美国颠覆、科技在亚洲的制造史、不确定未来中的需求力量
美债市场压力似乎在股市抛售期间引起了Trump的注意。 Trump看到投资者“有点不安”后，将广泛对等关税暂停90天并维持在10%，同时把中国关税提高至125%；Ben表示，10年期美债出现失序波动，可能触发追加保证金、基差交易平仓，最终演变成信贷冻结。“归根结底，这总是引起任何人注意的最佳方式。”
Ben的核心判断是，旧贸易体系不可持续，但一刀切关税会摧毁本可重塑该体系的美国需求。 美国在战略上依赖一个同等量级的对手，但不加区分地对进口零部件征税，会推高本土制造商成本，削弱其供应链多元化能力。他的方案是：识别关键投入品，公布可信的关税时间表，并以保底采购建设美国或可信盟友供应商——“发挥你的需求杠杆”。
Apple对中国的依赖，除了工资成本，也关乎灵活性和能力。 把制造风险移出资产负债表，帮助Apple走出1990年代末的危机、压降库存，并将一款iPod从2月提出想法推进到8个月后出货。令人不适却无法回避的结论是：“有充分理由认为，没有中国就没有Apple。”
中国已经成为制造业的TSMC：一个模块化平台，通过总量累积不断复利式增强专业能力和研发能力。 美国早期工厂围绕单一一体化产品优化，而中国供应商演变成横向扩张的客户服务体系，几周内就能服务创业公司，也能容纳Apple最后一刻的重新设计。这套生态无法简单“搬运”到美国，因为它的优势来自众多客户累积形成的需求。
Apple理性的供应商策略，为美国及其盟友带来了战略代价。 这套策略帮助中国竞争者挑战Largan Precision，也资助了Yangtze Memory的发展，使其成为SK Hynix和Micron的替代者；Ben表示，相关能力如今已延伸至HBM和Huawei。中国企业在Apple供应商名单中从少数变成多数，因为“需求控制供给”。
全面回流制造与降低对中国的依赖，是两个不同目标，需要不同的关税设计。 一刀切关税只有让一切东西都“贵得离谱”，才能迫使制造业广泛回流，同时还会抹去把零部件生产迁出中国的激励；有针对性的措施则可能建立一个可信制造圈。Ben的判断没有留下余地：“全面回流制造不可行，也不可能。”
台湾芯片依赖让脱钩可能更危险，而不会自动变得更安全。 中国采购台湾芯片，意味着北京在冲突中有所损失；把中国排除出全球经济体系，则会移除这层约束，而中国疲弱的消费市场和过剩工业产能已经像一座“战时经济”。更长期的威胁还将延伸至AI驱动的机器人领域，中国几乎主导了实体供应链中“所有相关环节”。
1. 美债压力在股市下跌期间引起Trump注意
Trump周三发帖称，中国关税“立即生效”提高至125%；他表示，已有超过75个国家致电美国代表，且没有采取报复措施。随后，他批准暂停90天，并在此期间将对等税率降至10%。即便如此，如果没有此前一周的混乱铺垫，这套贸易制度本身也会显得极端。
Trump后来承认自己在关注债券市场：“我昨晚看到，人们开始有点不安了。”Andrew也是其中之一：在他看来，周二10年期美债走势图可能意味着一个“雷曼时刻”，而非普通的股市回调。
Ben解释了传导机制：当追加保证金和自动化风险限额同时触发时，剧烈波动可能引发杠杆基差交易平仓，这或许解释了午夜前后的抛售。由于不完美对冲可能爆仓，并向信贷市场扩散，“尤其是10年期美债的大幅波动很糟糕。朝错误方向的大幅波动则非常糟糕。”
2. Ben的重写稿厘清了此前乐观论调遮蔽的问题
Ben称自己周一发表的文章是较弱的一篇：他坚持其中的实质内容，但文章缺乏统摄全篇的主题和结论。他重新找回的一条规则既私人又具有现实后果——“在我感觉良好、确实有话想说之前，我不该发表”——而周三的文章就是他的重写稿。
他也修正了此前围绕“A Chance to Build”的语气。为了以乐观收尾，他“遮蔽了我的警告和预测”：当前安排很危险，需要采取行动，但这并不意味着Trump的一刀切关税方案自洽。
3. 廉价亚洲硬件是美国软件的基础性互补品
Ben的框架从互补品出发：Android让高端智能手机变得更普及、更便宜，Facebook因此受益；同样，整个美国软件行业也受益于亚洲竞争者把硬件零部件做得更便宜、更好、规模更大。一旦这层基础遭到威胁，“即使你是一家企业SaaS供应商，也会受到伤害。”
贸易逆差视角严重错估了iPhone价值的分配。中国组装环节可能只增加“6至7美元”，但这部设备在统计上却会计入对华约400美元逆差；与此同时，Ben估算Apple利润约为500美元，这部分价值无法被只计算货物的口径捕捉。
这种遗漏让政策在机制上不利于美国最强的行业。如果关税只围绕实物商品贸易平衡进行优化，就“按定义而言，会对所有在美国实现正向价值的部分都不是最优方案”。
4. 中国帮助Apple用灵活性替代固定制造风险
Apple对中国的依赖颇具讽刺意味，因为它曾比大多数科技公司更努力地在国内制造。Steve Jobs在Fremont建了一座先进、高度自动化的Mac工厂，但Mac需求不足让闲置产能变成“白象”；他在NeXT重复了这一错误，几乎把这家创业公司推向破产。
当Jobs回到一家可能距离破产只剩60天的Apple时，Tim Cook开始清理库存、关闭德州工厂、精简爱尔兰业务，并把制造转向中国。库存是“每天都在贬值”的已付款产品；外包把这部分风险转移出去，也让Apple能够更贴近实际需求采购。
Ben把这一决定比作使用AWS。稳定、类似Basecamp的工作负载，可能通过自有硬件节省成本；但需求不确定或具有季节性的企业，会为了快速扩容而支付云服务溢价。Apple购买的同样是灵活性，而不只是更低的单位生产成本。
iPod就是典型案例：Jon Rubinstein在2月发现Toshiba有一款无人问津的小型磁盘驱动器，Apple围绕它设计产品，并在8个月后出货。后来，这条供应链还容纳了iPhone最后一刻从塑料改用玻璃的调整，支持Ben那个尖锐的结论：如果没有中国，Apple可能无法熬过1990年代末。
5. 中国成为模块化制造平台
传统美国制造通常与单一产品深度整合：“一家生产这个特定东西的工厂。”中国则逐渐演变为客户服务模式：横向扩张的工厂可以接下创业公司的订单，并在几周内交付。
Ben用Intel与TSMC的类比解释了其中的取舍。Intel把设计和制造整合在一起，可以更紧密地优化每一颗芯片；TSMC则提供标准化、类似乐高积木的组件。几乎所有人都采用TSMC之后，TSMC更大的规模为更多研发提供资金，并把制造前沿推到Intel内部服务体系之外。
中国实际上是“整体意义上的TSMC”。共享需求让这里成为一个协调焦点，客户共同为学习、灵活性和工具投入提供资金，这正是这套生态无法简单搬到美国的原因。
6. 低工资开启颠覆，能力积累完成颠覆
早期芯片组装仍需要工人用镊子接线、测试设备。亚洲劳动力成本可能是每小时11美分，美国则为2.25美元；与此同时，工会合同与行业快速改变产出、迅速扩张的需求相冲突。不过Ben强调，成本只是起点。
Intel在香港获得了很大进展，亚洲女性工人做出的产品更稳定、良率更高：“不只是因为更便宜，她们做得也更好。”美国政策在一定程度上鼓励这种转移，是为了在亚洲各地建立对共产主义的经济替代方案。
跨洋运输、国际电话和标准化集装箱随后促成了经典的颠覆周期。亚洲制造业先是在不同性能维度上低于定制化的本地工厂，随后“沿着这条价值链不断向上、向上、再向上，直到某种程度上替代了原有体系”。
7. 在美国生产iPhone会造成负和式劳动力再分配
Ben的反对理由比培训或建厂更根本：美国人已经拥有收入更高的工作，而且通常不想连接iPhone零部件。要支付足够高的工资来重新分配他们，就意味着从更高价值的活动中抽走劳动力；而按所需规模引入外籍工人，也不太可能得到Trump政府的认可。
贸易可以带来正和结果，因为每个经济体都进行专业化分工。要求每个环节都在国内完成，“隐含着负和”：投入美国iPhone组装的劳动力，就无法用于另一项美国活动。
Ben承认，更多培训和其他干预措施可能有所帮助。但他的观点是，这些措施无法靠一次关税切换逆转数十年的专业化分工。
8. Apple的供应商杠杆加速了中国技术能力积累
Apple对待供应商的方式，类似TSMC对待设备厂商：培育替代供应商，避免单一来源依赖，并压出更好的价格。TSMC可能帮助Lam Research和Tokyo Electron达到相同能力，再拿两家供应商的报价相互制衡；Apple在整个iPhone供应链中采用了同样的杠杆。
Largan Precision说明了这一点。它的镜头模组利润极高，曾一度达到Foxconn市值约五分之一，尽管只供应一个零部件。美国摧毁Huawei的智能手机业务后，Apple利用计算摄影保留6镜头设计，同时培养中国的Sunny Optical作为定价替代者。
Sunny Optical遭遇生产问题，Largan仍然占据主导，但战略模式延续了下来。对于Yangtze Memory，Apple提供研发支持，并给出先进存储器未来获得保底需求的可能性，使其能够挑战SK Hynix和Micron；后两者的存储器是iPhone最昂贵的部件。
美国施压要求Apple不要采购Yangtze Memory的存储器，但相关能力已经保留下来。Ben将这笔投资向下游连接到Yangtze的HBM业务和Huawei供应，同时指出，中国企业已经从Apple供应商中的少数变成多数，取代了台湾、日本和德国供应商。
9. 产业政策应结合有时间表的压力与保底需求
Ben会先识别来自中国的关键投入品，然后宣布分阶段实施的关税，同时在美国或可信盟友中建设相关能力。时间表很重要，因为企业会基于可预测的未来经济条件进行投资；但本周制造出的更多是“对疯狂局面的恐惧”，而不是可信度。
Andrew补充说，4年一届的总统周期让持久的产业战略很难维持；Ben则指出，使用紧急权力征收关税并取得财政收入，涉及尚未解决的宪法问题，因为这 traditionally 属于国会权限。两重复杂性都削弱了制造业投资所需要的长期承诺。
更强的胡萝卜是保底采购：“推动创新的方式不是铲钱，而是用需求作为胡萝卜。”中国会不喜欢一项明确要替代它的计划，但仍会继续销售，因为房地产泡沫转化为出口投资后，中国工厂仍然需要客户——这使美国可以“从中国采购零部件，帮助我们摆脱对中国的依赖”。
10. 一刀切关税保护最终组装，却固化中国零部件
一位来自Cleveland的听众点出了矛盾：美国进口中约一半是美国制造业使用的投入品，对这些产品征税，可能在替代工厂出现之前就造成裁员和产出下降。即便暂停了更广泛的措施，对中国征收125%的税率仍未解决大部分问题。
越南说明了原产地核算为何会误导。AirPods的一批货可能计入约150美元的越南逆差，即便越南通过组装环节只增加约10美元，而底层大部分价值来自中国。
