達(dá)利歐：“自由意志將死,，AI意志將立”
Ray Dalio
2025-06-07
在我看來,，我們已經(jīng)到達(dá)了一個新時代的邊緣。在這個新時代,，機(jī)器思維將在很多方面補(bǔ)充甚至超越人類思維,，就像在工業(yè)時代，機(jī)器補(bǔ)充甚至超越了人類的勞動力那樣,。隨著計算工具的發(fā)明,，靠人腦來進(jìn)行數(shù)學(xué)運算和存儲記憶的重要性變得大大降低。現(xiàn)在我們查找信息會用谷歌這種搜索引擎,，而不是用更傳統(tǒng)的方式,。同樣，或許用不了多久,，當(dāng)我們在各種情境下需要進(jìn)行理性決策時,，我們也會尋求計算機(jī)的指引，因為它能給出比我們自己的大腦更快,、更好的指導(dǎo),。
未來五年，我們將會見證大多數(shù)領(lǐng)域的顯著進(jìn)步,。創(chuàng)造人工智能只是應(yīng)用人工智能的第一步,。我和橋水公司專業(yè)從事AI投資已經(jīng)有幾十年了,，我們現(xiàn)在正在發(fā)掘的機(jī)會，在普通人看來簡直難以置信,。人類靠自己的大腦做決策的時代即將終結(jié),。我和橋水公司的同事們就親身經(jīng)歷和見證了投資決策的智能化革命，因此我對即將到來的一切深感興奮,。
這些技術(shù)將影響到方方面面,，因此，無論是國家,、企業(yè)還是投資者,，只要能夠善于利用這些技術(shù)，就可以與其他國家和企業(yè)拉開很大差距,。換句話說,，善用AI者將如魚得水，不善于AI者將寸步難行,。而從投資的角度講,，相對于投資研發(fā)這些新技術(shù)所需的成本，我們目前尚不完全清楚將會有多少資金流入這個領(lǐng)域,。
中美兩國是這些高新技術(shù)研發(fā)的兩大主要競爭者,。雖然還有幾個國家也在研發(fā)這些技術(shù)，并得益于這些技術(shù),，但是中美兩國憑借各自強(qiáng)大的經(jīng)濟(jì)和軍事實力,，仍然在這個賽道上發(fā)揮著最重要的影響力。雖然美國在最先進(jìn)的半導(dǎo)體芯片的研發(fā)上領(lǐng)先于中國,，但是產(chǎn)能較弱,；中國在先進(jìn)芯片的研發(fā)上緊追美國，在成本較低的中低端芯片的生產(chǎn)上領(lǐng)先美國,，而且在AI的部署上也領(lǐng)先于美國,。在這場AI軍備競賽中，中美雙方肯定都會偷過竊取或借鑒對方的技術(shù)和做法來超越對方,。不過我個人始終有這樣一個看法——總的而言,，知識產(chǎn)權(quán)保護(hù)是沒有什么效果的。有些核心機(jī)密在付出巨大努力后可以一時保密（比如原子彈的研發(fā)）,，但是任何一種公開使用的技術(shù)基本上都很很快被復(fù)制,。另外，各國法律體系在知識產(chǎn)權(quán)保護(hù)的執(zhí)法上做得也差強(qiáng)人意,。因此,，我們有理由假設(shè)，任何公開展示的而且受到歡迎的好點子,，幾乎一定會在6個月內(nèi)被復(fù)刻,。
當(dāng)然,，人工智能也不是唯一一項能決定大國實力的重要技術(shù)。除了芯片和人工智能以外,，中美兩國在量子計算,、基因編輯等生物技術(shù)、機(jī)器人,、航空航天等領(lǐng)域也是實質(zhì)上的主要競爭對手,。全球40個頂尖的計算機(jī)科學(xué)項目有20個在中國，所以中國在技術(shù)競爭中是美國當(dāng)之無愧的強(qiáng)勁對手,。
總之,，對于那些富有創(chuàng)造力和實踐能力的人，能夠為他們提供必要的資源來研發(fā)新技術(shù)（特別是新的AI技術(shù)）,，并且讓他們在有利于進(jìn)步的環(huán)境中創(chuàng)業(yè),，這讓我感到非常興奮和樂觀。當(dāng)然,，新技術(shù)是一把雙刃劍,。它們既提升了我們造福彼此的能力,，也增強(qiáng)了我們互相傷害的手段,。
節(jié)選自雷·達(dá)里奧新書《國家如何破產(chǎn)》，本書由Simon & Schuster公司旗下的Avid Reader Press出版社于2025年6月3日出版發(fā)行,。
瑞?達(dá)利歐（Ray Dalio）是一位作家,、全球經(jīng)濟(jì)學(xué)家，也是全球最大對沖基金橋水聯(lián)合基金（Bridgewater Associates）的創(chuàng)始人,、首席投資官導(dǎo)師,。（財富中文網(wǎng)）
譯者：樸成奎
在我看來，我們已經(jīng)到達(dá)了一個新時代的邊緣,。在這個新時代,，機(jī)器思維將在很多方面補(bǔ)充甚至超越人類思維，就像在工業(yè)時代,，機(jī)器補(bǔ)充甚至超越了人類的勞動力那樣,。隨著計算工具的發(fā)明，靠人腦來進(jìn)行數(shù)學(xué)運算和存儲記憶的重要性變得大大降低?，F(xiàn)在我們查找信息會用谷歌這種搜索引擎,，而不是用更傳統(tǒng)的方式。同樣,，或許用不了多久,，當(dāng)我們在各種情境下需要進(jìn)行理性決策時，我們也會尋求計算機(jī)的指引,，因為它能給出比我們自己的大腦更快,、更好的指導(dǎo),。
未來五年，我們將會見證大多數(shù)領(lǐng)域的顯著進(jìn)步,。創(chuàng)造人工智能只是應(yīng)用人工智能的第一步,。我和橋水公司專業(yè)從事AI投資已經(jīng)有幾十年了，我們現(xiàn)在正在發(fā)掘的機(jī)會,，在普通人看來簡直難以置信,。人類靠自己的大腦做決策的時代即將終結(jié)。我和橋水公司的同事們就親身經(jīng)歷和見證了投資決策的智能化革命,，因此我對即將到來的一切深感興奮,。
這些技術(shù)將影響到方方面面，因此,，無論是國家,、企業(yè)還是投資者，只要能夠善于利用這些技術(shù),，就可以與其他國家和企業(yè)拉開很大差距,。換句話說，善用AI者將如魚得水,，不善于AI者將寸步難行,。而從投資的角度講，相對于投資研發(fā)這些新技術(shù)所需的成本,，我們目前尚不完全清楚將會有多少資金流入這個領(lǐng)域,。
