2026年科技股还能入手吗？耶鲁芝大管院前院长×科技投资人的

2026年科技股还能入手吗？耶鲁芝大管院前院长×科技投资人的年终盘点来了

发布于 2025-12-26 · 时长 1:02:46

字幕摘录

时间	英文	中文
0:04	Welcome, everybody, to webinar number five, where we will sum up what are the key events from 2025 and look forward for 2026.	欢迎各位来到第五网络,我们将在这里总结2025年的关键事件,展望2026年。
0:16	I see we have 70 people.	看来我们有70个人
0:18	Well, Logan, Ted, welcome.	洛根泰德欢迎
0:20	Thank you, Jakob.	谢谢你,雅各布。
0:21	Thanks, Jakob.	谢谢贾可布.
0:22	Good morning.	早上好,你好吗?
0:23	On behalf of Yale Center Beijing, I just want to say welcome, everyone, and to Logan waking up at 430 a.m. in Silicon Valley, to Ted on the U.S. East Coast, and to Jakob.	我代表耶鲁中心北京,只想表示欢迎各位,欢迎洛根于凌晨4时30分在硅谷醒来,欢迎美国东海岸的泰德和雅各布.
0:40	Thank you very much for bringing this super popular webinar that you have done for the Global Network for Advanced Management to Yale Center Beijing and to our audiences, as well as the SOM alumni and friends audience.	非常感谢你们给耶鲁中心北京和我们的观众,以及SOM校友和朋友的观众带来了这个超级受欢迎的网络网络。
1:01	This has garnered a lot of attention, tens of thousands of views each time, and we can't thank you enough to bring rigor and analysis to one of the hottest topic du jour, the economics of high tech industries.	这引起了很多人的注意, 每次都有上万种观点, 我们无法感谢你把严谨和分析带到最热门的话题之一, 高科技产业的经济学。
1:22	And with both of you having a wealth of experience on the arguably the two biggest economies and countries that matter in this space, U.S. and China.	你们俩对这两个最大的经济体和在这个空间重要的国家,美国和中国,有着丰富的经验。

展开字幕全文（705 条）

序号	英文	中文
1	Welcome, everybody, to webinar number five, where we will sum up what are the key events from 2025 and look forward for 2026.	欢迎各位来到第五网络,我们将在这里总结2025年的关键事件,展望2026年。
2	I see we have 70 people.	看来我们有70个人
3	Well, Logan, Ted, welcome.	洛根泰德欢迎
4	Thank you, Jakob.	谢谢你,雅各布。
5	Thanks, Jakob.	谢谢贾可布.
6	Good morning.	早上好,你好吗?
7	On behalf of Yale Center Beijing, I just want to say welcome, everyone, and to Logan waking up at 430 a.m. in Silicon Valley, to Ted on the U.S. East Coast, and to Jakob.	我代表耶鲁中心北京,只想表示欢迎各位,欢迎洛根于凌晨4时30分在硅谷醒来,欢迎美国东海岸的泰德和雅各布.
8	Thank you very much for bringing this super popular webinar that you have done for the Global Network for Advanced Management to Yale Center Beijing and to our audiences, as well as the SOM alumni and friends audience.	非常感谢你们给耶鲁中心北京和我们的观众,以及SOM校友和朋友的观众带来了这个超级受欢迎的网络网络。
9	This has garnered a lot of attention, tens of thousands of views each time, and we can't thank you enough to bring rigor and analysis to one of the hottest topic du jour, the economics of high tech industries.	这引起了很多人的注意, 每次都有上万种观点, 我们无法感谢你把严谨和分析带到最热门的话题之一, 高科技产业的经济学。
10	And with both of you having a wealth of experience on the arguably the two biggest economies and countries that matter in this space, U.S. and China.	你们俩对这两个最大的经济体和在这个空间重要的国家,美国和中国,有着丰富的经验。
11	So with that, I'm really happy to be doing this last session for the year 2025, and we look forward to more.	因此,我很高兴能参加2025年的最后一次会议,我们期待着更多的会议。
12	We'll have a bit of a survey later on the call that Jakob will queue up to see what kind of interest we may have for keeping this going in a sustainable way.	我们稍后会调查一下雅各布会排队调查我们可能有什么兴趣以可持续的方式保持下去
13	So, you know, without further ado, Togan, we'll hand it over to you.	所以,你知道,没有进一步的ado,托根, 我们会把它交给你。
14	Thank you, Carol.	谢谢你,卡罗尔。
15	And you touched on one of the themes that as we look back on 2025, we really did focus on the G2, China and the U.S.	你谈到了一个主题, 当我们回顾2025年时, 我们确实专注于G2,中国和美国。
16	And of course, we focused on high tech and we feel like we got that right.	当然,我们专注于高科技, 我们觉得我们是对的。
17	So, Jakob, maybe you could put up the opening slide.	所以,Jakob, 也许你可以放开幻灯片。
18	As Carol did say, this is a look back at 2025 and looking ahead to 2026.	正如卡罗尔所说,这是回顾2025年,展望2026年.
19	Next slide.	下一个幻灯片。
20	And I'm joined, of course, by Logan Bender.	当然,我是洛根·班德
21	Logan, it's been a fun journey.	洛根,这旅程很愉快
22	We keep learning.	我们继续学习。
23	We keep trying to figure stuff out.	我们一直在想办法
24	Thank you for being so engaged and supportive of everyone in the community.	感谢大家如此参与和支持社区中的每一个人。
25	I hope you're doing well.	希望你表现的很好
26	Thank you, Ted, for the intro.	谢谢你泰德做介绍
27	And thank you, Carol, again.	再次谢谢你卡罗尔
28	Carol has been an excellent host.	卡罗尔是个出色的主持人
29	Her team has been awesome.	她的团队很出色
30	YCB, we're really excited about these lectures.	YCB,我们很兴奋这些讲座。
31	We try to do the best and put the best content forward.	我们努力做到最好,提出最好的内容。
32	So let's jump right in.	我们跳进去吧
33	Ted, go ahead and kick it off, please.	Ted 请继续踢吧
34	Yeah, we couldn't do it without you, Carol.	是啊没有你我们办不到卡罗尔
35	Again, thank you and your team.	再次感谢你和你的团队。
36	Amazing.	厉害
37	So the agenda today is do a quick look at the performance of some Chinese and U.S. stocks and then consistent with what our practice has been over the over the years.	因此,今天的议程是快速审视一些中国和美国股票的表现,然后与我们多年来的做法保持一致。
38	Let's talk about a current event, which is the Trump administration decision to authorize sale of NVIDIA's H200 chips to Chinese companies.	谈目前事件,即特朗普政府决定授权将NVIDIA的H200芯片出售给中国公司.
39	And then we are going to look at what we got right back in the previous webinars.	然后我们再看看我们在之前的网络研讨会上得到了什么。
40	And patting ourselves on the back, I'd say we got a lot right, not everything.	拍拍自己的背部,我会说,我们有很多是正确的,不是一切。
41	And then looking forward, what what do we think you should focus on in terms of big issues?	然后展望未来,我们认为在重大问题方面,你应该专注于什么?
42	And obviously, whether the U.S. is in the A.I.	显然,美国是否在A. I.
43	bubble is a big one.	泡泡是一个大的。
44	I think a yet bigger picture issue that I want to really talk about is are the U.S. and China moving towards divergent A.I.	我想真正想谈的一个更大的问题,是美国和中国走向不同的人工智能。
45	ecosystems or are they going to converge?	生态系统还是它们要汇合?
46	I think it's a big issue for not just our our discussion today, but not just for China and the U.S., but really for humanity.	我认为,这不仅是我们今天讨论的大问题,而且不仅是中国和美国讨论的大问题,而且真正是人类讨论的大问题.
47	Which country will be more supportive of A.I.?	哪个国家会更支持人工智能?
48	Which which of the two economies is more vulnerable?	两个经济体中哪一个比较脆弱?
49	And then as a favorite theme of Logan's, how how should we expect private and public markets to interact?	作为洛根最喜欢的主题我们如何期待私人和公共市场互动?
50	So that's our agenda.	这就是我们的议程。
51	Jakob, do you want to just cover the first one?	Jakob,你想要第一个吗?
52	How what do we see here with some Chinese and U.S. companies?	我们如何看待中国和美国的公司?
53	The key insight is that the main Chinese stocks have performed very well this year to the surprise of Western investors and media.	关键见解是,今年中国主要股票业绩非常好,令西方投资者和媒体感到惊讶.
54	And as you can see in the chart, we put the year to date increases in stock value for these particular companies.	正如你们在图表中看到的,我们把这些特定公司的股票价值的年增长率计算出来。
55	The Chinese tech giants have overperformed outperformed their main counterparts in the U.S.	中国科技巨头业绩超标超越美国主要同行.
56	Logan, back to you.	罗根回到你身边
57	What are you?	你是谁?
58	Yes.	对
59	So if we rewind to 2024 and 2023, there was some depression and valuation among some Chinese companies.	因此,如果我们回溯到2024年和2023年,一些中国公司出现了一些萧条和估价.
60	There was concern, growth concerns.	人们对增长感到关切。
61	There was regulatory concerns.	存在监管问题。
62	We've seen more positive outlook among both U.S. investors that invest in Asia and Asian investors.	在投资亚洲的美国投资者和亚洲投资者中,我们看到了更积极的前景.
63	And so we've seen the rally in the Chinese counterparts, the Nasdaq.	因此,我们已经看到集会在中国同行,纳斯达克。
64	As a reminder, Ted and I really like to track both ecosystems on East and West.	作为提醒,泰德和我真的很喜欢跟踪东西方的生态系统。
65	We feel that's very important to represent the global investor base, but not a lot of investors cover both.	我们认为,代表全球投资者基础是非常重要的,但并没有很多投资者涵盖两者.
66	So we like to offer insight into both ecosystems, which is why we bring charts of this nature up.	因此,我们想提供两种生态系统的见解, 这就是为什么我们提出这种性质的图表。
67	And we've seen other companies.	我们见过别的公司
68	Amazon's been relatively flat, even though they've had a lot of exciting growth initiatives.	亚马逊相对平坦,尽管他们有许多刺激性的增长举措.
69	Oracle's been up and down.	甲骨文虽上下.
70	We'll get into volatility later, but we've seen a lot of volatility in the U.S. tech space.	我们以后会遇到波动,但我们在美国的技术空间中已经看到很多波动.