一刀切关税可以阻止转口，但也会消除迁移零部件生产的激励：如果无论最终组装在哪里都要被同等征税，企业不如继续把困难的上游生态留在中国。Andrew认为，谈判可以要求越南限制中国投入品，但其边境条件和经济关系会让执行变得困难。
决定性区别在于，华盛顿究竟想让所有制造业回流，还是想降低对中国的依赖。一刀切关税只有通过施加巨额价格和痛苦，才能追求前者；Ben支持后者，并称普遍回流制造是“一项糟糕透顶的想法”。
11. 中国打破了美国设计、同时也被其利用的体系
Andrew认为，中国并非没有责任：出口导向的“以邻为壑”政策和不断恶化的工业产能过剩，正在掏空全球各地的行业。Ben同意，中国成为“有史以来最大的搭便车者”，但这套体系最初确实由美国设计。
他的红海例子具体展示了这种失衡：受到威胁的船只主要承载从中国运往欧洲的商品，但负责巡逻这条航线的却是美国。旧交易是，美国提供安全保障，贸易伙伴购买美国国债，由此把贸易逆差、财政逆差和资产市场缠绕在一起。
中国的规模让这套架构变得不可持续，但诊断问题比修复问题容易。“把整个系统炸掉，当然可以解决问题，”Ben说，随后以第三次世界大战作为重置体系的刻意恐怖边界案例。
12. 台湾让脱钩变成一场安全赌博
先进芯片是一项特殊依赖，因为大量关键产能集中在台湾，超出美国的直接控制范围，而中国也不愿在这段关系上作出妥协。Ben大约从2016年或2017年起就持续关注这一点，因为除芯片外，大多数产品最终都可以在盟友之间寻找供应来源。
但让中国采购台湾芯片，会创造其不愿入侵的激励。全球一体化最大的安全收益，就是北京有所损失；每一步把中国从体系中剥离出去，也都会降低“摧毁整个系统”的代价。
Andrew问，关税驱动的经济破坏是否可能压低中国的攻击意愿。Ben看到的却是相反风险：一个近乎衰退的消费市场，加上大规模过剩供给，已经像一座“战时经济”，而“消化过剩供给的一种方式，就是把它炸掉”。
未来的暴露程度可能更高，因为AI通过机器人进入实体世界，而中国几乎主导所有相关投入品。Ben最后的感慨既是文化性的，也是战略性的：美国应当创新并“靠增长走出去”，而不是让Elon Musk削减社会保障，同时让政策默认走向“切断、保护、筑墙”。
13. Suzuka说明，F1的高端产品是接近感，而不是比赛本身
Ben这段制造业主题之外的插曲始于名古屋Toyota博物馆：这里展示了Toyota从纺织业起步，发展到工作机械和现代汽车工艺的历程。他购买的F1 Experience套票还包括周四赛道乘车和维修区参观；女儿与Kimi Räikkönen自拍，一家人几乎见到了所有车手。
周五练习赛是最好的赛道体验，因为观众可以在Suzuka的“8”字形布局中自由移动，从发夹弯一路走到200R和Spoon，感受到电视画面会压平的高低落差和过弯过程。47起草地火灾基本毁掉了第二次练习赛。
周六带来了情绪上的回报。Andrew强调，Max Verstappen驾驶的赛车在大多数周末都比不上McLaren，但Max仍以约0.1秒优势拿下杆位；偏爱Honda和Red Bull的观众随即沸腾。周日则是“有史以来最无聊的比赛之一”，Max每圈领先约1.2或1.3秒，Andrew称之为整整60圈。
商业启示在于，F1通过分层定价实现“接近感”变现：卡车乘坐、维修区、以及价格逐级上涨的围场通行权限。Ben如今支持在受限的城市赛道举办冲刺赛，但在真正的赛道上则“绝不”支持，因为练习赛期间自由走动本身就是产品；Suzuka明显胜过Vegas，不过其令人疲惫的通勤也让Austin成为下一次体验的诱人选择。
Hello, and welcome back to another episode of Sharp Tech. I'm Andrew Sharp, and on the other line, Ben Thompson. Ben, how are you doing?
I can't decide if we timed this perfectly or—
Mm-hmm.
Terribly. We've gotten a lot of inbound inquiries: “Where is Sharp Tech?” We said we were going to miss last Thursday and Monday for reasons that we may get into later in the podcast, but here we are. Basically, the world's the same as it was a week ago, so we're all good, right?
It was a very chill week across the board. I am leaving my Bloomberg Terminal to talk to you for the next hour. I'll check back periodically throughout the pod. A wild ride. I'm excited to get into it. Any thoughts on the way in?
Well, it's been an interesting week, to say the least. I wrote 2 articles. I put a little thing in my second article of the week about how I'm breaking some rules. Basically, I was on vacation last week, and all of this news happens—these broad-based tariffs.
I think one of the challenges with the Trump administration—this was certainly the case the first time around, and it's the case the second time around—is that a good friend of ours has remarked continuously that his primary objection is the complete inability to rest and be normal for any extended period of time while Trump is president.
Mm-hmm.
That is certainly an impulse on one hand: Let's see what actually happens in the long run. On the other hand, there is the danger of crying wolf, of getting worked up so often that when something truly destructive happens, you want to have a full ammo pack to respond.
And so, on Monday—I was on vacation last week and didn't get back until Monday afternoon, for good reasons, which we may get into—I did feel compelled to write. I wrote an article that I didn't feel awesome about, and the longer it was up, the worse I felt. I got some feedback from people saying, “Look, this is one of your weaker articles in a while,” which didn't hurt because I already felt that way.
Mm-hmm.
It was another lesson that I temporarily forgot and needed to keep in mind: I need to not publish until I feel good and have something I want to say. I felt better about what I wrote on Wednesday. It was a mulligan. I don't try to take them too often, but here I am.
Mm-hmm.
They're like, “Oh, again.” But here I am. Like I said, I try to resist the urge toward solipsism. Too much writing in general is just people articulating their inner demons and feelings, and I'm like, that's not what strategy is. Strategy is not about me. But here I am.
Well, exactly. You had broken—
Doing it right now.
The seal on solipsism on Wednesday. You're keeping it going with this podcast. But you wrote that article on Wednesday. To answer your question at the top, I think it was perfect timing for a weeklong vacation, because the story was changing every 4 or 5 hours for the last week or so. I don't know that we would've been able to say anything coherent that wasn't immediately outdated.