中美兩國是這些高新技術(shù)研發(fā)的兩大主要競爭者。雖然還有幾個國家也在研發(fā)這些技術(shù),，并得益于這些技術(shù),，但是中美兩國憑借各自強(qiáng)大的經(jīng)濟(jì)和軍事實力，仍然在這個賽道上發(fā)揮著最重要的影響力,。雖然美國在最先進(jìn)的半導(dǎo)體芯片的研發(fā)上領(lǐng)先于中國,，但是產(chǎn)能較弱；中國在先進(jìn)芯片的研發(fā)上緊追美國,，在成本較低的中低端芯片的生產(chǎn)上領(lǐng)先美國,，而且在AI的部署上也領(lǐng)先于美國。在這場AI軍備競賽中,，中美雙方肯定都會偷過竊取或借鑒對方的技術(shù)和做法來超越對方,。不過我個人始終有這樣一個看法——總的而言，知識產(chǎn)權(quán)保護(hù)是沒有什么效果的,。有些核心機(jī)密在付出巨大努力后可以一時保密（比如原子彈的研發(fā)）,，但是任何一種公開使用的技術(shù)基本上都很很快被復(fù)制。另外,，各國法律體系在知識產(chǎn)權(quán)保護(hù)的執(zhí)法上做得也差強(qiáng)人意,。因此,，我們有理由假設(shè)，任何公開展示的而且受到歡迎的好點子,，幾乎一定會在6個月內(nèi)被復(fù)刻,。
當(dāng)然，人工智能也不是唯一一項能決定大國實力的重要技術(shù),。除了芯片和人工智能以外,，中美兩國在量子計算、基因編輯等生物技術(shù),、機(jī)器人,、航空航天等領(lǐng)域也是實質(zhì)上的主要競爭對手。全球40個頂尖的計算機(jī)科學(xué)項目有20個在中國,，所以中國在技術(shù)競爭中是美國當(dāng)之無愧的強(qiáng)勁對手,。
總之，對于那些富有創(chuàng)造力和實踐能力的人,，能夠為他們提供必要的資源來研發(fā)新技術(shù)（特別是新的AI技術(shù)）,，并且讓他們在有利于進(jìn)步的環(huán)境中創(chuàng)業(yè)，這讓我感到非常興奮和樂觀,。當(dāng)然,，新技術(shù)是一把雙刃劍。它們既提升了我們造福彼此的能力,，也增強(qiáng)了我們互相傷害的手段,。
節(jié)選自雷·達(dá)里奧新書《國家如何破產(chǎn)》,，本書由Simon & Schuster公司旗下的Avid Reader Press出版社于2025年6月3日出版發(fā)行,。
瑞?達(dá)利歐（Ray Dalio）是一位作家、全球經(jīng)濟(jì)學(xué)家,，也是全球最大對沖基金橋水聯(lián)合基金（Bridgewater Associates）的創(chuàng)始人,、首席投資官導(dǎo)師。（財富中文網(wǎng)）
譯者：樸成奎
It looks to me like we are now at the brink of a new era in which machine thinking will supplement or surpass human thinking in many ways, like how machine labor supplemented and surpassed human labor during the Industrial Revolutions. Just as we saw that doing math in our heads and remembering facts became much less important with the invention of computerized tools that do these things, and just as we have gone to Google (or its equivalents) to find information rather than gathering information in more traditional ways, we will soon be going to computers to get our instructions on what to do when we are in different situations because the computer will come up with better guidance more quickly than we can.
Over the next five years, we will see dramatic advancements in most areas. Creating the AI capabilities is just the beginning of the AI applications. I know that in my area of investing where I and Bridgewater have been doing AI investing through expert systems for decades, the opportunities that are being developed are nearly unbelievable. The days of people making decisions in their own heads are ending. I and others at Bridgewater have experienced and capitalized on this (r)evolution via the computerization of investment decision making, so I’m excited by what will be happening.
Because these technologies will impact almost everything, there will be exceptionally big differences between the performance levels of countries, investors, and companies who use them well. Those who know how to use these tools effectively will be rewarded, and those who fail to do so will be penalized. It is worth noting, however, that from an investment perspective, it is not totally clear how much money will come in relative to the costs that will go out to invest in and create these new technologies.