71	You know, big wins, big losses, big news, you know, regulatory issues.	你知道,大赢家,大损失, 大新闻,你知道,监管问题。
72	And so there's a lot there we'll unpack.	所以有很多东西要收拾
73	But here's a performance here today.	但今天这里有个表演
74	Ted, if you don't have any comment, we'll move on to the next slide.	泰德,如果你没有任何评论, 我们将转到下一个幻灯片。
75	Well, the performance of the Chinese companies has really been impressive.	中国公司的表现真令人印象深刻.
76	We should note that the Chinese tech stocks lagged U.S. tech stocks in previous years.	需要指出的是,中国科技储量在前几年落后于美国科技储量.
77	So this is a reversal.	因此,这是一个逆转。
78	And we know that stock prices reflect expectations about the future.	我们知道股票价格反映了对未来的期望。
79	So this is a significant change.	因此,这是一个显著的变化。
80	Logan, you mentioned volatility and, you know, your connection with Professor Schiller.	罗根,你提到了波动还有你和席勒教授的关系
81	This, you know, Professor Schiller, Nobel laureate at Yale, what he has emphasized to the world is that stock prices may be efficient.	这,你知道,席勒教授,耶鲁的诺贝尔奖得主, 他向全世界强调,股票价格可能是有效的。
82	We may not be able to outguess the market, but there's a lot of volatility.	我们也许无法超越市场,但波动很大。
83	And I think that's noteworthy here.	我认为这值得注意。
84	So let's go on to the next slide.	因此,让我们继续下一张幻灯片。
85	What did we get right?	我们得到了什么正确的?
86	Logan, you want to take the first two?	洛根,你想拿前两个?
87	Yeah, let me take the first two.	是啊,让我拿前两个。
88	So one major theme.	所以有个主题。
89	And this was debated in the Wall Street community, both on the sell side being the researchers and then on the investor side	华尔街社区对此进行了辩论卖家和投资者都是研究者
90	and among retail investors and academics like ourselves is what's going to happen every quarter?	而像我们这样的零售投资者和学者中,每个季度都会发生什么?
91	We've seen CapEx, which is capital expenditure.	我们已经看到CapEx,这是资本支出。
92	So companies and we're picking on the hyperscalers or the large cloud giants, the leading tech firms in the U.S. and Asia.	因此,公司和我们正在挑选超规模或大型云层巨头, 美国和亚洲的主要技术公司。
93	What are they spending every quarter on their capital expenditure?	它们每个季度的资本支出都花多少?
94	Mostly data center build out, buying GPUs, things of that nature, semiconductor related to power and grow the AI initiatives.	主要是数据中心的建设,购买GPU,这种性质的事物,与动力有关的半导体,并成长AI倡议.
95	And we'll look at a chart going back to 2020 in a little bit.	我们再看看2020年的图表
96	But the main idea is that when we got to the beginning of this year, 2025, that CapEx train had taken off.	但主要的想法是,当我们到达今年2025年初时,CapEx列车已经起飞.
97	The trend had inflected.	这一趋势有所反映。
98	This was into the tens of billions of dollars every quarter.	这是进入数百亿美元每个季度。
99	It seemed like these amounts were getting almost too big to stomach from a risk perspective in terms of deploying capital.	从部署资本的风险角度来看,这些数额似乎变得几乎太大,无法承受。
100	And we our view was that the risk of underinvestment and not being at the table to have a voice for the future of	我们认为,投资不足的风险,以及不能在谈判桌上对未来发表意见的风险,
101	AI in the tech ecosystem would be far too risky as opposed to trying to defend your capital expenditure decision.	在技术生态系统里,AI会太冒险而不是试图捍卫你的资本支出决定。
102	And we were right about that.	我们是对的。
103	We've seen record investment that's influenced other parts of the ecosystem, which we'll discuss later.	我们看到了影响生态系统其他部分的创纪录投资,我们稍后再讨论。
104	But record CapEx did continue.	但记录CapEx确实继续.
105	We believe it will continue in 2026 again for similar reasons.	我们认为,出于同样的原因,这一对话将在2026年再次继续下去。
106	And then could I interrupt?	我可以打断一下吗?
107	I mean, these companies, I think you're pointing out something profound, but it's also kind of sensical.	我的意思是,这些公司, 我认为你指出的东西深刻,但它也有点敏感。
108	They're looking at the risks of underinvestment and overinvestment and they see asymmetry.	他们正在研究投资不足和过度投资的风险,他们看到不对称.
109	They don't want to be left behind.	他们不想被留下
110	They don't want to be left without a seat at the table.	他们不想在桌子上没有座位
111	Having said that, and they get into these heavy CapEx, you'll come to this later.	说完这番话,他们又进入了这种重的CapEx,你以后会来看的.
112	But I just wanted to foreshadow that one reaction or corollary to that is you want to try to de-risk your investments.	但我只是想预示一种反应或必然结果就是你想尝试去冒险投资
113	And we're seeing a lot more partnerships and investments, reciprocal investments among these big spenders.	我们看到了更多的伙伴关系和投资, 这些大支出者之间的互惠投资。
114	So keep that in mind.	所以记住这一点。
115	Yes.	对
116	Yeah, thank you.	谢谢
117	Yeah.	对
118	We'll look at the circularity of investment and possible implications and into that.	我们将研究投资的循环性和可能的影响并由此而来。
119	But Ted's absolutely right.	但泰德绝对是对的
120	And I think in the back of your mind as well, we're going to hit on private markets a little bit less than some of our lectures.	我想在你的脑海中, 我们将会在私人市场打点比一些讲座少一点。
121	But in the private space, these companies, not only are they spending record amounts on CapEx for their own core business lines or new business lines,	但是在私人空间,这些公司, 他们不仅在CapEx上花费了创纪录的金额他们自己的核心业务线或新的业务线,
122	but they're also investing heavily in the venture ecosystem.	但他们也在对冒险生态系统进行大量投资。
123	They're doing corporate venture deals.	他们正在做公司风险交易。
124	They're doing other partnerships with private companies at a scale that we have never seen before.	他们与私人公司建立了其他伙伴关系,其规模我们从未见过。
125	And that was breaking a record that we saw in 2024 about allocation of VC dollars in the ecosystem.	这打破了2024年我们在生态系统中看到的维也纳公约美元的分配记录。
126	So again, it comes down to the idea of are you missing a megatrend if you don't invest?	再说一遍,如果你不投资你是否错过了大趋势?
127	It's easier to defend the decision of, hey, we spent too much money, but hey, look at what we could potentially gain from it.	我们花了太多钱但看看我们能从中得到什么
128	We'll talk about revenue traction in a minute.	我们一会儿再谈收入吸引问题
129	And I'll hit on number two that ties a little bit into number one is as those expectations have grown, valuations have followed in the public space.	我将关注第二点, 与第一点有关联的是随着这些期望的增长, 在公共空间里, 估价也随之而来。
130	And somewhat in the private space, too, we've seen inflection valuations, larger rounds of capital have been raised.	而在私人空间里,我们也看到过虚构的估价, 已经筹集了更多的资本。
131	OpenAI, Ted, I was reading as we were getting ready for this lecture, I think two days ago, an article from the Journal.	OpenAI, Ted, 我读了一下, 当我们正在准备这个讲座, 我想两天前, 一篇来自《日刊》的文章.
132	And we printed in our current events, I think in 2024, that OpenAI hit a valuation of $80 billion.	我认为,在2024年, 我们的当前事件显示, OpenAI的估值高达800亿美元。
133	And now it's $500 billion a year later.	今则一年五千亿.
134	I mean, these are mind blowing sums of money.	我的意思是,这些是心吹的钱。
135	And so valuations, we thought valuations would be pressured in the public space to go up and they did absolutely.	因此,我们觉得估值在公共空间会受到压力, 要求它们上台。
136	Though we've had volatility, they have risen.	尽管我们有波动, 他们已经上升。
137	And I'll bounce it back to you for three.	我会把它还给你三个。
138	Go ahead on three.	继续数到三
139	So just an aside, what did I get wrong?	所以,只是一旁, 我得到了什么错误?
140	I thought Amazon would do well.	我以为亚马逊会做得很好
141	They're a laggard in terms of valuations of U.S. major tech companies.	就美国主要技术公司的估值而言,他们是落后者。
142	But turning to number three, we stated and we remain of the following view.	但是,谈到第三,我们指出,我们仍持以下观点。
143	The semiconductor industry is the most important industry in the world.	半导体工业是世界上最重要的工业.
144	It's where all the action is in terms of investment in data centers, how to develop AI capabilities.	这是所有动作在数据中心投资, 如何开发AI能力。
145	It's also, of course, so important to the fate of China and the U.S.	当然,这也对中国和美国的命运如此重要.
146	So I think it was correct to focus both on the industry and on those two countries.	所以我认为,既关注工业,也关注这两个国家,这是正确的。
147	Forgive me for apparently not saying anything great about the rest of the world.	原谅我显然没有说任何伟大的世界其他地方。
148	But I think objectively, it's really China and the U.S.	但我客观地认为,真的是中国和美国.
149	We're not trying to ignore Europe.	我们不想忽视欧洲
150	We're not trying to ignore South Asia, Latin America, et cetera.	我们并不想忽视南亚,拉丁美洲等地区.
151	But it's really a sort of a heavyweight battle right now between China and the U.S.	但它确实是一种重量级的战斗现在在中国和美国之间。
152	And as noted here, we've also been correct to say, OK, we have to realize this competition is in the context of rising global tensions.	正如这里所指出,我们也正确地说, 好了,我们必须认识到,这场竞争是在全球紧张局势加剧的背景下进行的。
153	And we'll talk about that in a minute when we get to this current event.	我们等会再谈这个,当我们到达这个活动。
154	But I think we allocated our time well, and I hope that the audience has benefited from our focus on semiconductor, the G2, rising global tensions.	但我认为我们分配的时间是好的,我希望观众从我们关注半导体,即G2,全球紧张局势加剧中受益。
155	And then just note, we also stated along the way, if you were to pick one company in the world that was sort of at the epicenter of everything, we'd say it's TSMC.	顺便说一句,我们一路都说, 如果你在世界上挑选一个公司, 它在一切的中心,我们会说它是TSMC。
156	It's not the most valuable company, but it's the one that's really caught in these major trends in terms of investment in AI, CapEx and rising global tensions.	它并不是最有价值的公司,而是在AI,CapEx和不断上升的全球紧张局势方面真正陷入这些主要趋势的一方.