1. Trump Rewrites the Tariff Story
You wrote your article on Wednesday morning, and then, lo and behold, the story changed again midday Wednesday. That's where we're going to begin: a note from President Trump on Truth Social at 1:18 Eastern Time.
About 4 hours and 18 minutes after I posted, yes.
Exactly. It was perfect. He wrote, “Based on the lack of respect that China has shown to the world's markets, I am hereby raising the tariff charged to China by the United States of America to 125%, effective immediately.”
An incredible opening clause there, but yes, continue.
“The lack of respect that China has shown to the world's markets.”
Definitely no other lack of respect being shown to markets this week.
A little bit of projection. That's no problem.
“At some point, hopefully in the near future, China will realize that the days of ripping off the USA and other countries is no longer sustainable or acceptable. Conversely, and based on the fact that more than 75 countries have called representatives of the United States, including the Departments of Commerce, Treasury, and the USTR, to negotiate a solution to the subjects being discussed relative to trade, trade barriers, tariffs, currency manipulation, and non-monetary tariffs, and that these countries have not, at my strong suggestion, retaliated in any way, shape, or form against the United States, I have authorized a 90-day pause and substantially lowered the reciprocal tariff during this period to 10%, also effective immediately. Thank you for your attention to this matter,” which is how he's signing off a lot of his posts these days.
It does amuse me every single time.
It is.
It's a great catchphrase.
2. The Bond Market Sends a Warning
Trump, in a press conference later Wednesday afternoon, was asked whether the activity in the bond market spurred some of these changes and said, “I was watching the bond market. The bond market is very tricky. I was watching it, but if you look at it now, it's beautiful. The bond market right now is beautiful. I saw last night where people were getting a little queasy.”
I can attest that I was one of the people getting a little queasy looking at Treasury charts on Tuesday night. You had to explain to me what a basis trade was.
This is the Trump problem, right? Did you even know that a 10-year Treasury chart existed before, like, 48 hours ago?
I did, because I was aware it was a source of anxiety for members of the administration.
Yeah.
But 3 months ago, the answer is no.
By and large, if equity markets are in trouble, you would expect people to be fleeing to bonds.
Mm-hmm.
The safest investment in the world historically is US Treasuries. If people are going into the bond market, then the interest rate should be dropping because you're bidding or buying bonds, and what you're willing to pay is based on what those bonds are going to return. If there's tremendous demand for bonds, then you're not going to get a great interest rate on them because everyone wants them, right?
Mm-hmm.
The problem is that in this case, even though the equity markets were way down, bond prices were rising. That could be some combination of a large sovereign selling, and people were wondering if it was China.
Right.
It could be, and maybe most disturbingly, an expression of a lack of confidence in the US. Is this—
Are they going to pay their debts?
The 10-year even going to be repaid?
Yeah.
Then there's this basis trade, which is basically this hedge fund arbitrage where you're buying long-term bonds but then shorting them in the short term. The problem with this is that it's a way to eke out little bits of profit.
Mm-hmm.
But those are subject to margin calls and risk calculations. The fact that this big spike happened around midnight, when all these risk calculations are run, means it very well could have been a lot of automated selling based on these prescribed rules for when you have to get out of a trade.
Mm-hmm.
That's the unwinding of that trade. That was probably what was happening. But again, the problem with bonds is that it's also the most liquid market in the world.
Right.
It's like cash, right? I think someone asked me—it might have been you—“Do we know if China sells?” I was like, “No, we don't.” Do we know what cash is being exchanged in a drug deal on the southern border? No, it's cash. It's not tradable, right? That is kind of what bonds are to the financial markets.
Right. It would've been a big problem if a bunch of hedge funds had their basis trade blow up.
Yeah, this is the problem with these huge moves in the stock market: stuff blows up. That's just what happens, right? Of course, everyone has risk management, blah, blah, blah, blah, blah. The reality is that nothing is perfectly hedged. People don't expect drops of this magnitude in this short a period of time, and that's how these things tend to spiral.
There's a big drop, then something blows up, then things go on, and then, in the long run, you're looking at a real freeze-up somewhere in the system, with real credit issues. Basically, large movements in 10-year Treasuries in particular are bad.
Mm-hmm.
Large movements in the wrong direction are very bad. That is always the best way, at the end of the day, to get anyone's attention.
Exactly. It got my attention on Tuesday night. I knew that the market had been struggling, but looking at the charts on Tuesday night, it was like, okay, so we could have some sort of Lehman moment here, and I have no idea what happens next. Clearly, it got President Trump's attention, and the market rebounded on Wednesday. Given the updates after you published your article, do you have any general thoughts on where we are now and how this has played out, at least as of Wednesday night at 9:30 Eastern Time?
No, and I think my article is still fine. It still stands up. My point and goal, basically, in my Monday article—
Mm-hmm.
I stand behind what I wrote. It was just a poor article that didn't really have an overarching theme or a very good conclusion.
Mm.
My point there was the frustration I had over the weekend. I took it as a given: yes, this is bad, and I failed to fully articulate why this specific thing was bad, but I also felt a lot of the response was head in the sand. It was like, why can't we just keep things the way they were? What I wanted to point out is that the way things are is long-term problematic.
Mm-hmm.
The reality is that the US is deeply dependent on another country that is a peer.
Not just a peer—an enemy at this point.
Well, I mean, that's to be determined, I think. I still have a streak of optimism that goes back to when I wanted to revisit my article last November, which I think holds up very well. That article I wrote back in November, number 1, worries about Trump killing the golden goose. Number 2, it lays out why this is particularly bad for tech.
3. Tech Depends on Asian Complements
The particularly big problem for tech is that Asia goes hand in hand with technology. Technology is software. I talk about zero marginal cost. All my talking, by and large, is about software. But that is enabled by this old principle of making your complements cheap or free. The things that you need for your product, if someone else is making them, ideally they're widely available, cheap, or free.
Mm-hmm.
So if you're Facebook, and you want to connect the whole world, it's really good for you if everyone in the world has a mobile phone. The mobile phone is a complement to Facebook. Android, for example, is one of the greatest things that happened to Facebook because it dramatically drove down the price of high-end smartphones, such that there could be that many more people using these devices, and then they could be on Facebook. That's a classic complement.
If you zoom out, this is tech as a whole going back 40, 50, 60 years. The rise of Asia is about tech hardware, at least in a tech context.
Sure.
Having an entire region of the world dedicated to producing your complements, most of which are commodities and thus have prices driven down to the marginal cost, and competing with each other to make them better and scale out and all these sorts of things, is an amazing thing for US tech. For that to be threatened is bad for everyone.
Mm-hmm.
Even if you're a business SaaS provider, it's bad for you. The mistake I made—the reason I circle back to this—is that I was very worried about this in light of Trump's election.
Well, can I just say that I remember recording a podcast about that article you're referencing, “A Chance to Build,” and going into it, having read the title, I was like, “This is going to be really optimistic, and we'll see what America could build.” Then, reading through section by section, it was like, man, Ben doesn't sound all that optimistic about any of this.
Right.
And so—
No, I mean, in retrospect, I should have just leaned on the pessimism, right? Look, this is bad. I was trying to be optimistic. I want to be optimistic. I want to believe that America can figure this out. We put it on the tin: We are Americans. We want America to succeed, and so I wanted to end things on a positive note. What that actually did was obscure my warning and prediction.
Mm-hmm.
And so I self-sabotaged that one as well. This is why I really wanted to—so you go back to Monday. I wanted to make the point that this is a bad idea, but the answer isn't to do nothing. There is something that needs to be done, and the fact that nothing has been done is what has opened the playing field to someone to come in who doesn't really understand software or appreciate that an iPhone coming from China drives far more value to the US than it does to China.
Mm-hmm.
China's value add from assembly is around $6 or $7 or something like that. That's counted in trade calculations as a $400 deficit or whatever it might be to China. Meanwhile, Apple's profit on that phone is around $500.
The problem with over-indexing on these trade-deficit calculations is that they're missing a huge part of the puzzle, and that huge part of the puzzle is absolutely dominated by the US. If you're not cognizant of that, and if you're going to create a policy completely indexed on those specific goods numbers, you're going to, by definition, be suboptimal for all the parts that are positive in the US.
Mm-hmm.
Anyhow, that's why I came back Wednesday. I said, look, no, let's step back and walk through why a blanket assertion that this stuff can be made in the US not only isn't plausible but also isn't desirable. Let's do it in a framework that I think is both correct and resonates with the Stratechery audience.
All those things still hold today. I'm very fine with this change, particularly given that I have money in the stock market. Who doesn't? But I don't regret rewriting and putting that out there.
Okay. Well, let's start with what you wrote on Wednesday, and I want to start with a specific example you cited. You wrote, “Apple has probably done more than any other company to move China up the curve in terms of the ability to manufacture components, often to the detriment of suppliers in the US, Taiwan, South Korea, Japan, et cetera. From Apple's perspective, spending time and money to bring Chinese component suppliers online provides competition for its most important suppliers, giving them greater negotiating leverage. From the US's perspective, this means a host of technologies and capabilities downstream from the smartphone, which is to say nearly all electronics, including those with significant military applicability, like drones, are being developed in China.”