The US and China are now the main competitors in designing these powerful new technologies, and how effective they are will have big impacts on their economic and military powers, though several countries are also developing and benefiting from these new technologies. While the US is ahead of China in developing the most advanced semiconductor chips and weak in its production of them, China is close behind in the development of advanced chips, ahead in producing less advanced chips much less expensively, and ahead in deploying AI. There will certainly be a lot of effort from both sides to gain an advantage over the other in this race, both by stealing/borrowing what the other side has and trying to defend one’s own gains. I keep in the mind the principle that by and large, intellectual property protections don’t work. While deep secrets that are protected with great effort (like the development of the atomic bomb) might be able to be kept hidden, anything that is openly used can almost instantly be replicated. Also, legal systems do a poor job of enforcing intellectual property protections. For these reasons, we should assume that most good ideas that are openly shown and are liked a lot will be replicated in about six months.
I should also make clear that AI isn’t the only important technology shaping the relative power of nations. There are many technologies beyond chips and AI that the US and China are the main real competitors in, including quantum computing, gene editing and other biotech, robotics, space, etc. China, which is home to 20 of the 40 best computer science programs in the world, is a formidable adversary to the US in the technology competition.
In conclusion, I am very excited and optimistic about the revolutionary improvements that are likely to take place as the result of inventive/practical people being put together with capital that gets them the resources that they need (perhaps most importantly, these new AI technologies) and operating in great environments that are conducive to advancement. Of course, new technologies are double-edged swords. For example, they have advanced how we can do each other harm as well as how we can do each other good.
From How Countries Go Broke by Ray Dalio, published on June 3, 2025, by Avid Reader Press, a division of Simon & Schuster. Copyright ? 2025 by Ray Dalio.
Ray Dalio is an author, global economist, and founder and CIO mentor at Bridgewater Associates, the world’s largest hedge fund.
財富中文網(wǎng)所刊載內(nèi)容之知識產(chǎn)權(quán)為財富媒體知識產(chǎn)權(quán)有限公司及/或相關(guān)權(quán)利人專屬所有或持有,。未經(jīng)許可,，禁止進(jìn)行轉(zhuǎn)載、摘編,、復(fù)制及建立鏡像等任何使用,。