157	So and then Logan has really kept me up to date on private markets.	然后Logan一直让我了解私人市场的最新情况
158	Maybe you would cover the last topic and why we think it's important.	也许你会讨论最后一个话题以及我们为什么认为这很重要
159	Yes, no problem.	没问题
160	So paying attention to private markets.	所以关注私人市场
161	This really struck me in 2025 in particular when we looked at all the venture dollars that were summed up essentially and deployed into the U.S.	这真的让我印象深刻, 2025年,特别是当我们审视了基本上归纳出来并投入美国的所有风险基金。
162	venture ecosystem, AI related or AI specific bets, I guess, and then ancillary.	冒险生态系统,AI相关或AI特定赌注,我想,然后是附属的.
163	So hardware and software, different AI applications swallowed up over half the venture dollars in the U.S.	因此,硬件和软件,不同的AI应用吞噬了美国一半以上的风险美元.
164	And I think about a third of it in 2024.	我想到了2024年的三分之一。
165	And these are record sums of money.	这些是创纪录的钱
166	These are not trivial sums.	这些不是微不足道的金额。
167	These are in the hundreds of billions of dollars of capital.	这些都是千亿元的资本。
168	And so you have seen an extraordinary bet on a wager on the growth of AI being the next mega trend.	所以,你已经看到一个非凡的赌注一个赌注,AI的成长是下一个巨大的趋势。
169	And that has led to huge valuation increases in the private markets.	这导致了私人市场价值的大幅上升。
170	So much so that some of these private companies have gotten so big that they are starting to influence quite a bit the policies and operational decisions of public companies,	以至于一些私营公司变得如此庞大以至于它们开始影响公共公司的政策和经营决定
171	which is kind of flips the script from from before, where you had the S&P 500 and the tech giants they would not talk down to,	那种翻转脚本从之前, 你在那里有S&P 500和科技巨头他们不会说话下来,
172	but they would collaborate on their terms with private companies that were later stage or emerging competitors.	但是,它们将根据其条件与属于后期或新兴竞争者的私营公司合作。
173	And that's not necessarily the case anymore.	情况也不一定是这样了
174	In fact, you see private companies that are they would they would say that they are ahead or they are leading or they are the dominant firm for now.	事实上,你看到的是私营公司,它们会说它们领先,或者它们领先,或者它们现在是主导企业。
175	And Ted and I are very careful about using the term dominant in tech because that never really seems to be truly dominant long term.	Ted和我非常小心地使用在技术中占据支配地位的术语, 因为从长远来看,这似乎从来没有真正占据支配地位。
176	But you get my point.	但你明白我的观点
177	And so I think that it's important to understand how much how much ballast the private markets have right now in the tech ecosystem.	所以我认为,重要的是要了解私人市场在技术生态系统中有多少压载物。
178	It's more than we've seen in a long time.	况于久之所见.
179	So we'll conclude on that point.	因此,我们将得出这一点的结论。
180	I think unless you had another comment, we'll move on.	我想除非你另有评论,否则我们就继续前进.
181	Now, I think we should go to some of the proof points that we think are important on the global tensions.	现在,我认为我们应该谈谈我们认为对全球紧张局势很重要的一些证据。
182	I'll just draw your attention to the lower left hand headline.	我会提醒你注意左手下部的头条
183	The restrictions on the sale of Nvidia chips has led to a lot of tensions contributed to the global tensions.	对Nvidia芯片销售的限制导致许多紧张局势,助长了全球紧张局势.
184	It's also consistent with basic economics.	这也符合基本经济学.
185	If you put in a restriction, there are going to be efforts to evade the restriction.	如果你施加限制, 将会有人试图逃避限制。
186	So this particular headline claimed the story claim that there was a billion dollars worth of Nvidia chips being moved into China despite the ban.	因此,这个特别的头条声称这个故事声称,有价值10亿美元的Nvidia芯片尽管受到禁令却被搬进中国.
187	But let's go on to the next one.	但复说下一部.
188	This is a point for Logan's point.	这是罗根的观点。
189	And maybe you could talk through this slide.	也许你可以通过这个幻灯片说话。
190	This is one of our favorites.	这是我们最喜欢的之一。
191	Yeah, this is a slide.	是的,这是一个幻灯片。
192	We pulled this from the Wall Street Journal.	我们从《华尔街日报》找到的
193	And we've pulled this slide as we've gone on in the class that Ted and I taught at Yale, which is looking at the capital expenditure, which is what we referenced earlier.	我们在Ted和我在耶鲁大学教的课上完成了这个幻灯片, 研究了资本支出, 我们之前提到的就是这个。
194	This is CapEx by quarter for the above four names.	上面四个名字是CapEx的四分之一
195	我们有Meta,Alphabet,微软和亚马逊。
196	If we were to include Oracle and some other companies in there, we would see this obviously increase.	如果我们把甲骨文和其他一些公司都包括进去,我们就会看到这种明显的增长。
197	But this shows you the steady march in CapEx spend every quarter it goes up.	但是这显示了CapEx的稳步行进每四分之一都涨价
198	And if you look at the scale on the left axis, you see this is eclipsing 50 billion dollars per quarter.	如果你看看左轴上的尺度, 你就会看到这每季度有500亿美元被挤压。
199	These are huge sums of money.	这些是一大笔钱
200	And let's also remember.	让我们也记住。
201	And the reason why I want to hammer this home is because the revenue realized from these initiatives is much less than the spend.	我之所以想敲打这个家是因为这些倡议带来的收入远远低于支出。
202	We have not seen any of these companies achieve a positive ROI relative to the amount of CapEx that they've deployed.	我们没有看到任何这些公司取得了积极的ROI 相对于他们部署的CapEx的数量。
203	And so this is on the hope that this will be.	因此,这是希望这将是。
204	Yes, Andy Tan.	是的谭安迪
205	Yes, thank you.	是的,谢谢你。
206	That's a lot of money.	谓多财.
207	That's absolutely correct.	没错
208	That's what we're saying.	这就是我们所说的。
209	And but the revenue realized from these this spend hasn't been there hasn't been as much traction as some people hope, which we will then talk about in the potential a bubble.	但从这些支出中获得的收入并没有像有些人所希望的那样多, 然后我们再讨论潜在的泡沫。
210	If you could move to the next slide, please, we'll take another look at this from another dimension.	如果你可以移动到下一个幻灯片,请, 我们将从另一个维度再看一遍。
211	Let's put this in perspective.	让我们把这个角度。
212	These tech giants in this basket of companies are investing 40 plus percent of their revenues of their revenues on CapEx related to semiconductors.	这些技术巨头在这个篮子公司正在投资40+他们收入的% 与半导体有关的CapEx。
213	And so as Ted mentioned, TSMC and the conversation on the G2, the near and medium future is hinging on the semiconductor industry.	正如Ted所提到的,TSMC和G2上的对话, 近中未来是兴起半导体工业。
214	It's the most important.	这是最重要的。
215	I believe China is going to announce their five year plan.	我相信中国会宣布他们的五年计划
216	And let's also understand that semiconductors not only do they impact the AI, but they impact biotech.	我们还要明白,半导体不仅影响人工智能,而且影响生物技术。
217	They impact quantum computing.	它们影响量子计算.
218	They impact automotive.	它们影响汽车。
219	Ted and I had a lecture a couple of months ago about how much technology is is hinging independent on semiconductor content.	Ted和我几个月前曾做过一个讲座关于半导体内容方面有多少科技在独立运行
220	And we've seen the inflection really kick off in 2023.	2023年,我们目睹了这场战争的爆发。
221	And we've seen three record years of CapEx.	我们见过三年创纪录的CapEx
222	And then let's take a quick peek at 2026.	2026年我们再仔细看看
223	Next slide, if you could, please.	下一个幻灯片,如果你可以,请。
224	Let's look at 2026.	让我们看看2026年。
225	And so this is a little bit of an eye test.	所以这是一点点的眼睛测试。
226	This expands the basket.	这会扩大篮子。
227	CoreWeave is a data center company that has come to IPO fairly recently.	CoreWeave是一家数据中心公司,最近才来到IPO。
228	And so these are what are called hyperscalers, largest kind of cloud giants in the US tech basket.	这些是所谓的超规模的, 美国科技篮子中最大的云层巨头。
229	When we look at 2026, we see this is consensus estimate about a continuation of this CapEx trend getting getting even bigger.	当我们看2026年时,我们发现,这是关于这种CapEx趋势继续扩大的共识估计。
230	So as much money as they have spent year to date, quarter to date, last year, the year before, there's still spending.	因此,他们花费的金钱,如每年至今,季度至今,去年,前一年,仍然有开支.
231	And we believe we are with consensus in this way that we believe that CapEx will increase in 2026 and it won't slow down materially.	我们相信我们以这种方式达成共识,相信CapEx将在2026年增加,不会在物质上放慢速度.
232	And I think that wraps up that section.	我认为这段结束了。
233	说吧 Ted
234	Logan, we infer, even though we don't have the data on Chinese spend, we infer that the Chinese are spending as much or more.	洛根,我们推断,虽然我们没有关于中国开支的数据,但我们推断中国人的支出是同样或更多的.
235	Yes.	对
236	And why do we?	为什么?
237	That's right.	没错
238	You'll see this later in one of Logan's slides.	你以后会在洛根的幻灯片里看到这个
239	The the amount of power generation capacity that China has put in place.	中国已建立的发电能力量.
240	It's it's now growing at a really fast rate.	现在它正在以非常快的速度增长。
241	Total capacity in China far exceeds the US capacity.	中国总产能远远超过美国产能.
242	What is that capacity being used for?	这种能力是用来做什么的?
243	We believe it's for data centers, but we don't have data by company.	我们相信这是为数据中心, 但我们没有公司的数据。
244	So that's why we're not showing a companion slide here.	这就是为什么我们不展示一个同伴幻灯片在这里。
245	But this is not this is clearly not just a US phenomenon.	但这显然不仅仅是一个美国现象。
246	Yes.	对
247	All right.	好吧,我们走。
248	Let's go to the next slide.	下一张幻灯片
249	Ted, why don't you kick this one off?	泰德,你为什么不踢这个?
250	This is this is current event of the day.	这是今天的事件。
251	Yeah.	对
252	So so I called out that that headline about a billion worth of Nvidia chips being smuggled in.	所以我叫出了那条头条价值十亿的Nvidia芯片被走私进来
253	But this is the US has pursued what some people in the US call a containment strategy,	但这就是美国所追求的美国有些人称之为遏制战略,
254	where they've tried to limit China's development of leading edge technologies, especially those related to AI.	他们试图限制中国领先技术的发展,尤其是那些与AI有关的技术.