I think that is an interesting place to start, since China is still facing 125% tariffs here. With Apple, big picture, what does Apple look like in the 21st century if they never develop this supply chain in China?
4. Apple Needed China to Survive
One of the great ironies of Apple being so China-dependent, and arguably more than any US company the driver of Chinese technological development, is that Apple, more than any tech company, tried to be US-only.
They really did. Apple computers were made in the US. They had this big facility in Texas. Then Steve Jobs, when the Mac came out, presided over this—I mean, you think about Apple Park. Apple Park is this beautiful spaceship of a headquarters. You talk to any Apple employee who works at Apple Park, and they hate it.
Hmm.
They find it not ideal for the way that they work. It's great if you're an executive walking around. You have beautiful views of greenery, and you can check in on anyone. It's not so great for day-to-day operations.
Is it just too spread out?
Well, I think it's also not big enough, so there are too many people crammed in, and it's all open, which isn't ideal.
Hmm.
Every single Apple employee I've talked to who works there hates it.
Looks better in keynotes than it is in practice. Good note for all of us.
It's a great keynote stage. This was the Fremont Mac factory that Steve Jobs built in the '80s. It was this beautiful, pristine, futuristic factory, as advanced as it could be, and super-automated. It had real shades of when Tesla was building the Model 3 factory and had all these problems because they tried to over-automate things. Musk came back and said, “There needs to be a progression in getting—
Mm-hmm.
—this figured out.” There was this really famous anecdote about Steve Jobs specifying a specific type of paint that ended up screwing up a bunch of the machines or something along those lines. The core problem, though, was that they built this factory to produce…
A whole bunch of Macs, and they didn’t sell that many Macs. That’s a big problem. This, by the way, is an important thing to keep in mind: If you overbuild capacity, you need to use that capacity.
Mm-hmm.
If you don’t, it’s this albatross around your neck. And so Apple had this Mac factory albatross around their neck. Steve Jobs gets fired, the Mac doesn’t sell very well, and the costs of it not selling well were dramatically amplified by the fact that they had this factory to build way more Macs than they were actually selling.
He goes to NeXT and does the same thing. He wants to build his beautiful manufacturing facility. The problem is NeXT was a startup. That basically all but bankrupted them trying to do all their own manufacturing. You fast-forward, and Steve Jobs comes back to Apple. Apple at this point is a mess. They’re on the verge of going bankrupt. They were, what, 60 days away or something along those lines.
You have that Wired cover.
Right.
“Pray.” Steve Jobs hires Tim Cook. And over time, Apple wasn’t the first into China or the first to work with Taiwanese suppliers, who were really sort of the foundation of this. A lot of the PC industry had been going there. Why had the PC industry been going there? Well, you had this standardization of components that was happening around the Wintel core.
Mm-hmm.
And for all the reasons we’ve talked about, right? You want to gain scale. It is in the interest of Windows to drive, “We’re going to support everything.” Why? Because if we support everything, then everything can run Windows, and then we get a much greater return on our upfront investment.
These themes we talk about come up again and again and are applicable to things like manufacturing. So you have this shift over time, with all these components being made in Asia, particularly by these Taiwanese companies. Some manufacturing was done in Taiwan, but the Taiwanese companies were also the first into China to build the factories there.
This is why Taiwan is such a ridiculously wealthy country.
Yeah.
It’s kind of a weird economy. You have all this money from building in China that is just sitting in Taipei. And so Cook helped save the company. Apple had all these product lines, which Jobs famously shredded, but in collaboration with Cook.
They had all this inventory sitting there. Inventory—you’ve paid money for the stuff, and you haven’t gotten money from customers yet. It’s just diminishing in value on a daily basis. And so Tim Cook completely cleans up the supply chain—or, sorry, the inventory problem—and they move their manufacturing to China. They closed down their Texas facility. They had a large facility in Ireland that they dramatically slimmed down and basically made into a tax haven.
Basically, this was the shift that happened, and this was when it was just the Mac. This was even before the iPod came around. What you could do was offload all this risk onto other companies in other countries, and then you could rationalize your supply chain so you only bought what you could actually sell.
Right.
And that’s so obviously beneficial from a business perspective. We talk about this in the context of the cloud, right? To build up your own data center, number one, you’re not using all the servers, so that’s just a waste. But then you get a big rush of growth, and you can’t build a new data center fast enough to serve your new customers.
The ability to scale up and down with AWS or something like that is why you pay the big AWS bills. It’s not really about saving money. If you measure on a one-to-one basis, you can save money running your own hardware. The Basecamp guys have been doing this. They’ve shifted all their stuff to their own hardware, and they keep tweeting, “Man, we save so much money by doing this.”
Well, that works. Basecamp has a pretty stable customer base. They know how large they are. They know what compute they need. It makes sense. More power to them. If you’re an entity with moving loads, or you have scalability—say you always scale up during certain buying seasons and you scale down at other times—this flexibility is worth a lot.
Mm-hmm.
And that’s what Apple got by going to China. They got this sort of flexibility, and they could scale. Along with that scalability capability, you got things like being able to come up with the idea of an iPod in February, when Jon Rubinstein goes to Japan and Toshiba shows him this new disk drive that no one wants to buy because it’s slow and not very big. He says, “Well, we could put a device around that. That could be an iPod.”
Mm-hmm.
Eight months later, Apple can ship it. That is a function—
Unreal, yeah.
—of having a supply chain that can do that. And this is a key thing that’s different about manufacturing. We think about, worry about, and bemoan the loss of U.S. manufacturing, but a lot of the U.S. manufacturing that existed previously was integrated with the product itself. It was a factory that made this specific thing.
In China now, it’s much more flexible. Here’s another analogy: the Intel-TSMC one. Intel’s opportunity was that they both designed the chip and manufactured the chip. What this meant in practice was that they could often change things in the design to accommodate the manufacturing, so they could reuse old equipment, for example. Or they could work through a problem, and you could fix it in software—in this case, fixing it in the design of the chip to work around it.
You can push on the frontier the more pieces that you control. TSMC, on the other hand, was much more flexible and much more standardized. Here are the defined things that we can do. You can get these IP blocks. It became much more like LEGO building.
Right.
You put these pieces together, but on a one-to-one basis, the Intel chip would be more performant because it was fully integrated to maximize and pull out the potential of that. The problem with this is that because everyone ended up on TSMC, TSMC ended up with way more volume than Intel, which meant they had way more money and way more R&D. They could invest in pushing things further.
This scalability and flexibility means you become a shelling point for everyone to pile in and basically collaborate on R&D. Everyone collaborates to push chip production forward by buying from TSMC.
Mm-hmm.
And then TSMC invests all this money to push things forward. Intel is just serving themselves.
Mm-hmm.
And this was China as a whole—it was basically TSMC.
Exactly.
That’s the Chinese manufacturing base.
Exactly. That’s right, and so that’s why you end up—over time, this starts with more scalable things. We’re going to build the same motherboard for a gazillion people, or whatever it might be, and you have this standardization in PCs.
But with this standardization, you got more and more modularity in the system, so that today, if you’re a startup, you can call up China and there will be factories that can make you an order and get it to you in a matter of weeks.
Mm.
They’ve developed these systems that are super flexible because they’re meant to serve a whole bunch of people, and so they’re horizontally very scaled. This is how manufacturing works today: You have the broad—
It’s more of a client-service business today than it was 50 years ago in the United States.
That’s right.
Yeah.
Absolutely. To go back to the iPod, you could spin up the capability to make an iPod and actually ship it within 8 to 9 months.
Mm-hmm.
You fast-forward to the iPhone, and you have famous stories of Apple—I can’t remember which iPhone. I think it was the first iPhone. It was going to be a plastic screen, and they changed it to glass at the last minute.
Yep.
They did it, and it’s amazing that they did it. That capability is worth a lot. The reality—the true answer to this question—is we can sit here and critique Apple, and I have some criticisms of Apple in terms of their China dependence, but there’s a good argument that Apple does not exist without China.
Yeah.
They just go under in the late ’90s because you needed to offload the risk of manufacturing to even do the turnaround that they did.
And it’s interesting. Until I started hosting this podcast and hosting Sharp China, I was under the impression that Apple and China was purely about exploiting a more favorable cost structure and cheaper labor. But there was also a great Financial Times article a couple of years ago. Some of what you laid out there was new to me, and some of it was familiar, but the abilities that China had on the manufacturing side made certain technologies in the iPhone more viable. It’s just the sort of thing that wouldn’t have been able to be replicated in any other manufacturing market.
So it wasn’t really just labor. There was manufacturing and expertise that Apple spent billions of dollars investing in in China over the last 15 or 20 years, and the symbiosis has been part of the story the whole way.
