WATCH TAIWAN

“Google sold, I think, a million was it the v sevens, the Ironwoods, to Anthropic? DeepMind people are like, this is insane. Why did we do this? But then Google Cloud people and Google executives saw a different, like, thought process. They saw this dislocation. They negotiated a deal, and they were able to get access to these to this compute before Google realized. We saw capacity on anthropic or sorry, on TPUs go up by a significant amount over the course of those six weeks. Google even had to go to TSMC and explain to them why they needed this increase in capacity because it was so sudden. But that a lot of that capacity increase was for selling to Anthropic. Because Anthropic saw it before Google.”

“But then before the Zhang Youxia's downfall several weeks ago, what we do see, what we thought that is that Zhang Youxia seems to be the person that will be immune from those. But it seems that Xi Jinping just cannot stand with a person that could be powerful or could be potentially challenging his role about the PLA. So I think that about the Zhang Youxia, it's more about how Xi Ming feel that whether Zhang Youxia is threatening his positions or Zhang Youxia overstepped the authority of the PLA.”

“Used to be 1,400,000 smartphones were sold a year. Now we're at, like, 1.1. But our projections are we maybe get down to, like, 800,000,000 this year. And next year, like, 600 or 500,000,000. They see Xiaomi and Oppo are cutting low end and mid range smartphone volumes by half. Today, you already see all the memes, like, on PC subreddits and PC gaming PC Twitter is, like, cat dancing videos. And it's like, this is why memory prices has doubled, and you can't get a new gaming GPU.”

“And the plan is that should the blockade and the energy that really happens, Taiwan probably will reopen its nuclear power plant as the power for our energy supply. And also, we probably will also limit our usage of the natural gas and increase the composition of the coal for Taiwan's energy usage. And the possibility of reducing the production for the semiconductor as well as other energy-consuming industries in order for Taiwan to sustain our defense, that will be an important part of the measure that we have to adopt.”

“In isolation, you almost always wanna go with a smaller model that gets RL'd faster and gets deployed into research and development. So you can build the next thing and get more compute efficiency wins. And then this compounding effect of, oh, I made a smaller model that I r l'd more that I then deployed into research and development earlier, and I spent less compute on the training itself because I was able to allocate more compute to the research. This compounding effect of being able to do the research faster and faster is potentially a faster takeoff.”

“Dario, when he was on your podcast, was very, very, like, conservative. He's like, you know, I'm not gonna go crazy on compute because if my revenue inflects at a different rate, at a different point, I don't wanna go bankrupt. But in reality, you know, he's definitely missed the pooch in terms of, like, going, like, OpenAI, which was, let's just sign these crazy fucking deals. And OpenAI has kind of got way more access to compute than Anthropic by the end of the year. And so what does Anthropic have to do to get the compute? Well, they have to go to lower quality providers that they would not have gone to before.”

“And so you sort of have this, like, dynamic that is quite interesting in that. An h one hundred is worth more today than it was three years ago.”

“And recently, new development is actually in Korea. Not about Korea country itself, but the US presence in Korea. And then the probably China started to realize that the nearest base from the United States toward Beijing is actually a US air base in Korea. That's a new realizations. And I hope that the China will take that into heart.”

“We're looking at the evolution of Chinese policy towards Taiwan. Although China always said that their policy has been very consistent on Taiwan, like they are for the unification. But during the Hu Jintao era, basically China, the focus is on preventing Taiwan from declaring independence. But in the Xi Jinping era, the whole focus is bring Taiwan in, whether Taiwan like it or not, even though time does not declare independence. So Xi Jinping is more for unification and at my own speed.”

“We also noticed that the Russians' activity along with China on Taiwan increases dramatically in the last years. So that if the Russians prevail in its plan on Ukraine, at that time when China decided to attack Taiwan, we are going to face not just one country, like China, but the probably the Russians' involvement, whether directly or indirectly, that's going to be there. So that's double worry for us.”

“If you ship out all the process engineers and assuming it's, like, hot enough that you destroy the fabs, no one has all the fabs in Taiwan now, which is a big risk. These tools actually use a lot of semiconductors, which are manufactured in Taiwan. So it's like a snake eating its own tail sort of like meme because you can't make the tools without the chips from Taiwan, which you can't use without the tools in Taiwan. Just shipping out all the engineers and blowing up the fabs means China has a stronger semiconductor supply chain than the rest of the world.”

“I believe an iPhone has 12 gigabytes of memory. Each gig cost used to cost roughly 3 or $4. That's $50. But now the price of memory is, like, tripled. Let's call if it's now, it's $12 per gig for DDR. So now you're talking about a $150 versus $50. A $100 increase in cost on Apple. Apple has some margin. They're not just gonna eat the margin. So now that's a $100 cost increase. The NAND also has the same sort of, like, market. So in fact, you know, it's probably a $150 increase on the on the iPhone. That means the end consumer is paying $250 more for an iPhone.”

“One of the hedge funds I went to interview at, it was a spin off of SAC that did health care. The company is called Orthodontic Centers of America. I went home and I spent maybe, like, hours and hours going through the financial filings. It hit me that what they were doing was the simplest form of accounting fraud, which is just capitalizing expenses that should have been expensed in a big way. I posted online. Within a few hours, the stock started to go down. Next day, stocks down, like, 30%. That's how I got my job.”