255	How have they done that?	他们怎么做到的?
256	By picking out Nvidia chips and saying,	通过挑选Nvidia芯片并说,
257	oh, those are the things that really influence whether the US has a lead, whether China has a lead.	哦,这些都是真正影响美国是否有领先, 中国是否有领先。
258	And just in the last 10 days, the Trump administration has reversed that containment policy.	而就在最近10天,特朗普政府扭转了这一遏制政策.
259	And it announced that the US would, going forward, allow Nvidia to sell their H 200 chips to Chinese companies.	并宣布美国将允许Nvidia将他们的H200芯片卖给中国公司。
260	So this is a significant shift from prior policy Trump consistent with what he does a lot.	因此,这与先前的政策特朗普符合他所做的很多事情,是一个重大的转变.
261	He said we want to take a piece of this.	他说我们要拿一块这个
262	So there's in effect a company specific product specific export tax on these on these chip sales.	所以实际上公司对这些芯片销售征收特定产品出口税。
263	But Logan, maybe you could try to try to put this into perspective.	但Logan 也许你可以试着把事情看清楚
264	Where where do H 200 chips fall in terms of Nvidia's current product portfolio and future product portfolio?	从Nvidia目前的产品组合和未来的产品组合来看,H 200芯片在哪里落下?
265	Yes.	对
266	So the H 200 is not Nvidia's leading edge semiconductor, meaning what's coming in 2026 and 2027.	因此H 200型并非恩维迪亚的领先边缘半导体,意思是2026年和2027年的来历.
267	This was launched in 2024, maybe late 2023.	这部作品于2024年推出,可能晚于2023年.
268	And so an H 200 is a name of a chip.	所以一个H200是芯片的名字。
269	But what the Trump administration or the US is saying is that they are concerned about a performance threshold	但特朗普政府或美国所说的话是他们关心的是业绩门槛
270	that they're allowing China to use or allowing Nvidia to export to China.	他们允许中国使用或允许Nvidia出口到中国。
271	And so that's the concern is the amount of power that they're giving them that these chips have.	这就是人们所关心的, 也就是他们给他们的能量, 这些芯片有。
272	That doesn't mean that China can't buy a lot of them.	这并不意味着中国买不到很多.
273	But the H 200, one could argue, since it's not the leading node, it's not the most powerful chip to come.	但H200,人们可以争论,因为它不是领先的节点,也不是最强大的芯片来袭.
274	Is it OK to give it to China is kind of the counterbalance to that conversation.	把它交给中国,是不是可以算是与那个对话的平衡。
275	Instead of saying, well, we're giving away everything or giving away the farm,	而不是说,嗯,我们放弃一切或放弃农场,
276	and then we're going to be outcompeted because we're giving away our best chips and our best in breed horse.	然后我们就会失去能力因为我们把最好的筹码和最好的品种马
277	Not necessarily the case.	不一定是这样
278	This is not this is not the leading edge.	这不是领先优势
279	And so I think it's important to to note that.	所以我认为应该注意这一点。
280	And on the last point, it's it's potentially a huge boon for Nvidia.	最后一点,它可能给Nvidia带来巨大的好处。
281	Their stock price rose immediately after, not surprisingly.	他们的股票价格随后立即上涨,并不奇怪。
282	But there is some doubt about whether Chinese companies will be allowed to purchase these chips and which Chinese companies in which volumes.	但对于是否允许中国公司购买这些芯片,以及哪些中国公司有多少批量,还存在一些疑问.
283	And Xi has said he's been he's been very clear he wants the progress in China to be built on Chinese foundation, not the US foundation.	而习近平表示,他一直非常明确地希望中国的进步建立在中国的基础上,而不是美国的基础.
284	So there's I think there's uncertainty about what this really will mean in practice.	所以我认为这实际上意味着什么是不确定的。
285	But let's you know, I've been following the US reaction to this and the US reaction is negative.	但让我们知道,我一直在关注美国对此的反应,美国的反应是负面的.
286	And here here are just a couple of examples of US media saying this decision is wrong.	本文仅举几个美国媒体的例子,
287	And The Wall Street Journal, New York Times rarely agree on things like that, but they do on this issue.	而"华尔街日报","纽约时报"也很少同意这样的事情,但是他们在这个问题上确实同意.
288	They basically say that giving this allowing the purchase of these chips will give an advantage to China that they lack because the H 200 chips are better than leading Chinese chips.	他们基本上说,给予这种允许购买这些芯片的做法,会给中国带来他们缺乏的优势,因为H200芯片比领先的中国芯片要好.
289	That's an important fact.	这是一个重要的事实。
290	They're better.	他们更好。
291	So as Logan said, the performance of these chips is better and that allow applications to develop in China that otherwise would not develop.	所以,正如洛根所说,这些芯片的性能更好,并且允许在中国开发否则不会开发的应用程序.
292	And that that concern if you read the US policy commentary carefully, a lot of people who share this consensus negative view are focused on military, national security, data security issues.	如果你仔细阅读美国的政策评论,那么,许多赞同这种共识的负面观点的人,都专注于军事,国家安全,数据安全问题.
293	So that's basically the negative consensus.	所以,这基本上是消极的共识。
294	Let's go to the next slide.	下一张幻灯片
295	But I I I'm not going to try to convince you whether the US consensus is right or wrong.	但我不打算说服你美国的共识是对是错
296	But I think Logan and I would like to lay out an alternative view of this and and critical thinking is needed here.	但我认为,洛根和我愿意对此提出另一种看法,在此需要审慎的思考。
297	And I think it begins with a basic couple of basic facts about chips.	我认为这是从一些关于芯片的基本事实开始的。
298	Nobody consumes chips.	没人会吃薯片
299	They buy chips, but chips are purchased because they're used in the context of hardware and software to produce services.	他们购买芯片,但芯片是购买的,因为它们被用于硬件和软件的背景中来生产服务.
300	So Nvidia makes well, they design these chips.	因此,Nvidia做的很好, 他们设计这些芯片。
301	TSMC is their major supplier of these chips.	TSMC是这些芯片的主要供应商.
302	But Nvidia is not just a supplier of chips.	但Nvidia不仅仅是一个芯片供应商。
303	They're they're not just a hardware company.	他们不只是一家硬件公司
304	They're not just even not just a hardware company, not just a software company.	他们不仅仅是个硬件公司不仅仅是个软件公司
305	They're they they develop ecosystems.	它们是它们开发生态系统。
306	So Logan, maybe you could explain a little bit about the so-called software stack and maybe the hardware stack so that we can get a deeper appreciation of this point.	所以Logan,也许你可以解释一下所谓的软件堆栈和硬件堆栈这样我们就能更深刻地理解这一点。
307	Yeah.	对
308	So, yes, for for context.	所以,是的,为上下文。
309	And so we look at a company like Nvidia, which is essentially an IP designer.	所以我们看看像Nvidia这样的公司, 它基本上是一个IP设计师。
310	By the way, they have a marvelous venture portfolio.	顺便说一句,他们有一个了不起的冒险投资组合。
311	They've invested in a lot of interesting technologies.	他们投资了许多有趣的技术.
312	So I think Nvidia is branching out.	所以,我认为Nvidia正在分化。
313	Obviously, the robotics is phenomenal.	显然,机器人是惊人的。
314	But I think what's missed on the news when you read articles and things like this and sometimes in the analyst community, people often miss that Nvidia's programming language, CUDA, their CUDA ecosystem has grown a generation of engineers, not one or two	但我认为,当你读到这样的文章和东西时,有时在分析界,人们常常会想念恩维迪亚的编程语言CUDA,他们的CUDA生态系统已经发展了一代工程师,而不是一两年,或者软件更新或者补丁之类的.
315	That there are hundreds of thousands, millions, in fact, I think Jensen referenced in the millions of engineers that use the CUDA ecosystem.	事实上,有成千上万,数百万人, 我认为詹森在数百万使用CUDA生态系统的工程师中提到了。
316	And so that's a much more broad set of individuals working for interesting firms all over the world, US, Europe, Asia, etc.	全世界,美国,欧洲,亚洲等地的有趣公司都有一大群人
317	And so we don't want to constrict our audience to think that Nvidia is just the hardware that they're designing and then TSMC is printing out.	所以我们不想限制我们的观众认为Nvidia只是他们设计的硬件然后TSMC打印出来。
318	It's actually an entire ecosystem of software as well, Ted.	这也是整个软件的生态系统泰德
319	So it's it's an important understanding to have that context.	所以有这种背景是一个重要的理解.
320	So the economics here would emphasize the complementary investments that users make in a hardware stack, software stack.	因此这里的经济学将强调用户在硬件堆栈,软件堆栈中所做的互补投资.
321	And the most, I think, subtle question is, how does this affect the development of the Chinese ecosystem?	我认为,最微妙的问题是,这如何影响中国生态系统的发展?
322	Will the Nvidia chips be used and embedded in the Chinese ecosystem?	Nvidia芯片是否会被使用并嵌入中国生态系统?.
323	Or will there be a tendency to also see that the ecosystem in the US around these complementary investments will play a greater role in China because of this policy decision?	或者还会有一种倾向,看到美国围绕这些互补投资的生态系统会因为这一政策决定在中国发挥更大的作用?