5. China Climbed the Manufacturing Ladder
You know, it’s tricky: how did China get to this point? It did start with—I talked about chips in November—and there’s this reality: connecting wires for a chip, which was how you did assembly, meant you had the chip, and to connect it to the substrate, you actually had people in there with tweezers connecting wires.
Mm-hmm.
That’s how chips used to be made. I waxed poetic about zero marginal costs, and chips were the natural example. It’s called Silicon Valley even though it’s software now because chips were the first zero-marginal-cost product, and that’s always been a little bit of a lie.
You had this marginal cost, which was the labor to actually connect all these little wires and test all these chips, and the way that got shrunk down was absolutely by shipping it to Asia, where you had people who worked for 11 cents an hour. People in the US cost $2.25 an hour. People in the US wanted labor unions. Intel was vociferously anti-union.
Mm-hmm.
All these ideas of “we’re going to make it up in volume” and “we have to be able to scale quickly up and down”—there’s real tension there with a labor-union sort of parts-based system. You think about a lot of union contracts that sort of lock in what is to be produced and when. It’s that lack of flexibility that is actually completely anathema to these products in figuring out what works and when it works, whether you can scale it super quickly, whether you can make changes, and all these sorts of pieces.
So, obviously, any labor question is going to be very controversial, but there were good, valid reasons that Intel was anti-union—
Yeah.
At that point. And you also had—if you want to get really risqué and dangerous, if you want to tiptoe—
Let’s get dangerous.
—into some cultural aspects, this is something that generally… Intel actually got a lot of traction in Hong Kong with Asian women who were really good at this. They would work long hours and connect all these wires really well and reliably, with much higher yields than they were getting anywhere else.
Hmm.
It wasn’t just that it was cheaper. They also did it better. And that started it. Meanwhile, the US was enabling this. They cut the tariffs that impacted this. The laws were formulated to encourage this because you had Vietnam happening. You had South Korea a decade before that. There was a concern about communism in Asia moving into these countries, and giving them these economic opportunities was seen as a way to push back against that.
Yeah.
You look at Hong Kong, you look at Singapore, you look at Taiwan—that was successful. But it’s worth noting this did start with the drive to save money.
Mm-hmm.
And so I don’t want to dismiss that.
No, I don’t want to dismiss that. But what’s interesting is that it has evolved over the last 20 to 30 years, and China and Chinese manufacturers have moved up the value chain themselves and become indispensable to Apple, at least historically.
Right. Well, that’s the disruption point, right? What is disruption? You have some sort of technological change that enables a new way of doing things. In this case, the technological change was that you could fly across the ocean, you could make phone calls across the ocean, and you had standardized shipping containers.
The shipping containers didn’t matter much for chips. Those are mostly flown in planes, given how small and valuable they are and how important time is to them. But this technological change made this shift possible, and you had this new market entrant, which was Asian manufacturing, competing on a different variable than the existing ones, which were bespoke, customized, tuned, local—all the things that came with local manufacturing.
Mm-hmm.
What you had was this very classic story of the low end moving up and up and up the value chain until it sort of swapped what was there. The reason it’s worth observing this is that we do need to make changes, but they need to be made with the awareness and humility that you don’t just lift this entire ecosystem up and—
Exactly.
—move it to the US.
Yeah.
And if you don’t get that, if you don’t understand how we got where we are and the sheer impossibility of going down—and it’s not that Apple wants to make more money. It’s literally impossible. You and I are going to continually choose to sit here and podcast and not connect wires on an iPhone, right?
It’s not that we’re saying, “Well, yeah, my job’s important,” but other people will go do that. Work your way all the way down. There are all these jokes online—
Well, there are also skills as far as manufacturing is concerned. You couldn’t—
No, but—
—flip a switch.
—even setting aside the skills, right? People will say, “There should’ve been more training. There should’ve been all these sorts of things.” All of which are true, by the way. I’m not dismissing those.
Yeah.
I’m just saying, on a fundamental economic level, no one in the US wants to do these jobs, and for good reason. There’s this idea that you’re going to be making iPhones in the US, so you’re going to have to get people to shift from jobs right now that are much higher-paid. Apple’s going to have to pay way more. But then those people are working in these factories; they’re not doing these other jobs, so you’ve diminished your capability otherwise.
What are you going to do? Are you going to import a bunch of immigrants? That doesn’t sound like President Trump’s— that’s one solution. I’m not sure that’s amenable to the Trump administration. Trade can be positive-sum. Everyone can win. Saying everything has to be in the US is implicitly negative-sum. Stuff that goes to manufacturing these iPhones in the US is stuff that’s not going to something else.
Yeah, exactly. It’s one of the reasons people are breathing a sigh of relief, at least to some degree. I don’t want to get too excited, because everything could change next week, on Apple. I want to come back to the reindustrialization point because I think that’s an interesting conversation in its own right with respect to China.
But as far as Apple’s uncertainty here, do you have any sense of how complicated this could be? There was a report that Apple plans to shift more iPhone production to India to avoid US tariffs on Chinese goods, which threatens $70 billion that China earns annually from Apple. But the future looks cloudiest for them and perhaps Tesla after this week.
6. Apple Built Its Chinese Rivals
Yeah, Apple probably, at some point—it’s hard because, on one hand, this ecosystem sort of developed organically. On the other hand, Apple has also pushed this ecosystem forward, as I mentioned, right?
Mm-hmm.
You see this with any sort of supplier, or you see this with TSMC. Let’s use TSMC as an example again. One of the reasons why you have the big 5 semiconductor equipment manufacturers—and they used to be more specialized in what they did, but now they all kind of do everything—is that TSMC continually sets them against each other because they don’t want to have a single-source supplier, right?
TSMC doesn’t like the situation where they have to pay ASML a bunch of money because they’re the ones that make EUV. When it comes to etching, when it comes to mask making, when it comes to all these different pieces, they will say, “Okay, we got this working with Lam Research and Tokyo Electron. Why don’t you get it to work, too? Here’s how you should do it. Here’s XYZ.”
And they’ll bring them up to speed, and then they go back to Lam Research and say, “Yeah, Tokyo Electron actually figured it out, too, believe it or not. And they’re willing to offer this price. How about you lower your price here?”
Mm.
This, again, is just the way TSMC is at the center of everything. Whoever is at the core of the value chain gets to push their way around. This was sort of like Windows back in the day, right? Windows was standardizing everything. They loved it when everything was competing with each other because that meant there were more PCs, right?
Yeah.
TSMC loves when their suppliers compete with each other. This is what Apple does with the iPhone. They do it everywhere in the iPhone. I brought this up just because I’ve always found Largan Precision interesting. It makes the lens assemblies for the iPhone.
One of the more egregious Apple bits, if you’re an Apple customer, was that Largan Precision dominated lens assemblies. They were this tremendously successful Taiwanese company with huge profit margins. I pointed out years ago that they were worth almost as much—or, like, a fifth as much—as Foxconn, despite the fact that they made 1 component of the iPhone and Foxconn assembled the whole thing.
Mm-hmm.
That just speaks to how valuable this could be.
Apple hates this. Every dollar that goes to Largan Precision—
For the love of God—
...is a dollar that hurts them.
...we need an alternative.
That's right.
Yeah.
And so Apple has been working for years to get an alternative Chinese supplier online here. There's lots of stuff that's not super clear here, but as I understand it, Largan Precision pushed forward to make a new 7-lens assembly instead of 6, and they had a committed partner in Huawei to buy it. The U.S. kills Huawei's smartphone business. Suddenly, they're like, "Oh, no." They're like, "Well, Apple will buy it."
Mm-hmm.
Apple's like, "You know what? We're getting really good at computational photography. We'll make 6 work, actually. And also, by the way, Sunny Optical in China has figured out how to make 6-lens assemblies, so let's get some negotiation on price here." And, of course, Apple was the one pushing Sunny Optical to be able to do that.
Right.
Now, this is one where Apple—it's not the best example, because Apple did get burned. Sunny Optical had some real production problems. I don't know if this was combined with some of the issues they had with Face ID and stuff like that. This might be a different one. There was something that happened there too, I think, with a supplier. But basically, Largan ended up coming back, and they still do dominate. But supposedly Sunny Optical's going to be back this year in maybe some MacBook lenses or something like that. I don't know.
The key takeaway is that Apple has been investing in Chinese competitors to other—
Right.
...companies around the world.
Another well-known one is Yangtze Memory—
Right.
...where Apple basically funded all the R&D. What was Apple's reason for them to do this? Apple's a guaranteed buyer. Apple is a guaranteed buyer, and so Yangtze, on this super advanced DRAM that would've given Apple an alternative supplier to SK Hynix and Micron—and memory's the most expensive component in an iPhone—it's very much in Apple's favor to have—
Mm-hmm.
...a Chinese memory supplier that's going to undercut these other suppliers. This was one where it got the U.S.'s attention. They're like, "Apple, you'd better not buy memory from Yangtze Memory." But the problem is, all that R&D existed. Today, Yangtze is doing very well, or they're starting to figure out HBM, high-bandwidth memory, which is important for AI chips. They're supplying Huawei. This is downstream of Apple investment.