“The pattern recognition, to answer your question, for me, philanthropic was just reading Dario's essays and listening to him on podcast. When I look back at my career and look back at the companies we missed, Amazon in the early days, and I think what could I have seen? The only telltale sign was reading Jeff Bezos' nineteen ninety seven shareholder letter, which was like the clarity of thought. Dario struck me like that. I place a lot of weight really wrongly on clarity of thought and the ability to communicate as a CEO.”

“My wife begs me all the time to stop shorting stocks. It's a bad business. You really have to be intellectually stimulated by it. Most people in the market are just not fundamentally based, period. There are tons of people investing in things that are just based on stories, like, because of social media, because of Robinhood. And there's just endless amounts of shorts if you have duration and if you take a fundamental view.”

“Space data centers effectively are not limited by, you know, hey. We have this energy advantage. It's actually just limited by the same contended resource. We can only make 200 gigawatts of chips a year by the end of the decade. So what are we gonna do to get that capacity? It doesn't matter if it's on land or in space. Elon doesn't win by doing, you know, 20% gains. Elon never wins that way. Elon wins when he swings for the fences and does 10 x gains.”

“The thing that troubles me the most, frankly, is I think we are on a collision course with China over semiconductors. Taiwan produces 90 something percent of the most advanced semiconductors, and everything we use is semiconductors. It's almost as if you went back fifty years, if there's only one country that produced oil. That supply chain is fragile. If that supply chain were to get screwed up, we would have, like, an incredibly bad economy on the order of depression type economy.”

“With my wife right now, I was a pretty bad boyfriend in college and that, like, I was busy doing other things, and I was not very attentive. We broke up, and I remember I sat down with her and we went to get a drink just to catch up as friends. And I said, I got a job at Bear Stearns. And I just remember, like, she just started crying. She just started, like, crying with tears of joy, and I was like, wow. Immediately, I walked out, and I was like, I'm gonna marry that girl.”

“I definitely do not ever aspire to having hundreds of employees or something like that. I think hedge funds are good business. Objectively, our business, like, is horrible. It's amazing cash flows, has no terminal value. I told this to my companies I invested. I'm like, you have no cash flows and tons of terminal value. I have tons of cash flows and no terminal value. So we're good together. Like, we can kinda arbitrage that.”

“When I was speaking to the executives at the LLMs, the way I framed this, I said, look. I think your business is some kind of combination between Netflix and Spotify. Netflix in that, unlike other tech companies, you are spending a ton of money upfront to train these models. Once these models are trained, you go sell them at extremely high incremental margins. The music on Spotify is no different than Apple Music or Amazon Music. Theoretically, it's a pure commodity. What makes Spotify have pricing power? It's because it's personalized.”

“SpaceX was pretty obvious to me that the launch business at a minimum was gonna be a very good business. The proof for reusability and scale, like, yeah, okay. Starship is a game changer, which we knew about fairly early on but didn't know if it would work. And what that means very simply is that the cost of launching everything goes down dramatically, 97, whatever, percent. I think that in a relatively short amount of time, they are gonna be dramatically cheaper than any other form of delivering broadband.”

“It's incredibly difficult. We went from being top of the world. Everyone thinks we walk on water to being, like, everyone thinks we're gonna go to business. The most important moment was we do semiannual investor dinners with our LPs. Jeremy, the president of our firm, he said to me, he said, like, we can't do these dinners. Like, you know, this is gonna be a bloodbath. To me, it was, like, really clear. I said, no. We have to do these dinners. The analogy I gave was, like, we're gonna hit singles and doubles.”

“But when we look at the invasion capability of the PLA, we do not think that before year 2027, they could have that capability because they just do not have the kind of projection capability as well as the air and the sea lift capacity that's necessary for the amphibious operation against Taiwan. However, about the PLA capability, we do think that should they want to launch the blockade, they probably will be able to do that. And as well as the quarantine operation.”

“Yeah, I think the second Trump administration, what we learned so far is that it's very different from the first Trump administration, that all the assumption that we have about the first Trump administration probably will no longer apply for the second Trump administration. And when we look at the National Security Strategy, we do find that there's a reordering about the priority, especially on the foreign affairs, and it is more focused on the domestic as well as the Western Hemisphere.”

“I am more confident in the thesis that the hyperscalers are a worse business model going forward. The problem going forward is that I think that economically, it's highly unlikely that LMs are not very concentrated in the hands of four or five companies. At some point in the next five to ten years, they will be generating enormous amounts of free cash flow. When that happens, I think that they're likely to in source the compute.”

“ASML makes the world's most complicated machine, I e, an EUV tool, and the selling price for those is $304,100,000,000 dollars. And currently, they can make about 70. Next year, they'll get to 80. Even under very aggressive supply chain expansion, they only get to a little bit over a 100 by the end of the decade. ASML's commented their supply chain has over 10,000 people in it.”

“Huawei is arguably the only company in the world that has all the legs. Huawei has cracked software engineers. Huawei has cracked networking technologies. That's, in fact, their biggest business historically. And they have cracked AI talent. But furthermore, beyond NVIDIA, they actually have better AI researchers. And furthermore, beyond NVIDIA, they have their own fabs. It's very arguable that Huawei, if they had TSMC, would be better than Nvidia.”

“Yeah, first talk about the cyber, right? Taiwan experienced like basically the warlike situation by China on Taiwan about the cyber attack. In year 2004, the average daily attack by China on Taiwan, daily, we have about 2.4 million attacks every day by China in the cybers. And then last year, it had been slowly upgraded to 2.63 million attacks a day.”