324	And the counter argument to the US consensus being negative is by giving access to these chips, we're going to encourage investments,	美国共识的反面论点是提供这些芯片我们鼓励投资
325	complementary investments in software and hardware that match and relate to the US ecosystem.	对符合和涉及美国生态系统的软件和硬件进行补充投资。
326	So if you think about this not as a battle over who's got the leading chips, but who's developing the better ecosystem,	所以如果你不认为这是一场战斗谁有领先的筹码, 但谁正在开发更好的生态系统,
327	what we what we need to understand is that there are transitory issues here.	我们需要了解的是,这里存在着过渡性问题。
328	Yeah, we're definitely giving China the opportunity to buy better chips.	是啊,我们肯定给中国机会买更好的筹码。
329	But we're also maybe over the long term encouraging Chinese investment in complementary products, software and hardware that will sustain US leadership in the overall AI space.	但我们也可能长期鼓励中国在辅助产品、软件和硬件上投资,
330	That's that's the counter argument.	这就是反驳。
331	And another counter argument is we know China is working diligently.	另一个反驳论点是,我们知道中国正在勤奋工作。
332	They're allocating massive amounts of financial capital and resources to developing chips that are better.	他们将大量金融资本和资源用于开发更好的芯片。
333	And I can't predict that.	我无法预测
334	I can't predict their their path of success.	我不能预测他们的成功之路。
335	Deep Seek has said Chinese chips are definitely behind the US chips.	深探说中国芯片绝对是美国芯片背后的
336	But a question that I think is relevant to this policy question is how long would the advantage last for the US if we deny chips,	但我认为与这个政策问题相关的一个问题是如果我们拒绝筹码美国的优势会持续多久
337	try to put a clamp on all the smuggling, that's going to encourage greater Chinese efforts to develop chips that match and potentially leapfrog and exceed the capabilities of Nvidia chips.	这将会鼓励中国加大研发芯片的力度使其与Nvidia芯片相匹配并有可能跳跃并超越其能力
338	So I don't think this is a easy analysis, but simply for the for the US naysayers, the people who are negative to say this is obviously bad, I think it misses a lot of these important points.	因此我认为这不是一个简单的分析, 但对于美国否认者来说,
339	All right, let's let's move on to 2026.	好吧,让我们继续2026年。
340	I think this is the first one is on everybody's mind, right, Logan?	我想这是每个人心中的第一个对吧,洛根?
341	That's right.	没错
342	So everybody wants to talk about today.	所以大家今天都想谈谈
343	Let me let me start it off by saying, given what Logan said, we got these massive investments by hyperscalers in the US.	首先让我说,鉴于洛根的话, 我们得到了美国超大规模投资。
344	We have to ask the question, will they pay off?	我们必须问这个问题,他们会还清吗?
345	And to underscore this question, for those of you who followed Logan and my teachings over the last several years, there's a phenomenon called hockey stick high tech companies.	为了强调这个问题, 对于你们当中那些在过去几年里遵循洛根教义的人来说, 有一个叫做冰球棒高科技公司的现象。
346	And usually we think of the hockey stick profile as being for startups.	通常我们会认为冰球棒的简介是初创企业。
347	And the idea of the hockey stick is the company's startup.	而"曲棍球棒"的构思则是公司的启动.
348	They spend a lot on capital.	他们在资本上花了很多钱
349	They make a lot of investments in their products and services.	它们对其产品和服务进行大量投资。
350	So they have negative cash flows.	所以他们有负现金流。
351	They have negative profits.	他们有负利润。
352	And the idea of the hockey stick is hopefully that downward negative profit trend gets reversed as you get into the sales and get penetration into the accessible market with the result that the profits go from negative to not getting as negative to flatte	而曲棍球棒的想法是,希望随着你进入销售并进入可及的市场,负利润趋势得到扭转,结果利润从负增长到不至于负增长到扁平增长,然后出现。
353	Why do we bring this up?	我们为什么要提这个?
354	I think I think this what Logan said is really important.	我觉得洛根的话很重要
355	The capital expenditures by these US firms are so high.	这些美国公司的资本支出如此之高。
356	They're far exceeding what they're bringing in now in terms of return on investment and revenue generation.	从投资回报和创造收入的角度来说,它们已经远远超过了它们现在带来的结果。
357	Hence, it's correct in our view to think of these US hyperscalers as having within their product portfolios, a startup, a massive startup that's a lot like a hockey stick profile.	因此,我们认为把这些美国超规模机看作是在它们的产品组合, 一个启动,一个巨大的启动, 这就像一个曲棍球棒的简介。
358	Now, that means that means a lot of things.	意思是,这意味着很多事情。
359	But one of one of the things that investors are concerned about how many of these top US companies will make the turn.	但投资者关心的一件事是这些顶尖的美国公司中有多少会转弯。
360	And for that line of business, if you want to think about it that way, how many of them will actually be able to make the turn into profitability for that segment, that all important segment of their business?	对于这一行来说,如果你想这样思考的话, 他们中有多少人能够把这个部分变成利润, 也就是他们所有的重要部分?
361	That, I think, is what's behind all these concerns about is the US in a quote AI bubble?	我认为,这些担心背后的是什么美国在引用AI泡沫?
362	Yes.	对
363	All right.	好吧,我们走。
364	So let's jump into this.	所以,让我们跳到这个。
365	Can we please go one slide?	我们可以走一个幻灯片吗?
366	Let's look at the 2026 items just so the viewers can get a scope of what we're going to cover today.	让我们看看2026项只是为了让观众得到今天要覆盖的范围。
367	So we'll do the AI bubble in just a second here.	我们马上来做人工智能泡沫
368	We're going to discuss Ted's going to weigh in on this as well.	我们讨论一下 Ted也会参与其中
369	Are we moving towards a divergent ecosystem or are we converging into a primary ecosystem?	我们是走向不同的生态系统,还是正在汇合成一个主要的生态系统?
370	Country, which country would be more supportive of AI development?	哪个国家会更支持人工智能的发展?
371	That's top of mind as well and important, especially in the news these days.	这同样也是头等大事,特别是在现在的新闻中.
372	And Ted's going to lead which of the G2 economies is more vulnerable?	Ted要领导哪个G2经济体更脆弱?
373	All this being said, and then I'll lead and Ted will comment on the roles of public and private markets as we step into 2026.	所有这一切,然后我带领和泰德将评论公共和私人市场的作用当我们进入2026年。
374	Sound good?	听起来好吗?
375	All right.	好吧,我们走。
376	Let's go.	我们走吧。
377	Jakub, go ahead and go to the next slide, please.	贾库布,请上下一张幻灯片
378	Pause here for just a second.	暂停一下
379	No, on the on the title slide.	不,在标题幻灯片上。
380	So one more slide back.	再来一次
381	Yes, that one, please.	是的,这个,请。
382	All right.	好吧,我们走。
383	So let's think about let's just frame this for a second.	因此,让我们想想,让我们只是设定这一秒钟。
384	So there's a lagging statistic, but an important statistic that we put in one of our lectures that the energy consumption of the state of Virginia,	因此,有一个滞后的统计数据, 但一个重要的统计数据,我们放在我们的讲座之一, 弗吉尼亚州的能源消耗,
385	almost one quarter of that goes to power data centers alone.	将近四分之一的电源仅是电力数据中心。
386	Now, I pick on that because Virginia is a very data data center friendly state.	现在,我选这个因为弗吉尼亚是一个非常数据数据中心友好状态。
387	You have a lot of water and land and all the stuff that's near the government.	你有很多水和土地还有政府附近的东西
388	However, that's a tremendous amount of energy.	然而,这是巨大的能量。
389	That's a tremendous amount of resources to build it.	这是巨大的资源建造它。
390	We're now seeing data centers pop up in North Dakota all over the place.	北达科他州到处都是数据中心
391	And so it's not just the idea of a data center housing computers.	所以,这不仅仅是一个数据中心容纳计算机的想法。
392	These are changing the local economies and the energy infrastructure into the foreseeable future.	这些正在使当地经济和能源基础设施在可预见的未来发生变化。
393	So this is a megatrend.	因此,这是一个巨大的趋势。
394	And to build what?	和建设什么?
395	To power what?	什么权力?
396	When we do tokens.	当我们做标记。
397	I help my daughter with her algebra homework every night and I'm consuming data center tokens and somebody's electrical bill is going up in Virginia.	我每天晚上帮我女儿做代数作业消耗数据中心的代号有人的电费在弗吉尼亚州上涨
398	These are that may example may seem trivial, but this build out is historic.	这些可能看起来是微不足道的,但这种建设是历史性的。
399	And this amount of money, I think, exceeds all of the money the U.S.	我认为,这笔钱超过了美国所有的钱。
400	spent on the entire U.S. interstate system for like 60 years or something like that in the last two or three years.	在过去的两三年里在全美国州际体系上花费了60年左右
401	There's a great quote in one of the news articles recently about that.	最近的新闻文章中有一段很精彩的引用.
402	And so let's just let's just imagine that this is a huge event.	让我们想象一下,这是一个巨大的事件。
403	This is deploying, you know, tens of thousands of individuals, hundreds of billions of dollars of capital.	这是在部署,你知道,数万个人, 数千亿美元的资金。
404	And it's accelerating.	并且正在加速。
405	It's not just we were building it now.	现在不只是我们建造它。
406	We're going to stop.	我们要停下来。
407	It's going like this.	它是这样的。
408	So with that in mind, let's go to the next slide.	因此,考虑到这一点,让我们去下一张幻灯片。
409	Let's look at the circular investment ecosystem that we're in right now.	让我们看看我们现在所处的循环投资生态系统。
410	All right.	好吧,我们走。
411	So increasingly, we have seen what's referred to as circular investment.	因此,我们越来越看到所谓的循环投资。
412	I'll give you an example.	举个例子
413	NVIDIA同意向OpenAI投资1000亿美元.
414	OpenAI同意购买3000亿的计算能力以及Oracle和NVIDIA 以及OpenAI和Oracle 正在互相交易
415	Amazon is investing buying NVIDIA chips.	亚马逊投资购买NVIDIA芯片.
416	微软投资OpenAI.