Mm-hmm.
Apple wasn't investing in high-bandwidth memory specifically. They were investing in Yangtze generally.
Right.
Which brought them up to speed as a competitor. And this is a very good example of how Apple wasn't doing anything wrong from a corporate perspective, but this was pretty damaging to—
Yeah.
...you talk about things like chip controls, right? You talk about things like drones. All these pieces of investing and pushing. And you can look—Apple puts out the supplier list for the iPhone every year, and Chinese companies used to be a minority. Now they're a majority. By definition, that is replacing Taiwanese companies.
That's replacing—
Yeah, exactly.
...Japanese companies—
...suppliers.
That's replacing German companies, right? And Apple's doing it because they have the power. One more point on this: demand controls supply. Apple promises massive demand and can bring to bear money to support that because they control demand. This gets into my whole thing about Intel, about chips, about how the U.S.—the way you spur innovation is not by shoveling money. It's by the carrot of demand.
Mm-hmm.
By bringing to bear that aspect. That's the point of the article generally. I mean, there is an argument to remake trade for national security reasons, if nothing else. But you have to understand that it's not going back to the U.S. Formulate a sphere of influence that you trust, and start directing things using demand and guaranteed buying. The U.S. does have demand, and demand does control supply at the end of the day.
The problem with these tariffs is that they were demand-destructive. They were destroying the U.S.'s capability even as they were like an atomic bomb for China. So, yes, changes need to be done. Not this change. Leverage your demand. Lean into it.
Right. Well, let's talk about U.S. manufacturing, because Alex wrote in and said, "I live in Cleveland, where I drive past hollowed-out warehouses and factories every day. I desperately want to see new factories and jobs flooding into the Rust Belt, but it's unclear how these policies do that. Half of U.S. imports are used by U.S. manufacturing. Those imports just got more expensive, and they're going to translate to layoffs and reduced manufacturing output. How does shutting off U.S. manufacturing from the global supply chain address our strategic problems with China?"
So you wrote about this on Wednesday. How are you thinking about the problem that Alex highlights there? His email came in before the broad-based tariffs were paused for 90 days, but a lot of the manufacturing inputs come from China.
Yeah.
So I don't think that problem goes away all of a sudden.
Nope.
What do you think? How should that factor into the equation here?
The problem with talking about this is—
Mm-hmm.
...you're sort of veering dangerously close to centralized planning, right? It's like—
Yeah.
...we're going to choose to control here and control there, but with that sort of out of the way—that acknowledgement out of the way—you realize and identify what inputs you have that are dependent on China, and you put scheduled tariffs on those. These are going to be tariffed over time. And the problem with that is a schedule needs credibility.
Mm-hmm.
I'm not sure this week did a lot to engender credibility. It may have engendered fear of craziness, but I'm not sure it did a lot to engender credibility. But at least, with that acknowledgement out of the way—
Well, it's also, frankly, a challenge of the U.S. democratic system. Presidents change every 4 years.
Yep.
And so it's hard to develop a coherent industrial strategy.
Well, there's the whole question of whether these tariffs are even constitutional, right? They are revenue-raising. That's supposed to be Congress's job. They're being done under an emergency-act authorization. At what point—
Seems like a broadened exercise of authority from President Trump over the last couple of weeks, but, yeah.
But I mean, it is kind of an emergency. We do have a national security problem, right? And so, yeah, that's a whole other vein, which I'm not going to wade into—the constitutional questions. But, yeah, what you do is you acknowledge, realize, identify. You start one channel to spur the creation of these capabilities in the U.S. or in a trusted ally. There does seem to be an overarching—
Mm-hmm.
You look at the list of trade numbers.
Mm-hmm.
All the 10% ones were, like, the Anglosphere, the Americas, and the Middle East. That's kind of a signal, I think, of what the U.S. sees as its sphere in the long run, and this is the power of demand.
Yeah.
So you're on the China side. You're like, "Okay, they have this scheduled tariff. They're going to keep buying this from us, but they're clearly trying to usurp us. And you know what they're going to do?"
Keep selling.
They're going to keep selling you the stuff, right?
They need customers.
Listen to this. Think about newspapers putting their stuff on Google or going onto Facebook, right? All the things that we talk about in the media—the power of an aggregator is that you deliver so much demand that suppliers have to come to you on your terms, even though they hate it.
Yeah.
Now, again, this isn't an aggregation story. Aggregation is very explicitly about zero-marginal-cost items. Manufacturing items are not that. But the power of demand—this is hardly original to me—Keynes's whole thing was that demand matters more than supply, right?
The power of demand is, I think, viable to actually have a structured plan that advertises the fact. The point of this plan is to diminish our reliance on China, which we have right now, so we're going to keep buying from China and buy parts from China to help us get away from China.
Mm-hmm.
And China's going to look at that and say, "Yeah, that really sucks. Well, guess we'll sell to you anyway because we have to sell. We have the factories. We have the capability. We rerouted our property bubble into export investments. We have to keep selling."
So here's my question, though, because that seems to be in conflict with what you were laying out about binding China to the chips in Taiwan as a hedge against an invasion anytime soon.
Yeah. The chips thing is its own thing. That's the Trump card, no pun intended, that China has, and it's the most horrific Trump card in the world because it entails—
But I just mean in terms of having an existential urgency in order to get some of these industries off the ground may be what's required to move forward, because it's not all that attractive as an investment.
7. Chips Could Deter a Taiwan Invasion
Yes, but I think chips are special. Chips are so essential, and we're so dependent on Taiwan that here you really do need—you ideally want to bind. Look, the best thing for everyone, economically anyway, is that the system keeps—the reason why—
I was going to say—
This is so hard—
The last 30 years—
—is we have a perfect system, right?
Yeah.
Asia makes all the stuff. We make all the software. Everyone gets rich. The US gets richest of all. Everything is amazing except for the national security problem of us having total dependency on what I will call our peer and what you will call our enemy, and that's the fly in the ointment.
Adversary. Let's be a little more diplomatic. But, yeah, I mean—
Well, I think that's implied in “peer,” right? It's not someone that we get to push around. And I think peer would be a big step for the US. There's a lot of thought about China as this upstart little brother, and they're not. They can make stuff that we can't, and a lot of that stuff has national security implications. And you look at things like AI. AI is going to be manifested in the real world through robotics. Almost everything that goes into robotics, China dominates.
Mm-hmm.
It's not just that we're challenged with what exists today. We're behind the eight ball when it comes to the future, and I think the way to solve that is to start out by admitting where we are. When you admit where we are, that entails realizing that China has done tremendous things in these supply chains, has built up something that we can't just lift and take.
Yeah.
It's actually a problem. And this is going to require so much subtlety and long-term thinking, all of which is maybe a little fanciful to hope for. But look, I said at the end, cut China off. Pull up the ladder on chips. It's depressing to me.
Mm-hmm.
I thought we were Americans. I thought we could have a can-do attitude. We're going to settle the frontier. We're going to innovate. We're going to build new things. We're going to grow our way out. Why is Elon Musk going to waste his time trimming Social Security payments? What a waste.
Well—
Elon Musk builds stuff. America builds stuff. Why aren't we trying to innovate and create new things and grow our way out of our problems? That's how we solve things in America, and so many of our solutions are all predicated on just cutting off, protection, building up walls. That's not who we are. That's not who we've been, and that's—I find it a bummer.
Yeah. Obviously, there's the risk of an invasion with Taiwan, which severs all of the trade relationships that the US has with China, and it seems like some of these policies, if they remain in place—again, she could call Trump Friday and all of this could look different a week or two from now. I'm not sure it can be walked back at this point, but that is one risk.
The other concern is that China has become so dominant in manufacturing that they're hollowing out industries all over the world, and that is part of what the US is trying to counter with the tariff regime here. Obviously, blanket tariffs all over the world would not serve anybody, and I never took that seriously as a long-term plan for the Trump administration, even though members of the Trump administration said that they were going to do that.
Yeah. A lot of what happened with the Trump tariffs in the late 2010s is all that manufacturing. People are like, “Well, why is Trump blocking Vietnam?” Well, because Vietnam is full of factories.
It was a transshipment hub for the last couple of years.
And to be clear, there's stuff happening there. They're adding value. But a lot of these components go to Vietnam, are assembled, and then shipped on, and now it's from Vietnam. This gets at the whole calculation problem. The value added in Vietnam is very low relative to the value of the device as a whole, but when an AirPod comes from Vietnam, that's $150 negative in the trade deficit, even though Vietnam added $10 worth of value—maybe not even that much. A lot of that value is actually China.
And so, yeah, blanket tariffs stop that, right? Everything is going to be higher-priced. But that is so damaging to the US, too, because the problem is that it doesn't actually shift the real issue, which is the components. If final assembly is going to be tariffed everywhere, then just keep all the component manufacturing in China.
