.

但是，人工智能和整個計算現象告訴我們，我們真正需要的只是定義我們想要的東西。

Then computation, AI, automation can make it happen.

那麼計算、人工智能和自動化就能實現這一點。

And yes, I think the key to defining in a clear way what we want is computational language.

是的，我認為明確定義我們想要什麼的關鍵在於計算語言。