417	And so you see historic amount of deals being done over a trillion dollars of deals just with OpenAI alone.	所以,你可以看到历史上的交易量超过一万亿美元的交易仅仅与OpenAI。
418	OpenAI 不公开交易.
419	It's a private company.	这是私人公司
420	We've never seen that behavior before.	我们从来没有见过这样的行为。
421	And on the right hand side, the first point I have is lane convergence expanded.	在右手边,我的第一点是车道的汇合扩大。
422	What would Ted and I mean by lane convergence?	泰德和我在车道汇合是什么意思?
423	Well, when the AI inflection really took off in 2023 and has grown since then,	当2023年AI事件真正开始并且从那时起,
424	obviously, the behavior that we saw among the mag seven in particular was that you had these companies go from having their core lines of business.	很明显,我们从七强中特别看到的行为是,你让这些公司从核心业务线。
425	Google does 95 percent of the revenue from search.	Google提供95%的搜索收入。
426	You know, Amazon does e-commerce, etc.	你知道,亚马逊做电子商务等等。
427	To now you have kind of an open field.	到现在你还是有点空旷
428	Anybody can make a search engine.	任何人都可以做一个搜索引擎。
429	Anybody can have an LLM and they can be quite competitive with the others.	任何人都可以有一个LLM, 他们可以相当竞争其他人。
430	And so instead of everybody staying in their individual lane, it's the analogy we use as it goes from swimming lanes to more like water polo.	与其说每个人都留在自己的个人车道上, 不如说就是我们从游泳车道到更像水球的类比。
431	Now you have this kind of beehive effect where a lot of different companies are competing for similar lines of business or emerging lines of business like they never have before.	现在你们有这种蜂窝效应, 很多不同的公司都在竞争类似的业务或新兴业务, 像以前从未有过的。
432	Fast forward to 2025.	快进2025.
433	What we're also seeing is we're also seeing these very large private enterprises, private companies that are very well funded, also partly funded by the public companies that they may be competing against.	我们也看到的是,我们也看到这些非常大型的私人企业, 资金非常充足的私人公司, 也有部分资金来自它们可能与之竞争的公共公司。
434	And so we have this lane convergence expanded idea.	因此,我们有了这个通道的趋同扩大的想法。
435	And so what's the risk of a circular investment?	那么循环投资的风险是什么?
436	Well, we're seeing some of those risks emerge in valuation concerns.	我们正看到一些风险出现在估值关切中。
437	Oracle has been in the news recently.	甲骨文最近上过新闻
438	They took a big hit to their share price because they had delays in their build outs.	他们的股价受到很大打击,因为他们的建设推迟了。
439	That sends a shockwave or ripple effect through other parts of the ecosystem.	这会发出冲击波或波浪效应通过生态系统的其他地方。
440	Hey, they promised all this spend of this capex.	嘿,他们答应了花这些钱的披风。
441	Now they're pulling back or there's delays on it.	现在他们撤退了,不然就拖延了
442	So we're not going to materialize the revenue.	所以我们不会实现收入。
443	Hey, this deal that we're supposed to do.	嘿,这是我们应该做的交易。
444	Well, that's on the rocks now.	好了,这是在岩石上。
445	When I say on the rocks, I mean, it's pushed out or it's held there.	当我说在岩石上,我的意思是, 它被推出或它在那里。
446	It's not happening anymore.	这已经不发生了。
447	And so if you have circular investment that drives up expectations of earnings and drives up valuations and something stumbles, it can have a cascading effect through the ecosystem.	所以,如果你有循环投资推动对收入的预期, 推动估值和一些意外, 它可以产生连锁效应通过生态系统。
448	And so investors and analysts get worried and nervous when they see a lot of circular investment like a vortex.	当投资者和分析家看到像漩涡这样的大量循环投资时,就会感到担忧和紧张.
449	And we're seeing that right now at incredible scale.	而我们现在看到以不可思议的规模。
450	I mean, incredible scale.	我的意思是,不可思议的尺度。
451	There's just there's there's no no other way to describe how big these sums and expectations are.	只是没有其他方法可以形容这些金额和期望有多大.
452	We just haven't dealt with these sums of money before in the markets.	我们只是没有处理这些钱以前在市场。
453	And so expectations are very high right now.	所以现在人们的期望很高。
454	That leads to the idea that maybe this is publish.	由此产生了这样的想法,也许这个是出版的。
455	OK, so another important point is at no time in history have we seen such large private companies push storytelling and narrative because stories matter.	另一个重要点是,我们从未见过如此大型的私人公司因为故事很重要而推动故事和叙事。
456	What we're telling you expectations.	我们告诉你的期望。
457	Ted and I signed a deal for one hundred billion dollars over ten years.	Ted和我签了10年1000亿美元的交易
458	What if in year two we get divorced and we go our separate ways and that one hundred billion dollar expectation just goes poof in the wind.	如果在第二年我们离婚然后我们分开走 1000亿美元的期望就会被风吹走
459	And we've seen that writ large.	我们见过大字条
460	And so private companies are influencing public companies to make these spending decisions and it's driving investor behavior, driving expectations.	因此,私人公司正在影响公共公司做出这些支出决定, 它驱动着投资者的行为,驱动着预期。
461	And so the question that we have and I'll leave you with is, is this strategic interdependence or is this artificial increase in value?	所以,我们所要问的,我要留给你们的问题是, 这种战略上的相互依存性, 还是这种人为的价值增加?
462	And are we going to see circularity reduce risk or increase risk?	我们要看到循环减少风险还是增加风险?
463	And so think of the swimming lanes analogy.	想想游泳道的类比
464	Think of lane converters expanded.	想一想车道转换器的扩展。
465	I have a lot of different companies competing for browsers, ads, LLMS, things like this, different applications, applications of different software companies.	我有很多不同的公司竞争浏览器,广告,LLMS,类似的东西,不同的应用,不同的软件公司的应用.
466	You know, now you can go out, you can build a Microsoft Excel clone essentially in your apartment in downtown Manhattan.	你现在可以出去了你可以在曼哈顿市中心的公寓里建造一个微软的Excel克隆体
467	What a concept.	真是个概念
468	And so circular investment could lead to tremendous growth potential because they may all work together to create much more growth in productivity.	因此,循环投资可以带来巨大的增长潜力,因为它们可能共同创造更多的生产力增长。
469	That's on the positive side.	这是积极的一面。
470	On the negative side, you could see cascading effects when expectations start to pull back, Ted.	在负面方面,当预期开始退缩时,你可以看到连锁反应,泰德.
471	Let's look at another slide, Jakob.	让我们再看一张幻灯片,雅各
472	Yeah, Ted, could I slow you down there, Logan?	是的,泰德,我能慢下来吗,洛根?
473	So let's just pause a little bit on the artificial increase in value.	因此,让我们稍微暂停一下人工增加的价值。
474	I make a deal and I promise to invest in you and you promise to buy goods and services for me.	我做个交易,我保证投资你承诺为我购买商品和服务。
475	Neither of us are CPAs, accountants.	我们都不是顾问会计
476	But the issue that arises in that context is when do the sales get booked?	但是,在这方面出现的问题是,销售何时开始预订?
477	And how do the analysts evaluate the promise to buy maybe $10 billion worth of products and services?	分析家们如何评价购买价值100亿美元的产品和服务的承诺?
478	And what Logan is saying is, well, it may be over five years.	洛根说的是可能已经过了五年了
479	How does the accounting work?	会计工作如何?
480	What happens if the deal falls apart in year two?	如果交易在第二年破裂怎么办?
481	That's sort of an example, a little bit deeper insight into the risk issues.	这是一个例子, 更深入地了解风险问题。
482	And the other thing I would just point out, if there's uncertainty,	还有一件事,我要指出, 如果有不确定性,
483	which we believe there is about which of the U.S. companies are really going to succeed with AI,	我们认为,美国哪些公司真正会成功与AI,
484	then from a risk perspective, rather than just bet on yourself,	从风险的角度来看, 而不是仅仅赌自己,
485	it makes sense possibly to spread out your bets and bet on some of your cooperators, competitors, rivals, partners.	将你的赌注和赌注分散到一些合作者,竞争者,竞争者,伙伴身上,这也许是有道理的.
486	So Amazon just announced $10 billion investment in OpenAI.	亚马逊刚刚宣布对OpenAI投资100亿美元.
487	OpenAI may be the most successful company this past year because they went from nonprofit to profit.	OpenAI可能是去年最成功的公司,因为他们从非营利到盈利。
488	They've remained private.	他们一直保密
489	And as you pointed out earlier, their valuation is skyrocketing.	正如你早些时候指出的,他们的估价正在飞涨。
490	It's now $500 billion.	现五千亿.
491	So if you're watching AI and you're one of these leading companies, you're saying,	所以,如果你在看AI 而你是这些主要公司之一, 你说,
492	well, maybe I should bet on OpenAI.	也许我应该打赌OpenAI。
493	Maybe I should think about that as a way to reduce the risk of the competition for me.	也许我该考虑一下以此来降低竞争的风险
494	So I think that's a little more detail about why this is such a big issue.	所以,我想这更详细地说明了为什么这是一个大问题。
495	The circularity, is it really strategic interdependence?	循环性,它是否真的在战略上相互依存?
496	Is it just risk reduction or is there also artificial increase in value?	是否只是减少风险,还是也人为地增加了价值?
497	And I think, I don't know if you agree, Logan, it can be a mix of all three.	而且,我不知道你是否同意,洛根, 这可以是一个混合所有三个。
498	It's a mix of all three and well put.	三种并合而为妙.
499	And back to the idea that this is narrative driven.	回到这个想法,这是叙事驱动。
500	This is a story we have not seen realized revenues yet.	这是一个我们还没有看到实现收入的故事。
501	As Ted mentioned, how are we booking the revenue?	正如泰德提到的那样,我们如何预定收入?
502	I promised Ted $100 billion.	我答应泰德1000亿美元。
503	He promises me $100 billion.	他答应我一千亿美元。
504	Do we book that today and just assume that's going to all work out?	我们今天要订这个吗假设一切都会好起来的?
505	Not necessarily.	不见得.
506	But that's a story.	但是,这是一个故事。
507	We're making each other a promise, making each other a deal.	我们互相许下诺言互相交易
508	Materialization of that is TBD.	其实际化是TBD。
509	Let's take a step back and look at it from a simplified perspective.	让我们退一步,从简化的角度审视一下.
510	Again, narrative driven.	再说一遍叙事驱动
511	What's the narrative?	如何是叙事.
512	What's the story telling you?	什么故事告诉你?
513	Stories drive economic events.	故事推动着经济活动。
514	OK, this is a Google search.	好吧,这是谷歌搜索。
515	This is Google search of AI bubble.	这是谷歌搜索AI泡沫。
516	Look at the chart.	看图.
517	Look at the right hand side.	看右手边.
518	Look at the spike.	看看那个钉子
519	Oh, my goodness.	天呐
520	That's something.	不错
521	And so this is AI bubble.	这就是人工智能泡沫。
522	They've declined in recent weeks.	近几周来他们下降。
523	Or excuse me, in recent months since the spike in the summer.	或原谅我,最近几个月自从夏天的暴涨。
524	But they've ticked up.	但他们已经跳了。
525	But this could be anybody from your uncle at a holiday party to a person like Ted teaching a class like this.	但这可能是你叔叔在节日派对上的任何人给泰德这样的人教这样的课
526	Let's jump to the next slide and look at what's happening on the institutional side.	让我们跳到下一张幻灯片,看看机构方面正在发生什么.
527	So these are charts.	这些是图表。
528	I'll break these down.	我会打破这些。
529	So the chart on the left is the AI and AI bubble, as are mentioned on Bloomberg terminals.	因此左侧的图表是AI和AI泡,如彭博终端上所提到的.