You have no incentive to move it, right?
Yeah.
No, exactly.
Right.
But honestly, what I wonder is whether those reciprocal tariffs were a starting place for negotiation, and then part of the negotiations that will take place with a country like Vietnam will include limiting relationships with Chinese exporters and limiting transshipment. That's very difficult for a country like Vietnam to do because they have their own relationships—
They have a border with China.
—with China. Yeah.
Yeah.
Exactly. And so I don't know whether the Trump administration could possibly execute on that sort of an idea, but I do think that that is becoming the plan now. Was it the plan last week? Who's to say? But they're attempting to ring-fence China to some degree as part of a Cold War 2.0 scenario here.
I agree with you. It's sad that that's where we are, but I don't think China is blameless for ending up where we are, given what they've been doing on the export side with beggar-thy-neighbor policies for the last several years. It's gotten worse since people complained about it a couple of years ago.
Yeah, to be clear, China's the biggest free rider of all time, right?
Right.
I think a great example is the Houthi thing going on in the Red Sea. The ships going through there are ships going from China to Europe.
Right.
Very little US trade even goes through there. Why are we the ones fixing this problem, right? The reason we are the ones is that that was the deal. We're going to be the global police force, and you're going to buy all our Treasuries to pay for it, right?
China was just so large that it kind of screwed up the system. It doesn't make sense for a country of that size and capacity. The reason the US debt has exploded in conjunction with China entering the system—that's not an accident. Yes, China has stopped buying as many US Treasuries over the last 10 years. It's a liquid market, as we talked about. All the assets from this trade deficit have to go somewhere.
Mm-hmm.
And payments have to balance. That is the most important thing. These deficits are all intertwined together: trade deficits, country deficits, all these things are interconnected.
Yeah.
And China for sure has freeloaded off the whole thing. Again, by our design. All these problems go back to the ’90s, right? This is—
By our design, but our design is no longer viable or—
It's not viable.
It's not sustainable.
It's really not.
Yeah.
Right. The problem is, again, it's really easy to diagnose a problem. It's way harder to fix it. You can definitely fix the problem by blowing up the whole thing. That's what wars do, right?
Yeah.
If you want to fix the global trading problem, let's trigger World War III. That will fix it. That's one way to do it. I personally don't endorse that. One of the core things, to your point about blanket tariffs, is whether the goal is to reshore all manufacturing or to limit our dependence on China. Those are distinct, separate goals that require distinct, separate strategies.
I think full onshoring is not viable. It's not possible. The US has moved too far up the value chain. They're not going to go back down. If you did want to do that, blanket tariffs are a good idea, because they would just make everything crazy expensive and force everything onshore. Again—
Incur pain. Yeah.
Right. But I think that's a terrible idea. And so—
Well, and honestly, I don't know how serious that idea ever was, just for the record. I think Peter Navarro wanted to do that, and—
Ron Barbara or whatever his name is.
Yeah. There are all sorts of people who've been popping up on different cable news shows. At no point over the last week did I think that that was what would end up happening at the end of all this. I think there were always going to be negotiations, and the question is what the shape of those negotiations will look like in the weeks and months to come here, and how explicit they are about targeting China as part of an economic war. I guess we'll have to wait and see.
Yeah, I guess we'll have to wait and see.
I mean, again, this is where the Taiwan bit—and this is why I've been obsessed with the chip thing for years. I've been writing about this for 7, 8, 9 years. I wrote about Intel ages ago, but about Taiwan specifically, I think I started in 2016 or 2017, somewhere around there. Because this is the one big piece where we have a really big dependency that we don't control, and we need to get control of it.
Mm-hmm.
At the end of the day, we can, with other countries, figure out a way to make most things. Chips are a really important thing that is on Taiwan, which China doesn't want to compromise on. And the other thing about a global trade war is, you know, go back to my chips piece. Why do I want China buying chips from Taiwan? Because that's an incentive not to—
Right.
Invade Taiwan. This is the other problem with trying to reform the global system. There are so many downsides to China being tied into the world economy. The big upside is that it is in China's interest not to screw that up.
Mm-hmm.
And the more we remove China from the global economy and remove them from the system, the less they have to lose from blowing up the system. And that's just a—maybe this is me, but where's my American can-do attitude? Why am I just wanting to kick the can down the road? Roads can be longer than you think. Cans can go further than you think, and things can happen.
I mean, we had a friend that reached out—
AI is coming. What's that going to mean? I don't know.
Yeah. We had a friend that reached out to me as the host of Sharp China and was like, “Sharp, do you think this makes it less likely that China goes to war with Taiwan because they're just economically devastated by all this?” And I said, “Ah, I mean, that's one read on it.”
Hopefully, the opposite. Yeah.
But that's maybe too optimistic.
One way to look at China's supply surplus is that they already have a wartime economy. They have a consumer market that is kind of in a quasi-recession. It's not great. You add this massive excess of supply generation, which right now is being dumped on the world. One way to consume excess supply is to blow it up.
Mm-hmm.
Right? The war economies are not good. China's economy right now is not good. They're already structured for this. It's not great.
Yeah. Well, also, countries all over the world have been complaining about Chinese overcapacity, so the US is not the only country that has a problem with the current state of play in the global economy. I guess we'll see what happens. The market is back up, and that's really, I guess, all that matters at this moment.
What? No, it's the right statement, because I think the reality is that for a lot of people, that's true.
And all of this is so irrational. I mean, I want to be careful about assigning too much intent to the Trump administration, but you step back—Trump, what happened over the past week does follow his playbook to a T, where he sets out some insane and unreasonable ask, convinces everybody that he's crazy, and then convinces everybody that he's making concessions and finally behaving reasonably as he imposes 125% tariffs on Chinese exports and a blanket 10% tariff. Ironically, it's a very Chinese way to negotiate here.
And then the market is like, “Oh, my God, finally, we can breathe a sigh of relief.” If you had announced these policies 2 weeks ago, everybody loses their mind. But I guess for now, stability is here.
Yeah. It is true. The ball was moved, and markets are mostly back. I think—
And I don't have to spend all night checking the Treasuries market. So I'll take that as well.
Yeah.
8. Suzuka Delivers the F1 Experience
Before we close, though, we did get this note from Greg. He said, “How nervous is Ben that Yuki is going to take the fight to Max and expose Max for the mediocre talent that he is?” So that's just a ridiculous question to ask.
Well, also, clearly written before this past weekend at Suzuka.
Indeed.
So yes, I have a trip report, Andrew. I was at Suzuka over the weekend, at the race, which obviously was on Sunday. I could not get back on Sunday. That's why I got back Monday afternoon. So there you go. There's the answer that we've been talking about.
Mm.
And I do have to say, lots of great Yuki merch. I got a great hat.
Big thumbs up there.
Can I just confirm to everybody that Suzuka is an F1 circuit in Japan?
Oh, yeah. Sorry.
And Ben attended 3 days of F1 with his family in Japan. I was insanely jealous the entire time. Take it from here as you give us your travelogue.
Oh, yeah.
Well, it was actually 4 days.
Oh.
We stayed in Nagoya and went to Kyoto and did some things around Nagoya. There's this Toyota museum. That's the home of Toyota. There's this Toyota museum in Nagoya. There are actually 2. There's one where you can actually see the assembly line. We didn't go to that one. That's outside the city.
There's one inside the city that shows that Toyota started as textile manufacturing, and it shows the development of all these machines over time, which is amazing. Then they do the same thing with cars, but they mostly show modern processes. They have actual working machines there that aren't actually doing anything, but they're moving around, and you can see it. You can see these robots moving car pieces around and all these sorts of things. Awesome. Awesome museum. It could not be more up my alley. Extremely highly recommended.
Mm.
So, big thumbs up to that. But, yeah, we actually had 4 days. It's really hard to buy tickets in Japan. I've done it before. I think I did it with a Taylor Swift concert ages ago. Japan is just its own little world. Sometimes credit cards work, sometimes they don't. Everything's in Japanese.
So I did the lazy rich person thing. I bought from F1 Experience, and because I got that, I also got a Thursday pass. So we were there 4 days.
I was wondering how you got the Thursday passes. So that was through F1.
That was through F1.
Tell me more.
Suzuka is great. It's also not close to anything. It's an hour at best from downtown Nagoya, and then traffic can be insane, obviously, around the race. So it was a really tiring weekend, just the driving back and forth.
Can I tell you, I had a lot of FOMO watching you and seeing photos every day? A couple of months ago, you told me you were going to this race, and I thought about going. Then I did look up where exactly Suzuka is, and I was like, “You know, it's going to be a major pain in the ass. Fly into Tokyo, take another flight, and then get to Suzuka every day.”
Oh, take a train. But yes.
Well, I just didn't have the energy to do that. So I'm glad you did, though, and it was a shorter trip for you.
I just wanted to get tickets to the race. But it came with things that actually ended up being incredible.
Mm-hmm.