530	Bloomberg terminals are what we use on Wall Street to, one, communicate with each other.	彭博终端是我们在华尔街使用的, 一,相互交流。
531	It's a network.	腾讯网.
532	But also to look up any stock ticker or stock price in the world, any news event.	但也要查一查世界上的股票价格任何新闻事件
533	We use scrapers.	我们使用刮刀。
534	It is the central tool of an analyst.	它是分析师的核心工具。
535	You have your Bloomberg terminal.	你有你的彭博终端。
536	You have it on your phone.	你手机上有这个
537	You're connected to it 24 hours a day.	你一天24小时都和它有关
538	You live and breathe through your Bloomberg terminal.	你通过你的Bloomberg终端生活和呼吸
539	The reason why this chart is important, in my view, is that this shows you that institutional investors and analysts are also searching for this a lot.	我认为,这个图表之所以重要, 是因为这向你们表明, 机构投资者和分析人员也在寻找这个。
540	And they're reading about this AI bubble a lot.	他们经常读到这个AI泡沫
541	Then the chart on the right is also important because this is the amount of S&P 500 companies that are mentioning AI on their earnings calls recently.	右侧的图表也很重要, 因为这是500家S&P公司的数量,
542	And this chart goes back a ways.	这个图表可以追溯到过去
543	And so it's a bit of an eye test.	故略以眼验.
544	But if you look on the left axis, you see that about 50% of the Fortune 500 is mentioning AI on their earnings call this last trailing quarter Q2 would be.	但如果你看看左轴,你可以看到, 大约50%的财富500 在他们的收入上提到AI, 称这是最后一个末端的Q2。
545	And so that is it doesn't necessarily it's not indicative of bubble behavior, but it shows that this conversation has grown to the highest levels of corporate and it's trended up.	所以,这并不一定表明泡沫行为, 但它表明,这种对话已经发展到公司的最高水平,并且呈上升趋势。
546	It had a huge spike up after Chachi BT was released.	在Chachi BT发布后,它出现了巨大的突起.
547	You see that inflection 2022 Ted and Q2 2022 that spike up.	2022年特德和Q2 2022年的暴风雨
548	That's when Chachi BT was launched.	这时恰奇BT推出.
549	So these two charts, I think, illustrate what's happening in the investment world and the corporate world, as well as the retail world, which is the slide before.	我想这两张图表说明了投资世界和公司世界, 以及零售世界, 也就是之前的幻灯片。
550	If you have a comment, Ted, let it fly.	如果你有意见 Ted 让它飞吧
551	Otherwise, let's move to some different sources of capital, which I think are pressing right now.	否则,让我们转向一些不同的资本来源,我认为这些来源现在很紧迫.
552	Yeah, I think.	是的,我觉得。
553	No, let's move on.	不,让我们继续前进。
554	I think you're not surprising, given what we've said.	我想你并不奇怪,考虑到我们说过的话
555	US big tech companies need more capital, and they're looking at sources that in the past they didn't look towards.	美国的大型技术公司需要更多的资本, 他们正在寻找过去他们不期待的来源。
556	So this is this pretty much speaks for itself.	所以这几乎可以说明问题。
557	But again, think of this as individual high tech companies being hockey stick.	但再说一次,把这当做个别高科技公司是冰球棒。
558	They're attracting investors in the hopes to have this return, this penetration into the market and get revenues.	他们正在吸引投资者希望有这种回报, 这种渗透到市场并获得收入。
559	Let's go to the next.	来至下座.
560	Yep.	对
561	And I'll comment.	我会评论。
562	I'll comment on this quickly.	我很快会评论的
563	Historically, it's not common for tech companies to take on debt.	历史上,科技公司承担债务并不常见.
564	Right.	对
565	Just you know, debt is usually not part of the balance sheet item, even even at IPO.	你知道,债务通常不是资产负债表项目的一部分, 即使在IPO。
566	I mean, when you meet companies at IPO and maybe when they're later stage after IPO, they'll take on some debt.	我的意思是,当你在IPO遇到公司, 也许当它们后阶段IPO, 他们会承担一些债务。
567	But this is so we've invited a lot more sources of capital.	但这就是我们邀请了更多的资本来源。
568	You have asset backed securitization.	你有资产担保证券化。
569	You have straight up bond issuance.	你有直接的债券发行。
570	And so the pool of investors has expanded materially to fund these initiatives.	因此,投资者群体在物质上扩大了,为这些倡议提供资金。
571	And so this is a trend that I want the audience to pay attention to in 2026.	所以我希望观众在2026年关注这个趋势。
572	This is going to be important in my view going forward.	我认为这对今后的工作很重要。
573	All right, Ted, let's wrap this section up.	好吧,泰德,让我们结束这段。
574	I think we have one more slide on this.	我想我们还有一张幻灯片
575	Yes.	对
576	OK, so I think we should probably skip this, but I think you've already covered it.	好吧,我想我们也许应该跳过这个, 但我想你已经覆盖它。
577	I mean, the importance of narrative and will these companies be able to work their way out of the bottom part of the hockey stick?	我是说叙事的重要性这些公司能从冰球棒的底部走出来吗?
578	Mm hmm.	垛
579	Yep.	对
580	All right.	好吧,我们走。
581	Next slide.	下一个幻灯片。
582	I want to spend some time on this one.	我想花一些时间来做这个。
583	And thank you, Logan, for unpacking all that stuff about the bubble.	谢谢你,洛根, 整理了所有关于泡沫的东西。
584	I think it's you know, it is on everybody's mind.	我想这是你知道的, 这是每个人的心。
585	But hopefully now the people who are listening in have a better way to think about it.	但希望现在那些正在倾听的人有更好的方法去思考它。
586	You know, you pay attention to the private capital markets, the flow of funds, what interest rates are US companies paying for this additional capital, et cetera.	你知道,你关注私人资本市场,资金的流动, 美国公司为这额外的资本支付什么利率,等等。
587	So, you know, I think one of the benefits of being a little bit older than most people who are following the tech industry is that I have a longer term perspective.	所以,你知道,我认为比大多数跟随技术产业的人年长一点的好处之一是,我有一个更长远的视角。
588	And the AI discussion to me was was sort of muted five, ten years ago, but it became much more important.	对我而言,人工智能的讨论是五年前的,十年前的, 但它变得更加重要。
589	And I think it's proper to think about this as a major fork in the course of human history.	我认为应该把这看作是人类历史过程中的主要叉子。
590	You know, we humans have been around for a long time and it's all been natural what we refer to as natural intelligence.	你知道,我们人类在身边很长一段时间了我们所说的自然智能都是自然的
591	And now we have something called artificial intelligence that's moving not just into applications, but into generative AI, producing new ideas, works of art, music, writing, et cetera, and even going beyond generative AI to superhuman intelligence.	现在,我们有了一种叫做人工智能的东西, 它不仅正在进入应用, 而且正在进入基因智能, 产生新的思想,艺术,音乐,写作等等, 甚至超越基因智能到超人智能。
592	So given that we've already had this fork, natural intelligence and artificial intelligence, and given that we, you know, let's just pause it again.	所以鉴于我们已经拥有了这个叉子,自然智能和人工智能, 而鉴于我们,你知道,让我们再暂停一下。
593	AI is really right now, it's a battle between China and the US.	AI是真正的现在,这是一场中美战争.
594	It's the G2.	这是G2。
595	And the reason why we put in the Trump administration policy change is that I think within that fork of AI, we now have another fork.	而我们之所以把特朗普政府的政策改变是因为我认为在AI的叉子之内我们现在有了另一个叉子
596	And the second fork is are we going to have a convergent ecosystem for AI where even though we're rivals, China and the US share insights about AI.	第二个岔路口是,我们将为AI建立一个趋同的生态系统,在那里,尽管我们是对手,但中国和美国对AI有着共同的见解.
597	They share insights about how to invest in complementary assets like software and hardware, protocols, et cetera.	他们分享关于如何投资于软件和硬件、协议等补充资产的见解。
598	And there's exchange of ideas and leading edge human capital.	而交流思想和领先的人力资本.
599	Or are we going to get divergent ecosystems with China and the US?	还是要与中国和美国建立不同的生态系统?
600	I'll just I'll just reveal my hope and maybe my fear.	我只是会透露我的希望,也许我的恐惧。
601	My hope is these ecosystems don't diverge for the most part.	我希望这些生态系统不会大部分裂。
602	I think it's much better if we have one global ecosystem for AI.	我认为,如果我们有一个全球生态系统给AI更好。
603	And I'm concerned about the idea that we're going to maybe this is not, you know, this is not just saying the US doesn't want TikTok.	我担心的是,我们也许这不是,你知道的, 这不仅仅是说美国不想要TikTok。
604	It's not just, you know, China doesn't want to have Google, right?	这不仅仅是,你知道, 中国不想有谷歌,对不对?
605	As a search engine, this is much more profound.	作为一个搜索引擎,这要深刻得多。
606	This is how does each country move forward in developing AI and applying it?	这就是每个国家如何在发展和应用大赦国际方面取得进展?
607	Now, I think there's an argument for having a common ecosystem in large part because the two the two countries are emphasizing different strategies.	现在,我认为,有一个共同的生态系统在很大程度上是有道理的, 因为这两个国家强调不同的战略。
608	China is much more focused on applications, leveraging their huge advantage in manufacturing.	中国更注重应用,发挥它们在制造业的巨大优势.
609	And I think increasing that that advantage in the US is much more focused on generative AI and superhuman intelligence.	我认为,在美国,这种优势更加集中在基因AI和超人智能上。
610	So I don't think there's rivalry in terms of use applications as much as one might think.	所以,我认为在使用应用方面没有人们想象的那么多的对立.
611	So I think this is not so much just an issue for this coming year.	因此,我认为这不仅仅是明年的问题。
612	But think about the path dependency and where we're headed and ask yourself 10 years from now,	但想想路径依赖和我们要去哪里问自己10年后,
613	where are we going to be with these two different strategies and potentially two different ecosystems?	这两种不同的战略和潜在的两种不同的生态系统,我们在哪里?
614	What would that world look like?	这个世界会是什么样子?
615	What's the effect on societies around around the world?	对世界各地的社会有什么影响?
616	So that's my big look forward.	这就是我大前瞻。
617	Yeah, I'll jump in on that as well.	是啊,我也会跳进去的。
618	So one way we can break it down a little bit more is hardware divergence and software divergence.	一种方法我们可以更进一步地分解它硬件差异和软件差异。
619	We have seen hardware divergence.	我们看到硬件差异。
620	China does not have a native GPU.	中国没有本地GPU.
621	They have to buy them.	他们必须买下它们。
622	Their memory industry is lagging.	他们的记忆产业落后.
623	They don't have a native CPU.	他们没有本地的CPU.
624	But then on the software side, as Ted is pointing out, are we going to see two Internets?	但是在软件方面,正如Ted指出的,我们会看到两个互联网吗?
625	Are we going to see two digital economies?	我们将会看到两个数字经济吗?
626	Are we going to see two different AI ecosystems?	我们要看到两个不同的人工智能生态系统吗?
627	One could say that LLMs diffuse globally, albeit indirectly over time through academic research and people in use cases and things like that.	人们可以说,LLMS在全球传播,尽管随着时间的推移,通过学术研究以及使用类似案例和事物的人间接传播。
628	You know, talent is fairly moves around.	你知道,天赋是相当移动。
629	Are we going to see open open source models being a bridge between the two?	我们是否会看到开放源代码模型是两者之间的桥梁?
630	Possibly.	也许吧
631	But think about it also, in addition to Ted's comments, hardware divergence versus software divergence.	但想想看,除了泰德的评论之外,硬件差异对软件差异.
632	Back to you, Ted.	回到你身边泰德
633	So we started at 35 minutes after the hour and we said we'd go for about an hour.	所以我们在35分钟后开始, 我们说我们要去大约一个小时。
634	So maybe we could be a little bit quick with the remaining big issues and try to get done in six minutes.	所以也许我们可以快速一点解决剩下的大问题并在六分钟内完成。
635	But I think one question is which which country will be more supportive of AI development and all that it entails?	但我认为一个问题就是哪个国家会更支持AI的发展,
636	And I think let me just highlight number three.	我想让我强调第三点。
637	It should be number four, actually, the bottom one.	应该是第四,实际上,最下面的一个。
638	U.S. politics is starting to solidify and potential opposition against more data centers.	美国政治开始巩固和潜在反对更多的数据中心。
639	Senator Bernie Sanders has come out in favor of a moratorium.	参议员伯尼·桑德斯已经出面支持暂停
640	That means stopping investments in data centers.	这意味着停止对数据中心的投资。
641	Individual cities and states like Arizona have said no more.	亚利桑那州等个别城市和州都不再说这些话.
642	Georgia.	格鲁吉亚。
643	People are really upset about rising energy prices and they're blaming data centers.	人们对于能源价格的上涨感到非常恼火, 他们指责数据中心。
644	And that's probably right.	这也许是正确的。
645	So this support for AI is multidimensional.	因此这种对AI的支持是多方面的.
646	What effect is it going to have on jobs?	对工作有什么影响?
647	What is it going to do to society?	这对社会有什么影响?
648	What's it going to do to higher education?	这对高等教育有什么影响?
649	K through 12.	K到12号
650	Energy is a real flashpoint in the U.S.	能量在美国是一个真正的热点.
651	I don't have insights into how much support there is for AI development in China.	我对中国AI发展有多少支持没有见解.
652	But my guess is because it's being complemented by their strength in manufacturing, it's not as big of an issue.	但我的猜测是,因为它得到了它们在制造业中的力量的补充, 它不是一个大的问题。
653	That's just my intuition.	这只是我的直觉。
654	We'd love to get your insights.	我们很想得到你的见解。
655	In terms of the higher energy prices, I alluded to this earlier.	关于较高的能源价格,我早些时候曾提到这一点。
656	Whatever we're saying about data centers being added in the U.S., we infer from this slide that the investments in data centers by China are even greater.	无论我们说到美国增加数据中心, 我们从这张幻灯片推断, 中国对数据中心的投资更大。
657	Look at this chart.	看这个图表
658	I mean, this starts in 2010.	我的意思是,这始于2010年。
659	The lines cross.	线路交叉。
660	And China now has many more times greater capacity for power generation and therefore data center investment.	而中国现在的发电能力比现在大得多,因此数据中心的投资也因此增加。
661	Next slide.	下一个幻灯片。
662	So one question for the next calendar year and beyond is, you know, which of the G2 economies is more vulnerable?	所以下一个历年及以后的一个问题是, 哪个G2经济体更脆弱?
663	And I think the strength for China is their manufacturing advantage.	我认为中国的力量是它们的制造优势。
664	That may get so large that there's a backlash in Latin America and South Asia, Eastern Asia, Central Europe, Eastern Europe, Western Europe, et cetera, around Chinese manufacturing is leading to massive exports.	这可能变得如此之大,以至于围绕中国制造业的拉丁美洲和南亚,东亚,中欧,东欧,西欧等地区出现反弹,导致大量出口.
665	And I think there's naturally going to be tension where there's a domestic industry that competes against China.	我认为,在有国内产业与中国竞争的地方,自然会出现紧张.
666	So but but I think we'd start with that as an advantage.	所以,但我想我们从这个开始作为一种优势。
667	And then everybody knows China's vulnerabilities are in public debt, property crisis and weak consumer demand.	然后,大家都知道中国的脆弱性在于公债,财产危机和消费者需求疲软.
668	The U.S., its advantages in the areas I mentioned, I think a big concern that a lot of people have in the U.S.	美国,它在我提到的领域的好处, 我认为一个很大的关注很多人在美国。
669	is U.S. policy seems really chaotic, meaning haphazard, not consistent, not explained very well to the public.	美国的政策似乎真的很混乱, 意思是杂乱无章, 不一致,
670	Next slide.	下一个幻灯片。
671	So I think that's a big issue for the US and plus of course the AI bubble.	我想这对美国来说是个大问题当然还有人工智能泡沫
672	Let's go to the next one.	下一颂.
673	Logan, I think this is our last one.	洛根,我想这是我们最后一次了
674	Yes, I can cover this pretty quickly.	是的,我很快就能搞定这件事
675	We've highlighted a lot of these points, Ted, so I think we're good on this.	我们强调了很多这些观点,泰德,所以我觉得我们对此很好.
676	I'll just say that leading AI firms, they're influencing the conversation like they never have in the past with incredible size	我只是说,领先的AI公司, 他们正在影响谈话像过去从来没有过,以不可思议的大小
677	and investment from their co-operators, as Ted mentioned, or potential rivals, they're swimming in the same lanes.	和他们的合作者的投资, 正如泰德提到, 或潜在的对手, 他们游泳在同一条道。
678	We've seen lane convergence, so pay attention to that.	我们已经看到车道的汇合, 所以注意这一点。
679	Pay attention to the capital rounds that they're going to raise.	注意他们将要筹集的资本回合.
680	This will lead into the conclusions on the next slide with the IPO window coming.	随着IPO窗口的到来,这将导致下一张幻灯片的结论。
681	We're going to see some IPOs in 2026.	2026年见一些IPO.
682	So big things happening in the private companies.	私人公司里发生了很多大事
683	Ted, do you want to take number three?	泰德,你要拿三号吗?
684	No, I think we covered that.	不,我想我们掩盖了这一点。
685	I think we're good.	我觉得我们很好。
686	We don't have as much in the way of lines of sight and insights into the interaction between public and private companies in China,	我们没有太多的目光和深入了解公私公司在中国的互动,
687	but clearly both are important right now.	但显然两者都很重要
688	So, Jakob, should we turn it back to you?	那么,Jakob,我们该把它还给你吗?
689	Oh, conclusions first.	哦,结论第一。
690	We'll continue if we go forward, focused on the G2.	我们继续前进专注于G2
691	We'll continue our focus on semiconductors.	继续关注半导体.
692	This is Ted speak, but I believe what's happening right now is going to influence the course of human history.	这是泰德的演讲但我相信现在发生的事会影响人类历史的发展
693	I agree.	и種
694	We'll, of course, pay attention to the AI bubble.	我们当然会注意AI泡
695	The IPO window is open.	IPO窗口开通.
696	Private markets will continue to play a big role.	私营市场将继续发挥重要作用。
697	Logan is right.	洛根说得对
698	I mean, the narratives around AI bubble, they're so important to follow and understand who's got an authentic,	我的意思是,关于AI泡沫的叙述, 它们是如此重要跟踪和理解谁有一个真实的,
699	supported, fact-based narrative and who doesn't, especially given this interdependence and the circularity.	特别是考虑到这种相互依存性和循环性
700	You got to go beyond just the announcements of big deals, whether it's Oracle or Amazon, et cetera.	你必须超越仅仅宣布大交易, 无论是甲骨文还是亚马逊等等。
701	But we do think the CapEx boom is going to continue.	但我们确实认为CapEx的繁荣会继续下去.
702	I think the asymmetry that Logan pointed out, it's better to be wrong on the upside than wrong on the downside,	我认为洛根指出的不对称倒数倒数是错倒数是错
703	is driving this for U.S. companies and we suspect for Chinese companies as well.	我们怀疑是中国公司
704	OK, so, Logan, thank you for your insights.	好吧,罗根,谢谢你的洞察力
705	This is really a lot of fun to try to put this together.	这真的很有趣试图把这一点。

该视频共有字幕 706 条。解锁更多字幕为会员功能，请移动到价格

下载安装