Number 1, we got to do a truck ride around the track. It's amazing what you don't get from TV, particularly all the elevation changes. Then you watch it on TV later, and you understand really deeply what's happening in different areas. I would love to do one of those truck rides around the track at every single track. Fly me around the world. I just want the truck ride around the track, because it adds so much to the viewing experience.
Then we got to do a pit walk, where you walk down and see them working on the cars. Again, it's Thursday. They're not driving around. I warned my daughter—my daughter's gotten very into F1—“Look, assume we're not going to see any drivers.”
Mm-hmm.
Be pleasantly surprised if we did. We basically saw all the drivers. She actually got a selfie with Kimi Räikkönen, which is in her age range, which is very concerning. We saw Charles Leclerc a foot away, riding by on his bicycle. We drove on the track, saw Yuki on his bike, saw George Russell running on the track. We got to see a lot of drivers. It was super cool. Totally worth it.
And also, you got to buy merch and all those sorts of things when there's no one there, because it was Thursday.
Yeah.
Then I'm like, “Okay, that's going to be too much. Not going to go on Friday. Well, I might; I really wanted to go on Friday.” Just practice days. Best day. Friday was so great because you could go anywhere.
Interesting. Yeah.
And particularly, the first practice was awesome. I was with my son, and we sat at the intersection of the bridge because Suzuka is a figure-eight track, with the cars going underneath. We started at that intersection by the bridge and then slowly worked our way around. You go under the track at some points; there are all these tunnels and things like that. We went to the hairpin, then went down, then walked all the way down 200R, which was my favorite corner, and there's a spot right in the middle where you see the car coming right at you.
Mm-hmm.
The way they turn is amazing. Then we walked all the way down to Spoon, and there was a big pavilion where you could eat and get things like that. It was amazing to walk around during practice. We were going to do the other side during practice 2, and there were 47 grass fires. That was a big L. That was unfortunate, but anyhow.
Well, and the grass fires, again, for people who weren't watching, would just delay things for 20 to 30 minutes at a time.
Yeah.
Well, but the clock keeps running, too.
Practice 2 was a total dud. That was unfortunate. Okay, so we go to day 3. I'm like, “Okay, this is really good. We're going to go to the race on Sunday. This drive is so tough. We're going to be in the grandstand, the cars are going to go by fast, and we can watch qualifying on TV.”
Mm.
My wife, credit to her, was like, “No, I think we're going to go. We're going to do it.”
We're here. Let's do it.
She did not go on Friday, to be fair. So we get there, and Saturday is why you're a sports fan. It's why you go to live events. Just sitting there and watching—well, number one, it was startling how much faster they were. We were in the same assigned seats on Saturday. We got there in time for morning practice and then qualifying.
We were on the straightaway, and when they turn the engine modes up on those cars, it's tangible how much faster they go. It's pretty crazy.
Mm-hmm.
But the moment when you see it, it's going down. Q1 and Q2 are fine. Yuki's out in Q2. It was very sad. The crowd was very bummed. But then what happened in Q3 was that Max Verstappen pulled out this insane lap and won pole by 0.1 seconds in a slower car and all these sorts of things. The rush, and the whole crowd just erupting and pumping their fists super loud—it's like that—
So was it a Red Bull-friendly crowd? Was that the favorite team there—
Yeah, yeah.
—for sure?
For sure.
Because of Yuki?
Yep.
Okay.
Well, because of Honda. It's also the Honda team.
Mm.
So, once Yuki was out, everyone was full bore for Max. It's kind of an F1 thing, right? It comes down to tenths of a second. That tenth of a second, Max flies by, and his time pops up at the top of the sheet. I'm feeling emotional right now articulating this. The rush, the joy—it was awesome. That was worth the whole weekend.
And it is a good thing, because then our driver went totally the wrong way going back, and we were stuck in traffic for 3 hours.
3 hours.
And I had to use the bathroom really badly, and it was all worth it.
Ben was floating at that point. Well, and to lay out the stakes of exactly what happened, Max is in a car that is not competitive with the McLarens on most weekends this year, and he was not expected to get pole at Suzuka.
Watching at home, the McLarens took pole initially in the session. First it was Oscar, then it was Lando, and that looked like the end of it. And then Max comes in—
Well, in the stands, you saw them winding up, and Max was one of the last cars.
Exactly.
And so you kind of—
He was the last one—
You sense when it's coming. Yeah.
—of the actual contenders to cross the line, and he came up with one of the best single laps I've seen on TV in a really long time. As the Max diehard, I was happy that you were sitting in the stands to watch that. And then—
Yep.
He handled his business on Sunday. Spoiler alert.
One of the most boring races of all time.
Yeah.
My wife's like, “This is it?” I'm like, “Look, this was a pretty bad one, I have to admit.” No one changed order.
Apologizing on behalf of the sport.
Yeah. Max is up by 1.2 or 1.3 seconds every single lap of the race.
For 60 laps. Yep.
Sunday was pretty boring, not going to lie. We did get a nice view of the trophy presentation and things along those lines. I bought these specific tickets because I was worried that if my wife's coming and it rains on her, I'm going to be in really big trouble.
Yep.
So I have to get seats that have a cover. Sunday was pretty boring. I do have some broader F1 takeaways.
Okay.
This was so much better than Vegas. It was ridiculous.
Yep.
I can now appreciate the value of sprint races for city tracks. If you can't walk around and you're just going to sit in your seat, sure, do something more exciting. Have some stakes, because that can make it more interesting.
If you're at a real track, practice is the best. The idea of not having to really pay attention but being able to walk everywhere and see everything—I really want to go to Austin, to Circuit of the Americas, which you can do a similar thing on practice days. Don't make me—
Let's do it. Fly in—
—pay attention.
Mm-hmm.
I want to be there, see the cars going around, walk all over, and get a sense of the track. That is amazing.
My broader F1 takeaway is that I'm more in favor of sprints at road tracks, city tracks, and there should never be a sprint race at a real track. The whole experience is getting there and walking around. You're watching on TV anyway, right? They have big TVs there.
Definitely get the radio thing. I got one from American Express. Thank you, American Express. They were handing them out to cardholders.
It enhanced the qualifying session, right?
Well, it did, because it sounds like the—I did watch it back. Crofty and Sky were not paying very much attention to Max. The F1 channel, which I had the radio for, was very onto Max: “Max is looking pretty good. He could do this.”
He's got a shot.
You kind of had a sense it—
Yeah.
Yeah. You had a sense that it was coming. So, yeah, that's my overall takeaway. I also don't really have any desire to do that again. It was really tiring, that long drive every day.
The problem is the best tracks are all out of the way in the middle of nowhere because it's 6 kilometers long. You have to stick it somewhere. So, yeah, it was good. I'm glad I did it. I don't plan on doing it again. And that split second on Saturday—
Wait a second.
—is why you watch sports.
Don't plan on doing it again ever?
No, Suzuka. I don't plan on doing Suzuka again.
Suzuka. Okay.
Yeah.
Yeah.
Yeah.
Yeah. There are 23 other races on the calendar for you to check out.
And I don't think I would—I don't think I want to go to any city tracks. I want the real race tracks where you can walk around. That's where the cool stuff is.
The one thought I have before we close out: It is really interesting to me that F1 has its own sort of business line, basically bringing people inside the pit lane over and over again throughout a weekend and charging for it. To be clear, it's not cheap. But it's like if the NBA was selling tickets and allowed you to walk in the locker-room hallways. You don't actually go in the garages, but you're right there.
There are much more expensive packages that let you into the paddock. It's all selling access. This was definitely way too expensive, but also very pleb-level, just to preserve some degree of—
Right.
—common man this year, to be clear.
Yeah. No, there's some access you get, like—
Right. There were truck rides. On Friday there were truck rides between practices for the more expensive people. They didn't have to go on Thursday. We had to go on Thursday because we had the low-end package.
But it is—I mean, talk about the value of experiences. They've turned that into a pretty robust business over the last several years. They certainly charge through the nose in America for Miami, Las Vegas, and Austin as far as F1 experiences are concerned.
But Ben, perhaps we'll have to experience Austin together sometime soon. You can walk around down there.
No, for sure. It's on the list. Yeah, because that's the one—my impression of Austin is that it's kind of like the U.S. Suzuka in many respects.
Yes.
In that, number 1, it's a highly regarded racetrack. Number 2, it's kind of a pain to get to. But number 3, the reason why it's so great as a fan experience is the just-go-there, wander-around, go-wherever-you-want bit.
Mm-hmm.
Which is definitely the best part.
Yeah. We'll need VIP passes. I'm not going to wait all day to get up in that tower, but I am going to make it into that tower one way or another. That's my goal. For now, we will be back on Monday and discussing the state of the world, and also discussing some more traditional tech topics, I would assume. We have a lot of emails to work through in the weeks to come here, Ben. But it's good to see you, it's good to be back, and enjoy the weekend. We'll keep it rolling next week.
Sounds good. Talk to you later.