저자: JeremySiegel · 외국인
요약 보기
연준 회의 전 신용·통화량(M2 유사) 9% 급증을 금리인상 우려 요인으로 제기하며 단기 시장 제약을 언급. 게스트 Brett(ARC)는 AI가 데이터센터 투자→로보택시→새 산업의 4단계로 거시 성장을 견인, 2030년까지 실질 GDP 연 7% 이상 성장을 전망. AI 칩 수요 부족과 반도체 재고 상황을 강조. 마이크론은 100억 달러 규모 메모리 연구 이니셔티브를 발표하며 메모리를 AI 병목으로 규정, 장기 관점 강조. 자본지출 우려와 수익성 훼손 가능성도 함께 언급.
다듬은 글
Who, who is going to pay for all this? But something we should talk about is the Fed meeting this Wednesday. I'm starting to get a little more concerned about the growth of lending and the money supply. Over the past 6 months, it has grown by 9% year-on-year. This is the highest figure since the explosion that J.
Powell mistakenly provoked during the COVID crisis. And it's really hard not to worry about the rate hikes, that's all. Right now I don't think there will be any rate hike at all on Wednesday . There may be a couple of votes against, but if this credit continues to grow, it confirms the scenario that inflation is above target.
So those are three things: if we get some resolution to all of this, then, you know, the bull market is only in its early stages, but I think those are the concerns that are holding the market back for now. Yes, maybe you're right. We constantly debate about AI: are we underestimating its power and what we are witnessing, or are we already too far ahead because there is simply no money right now for many of the developments we are seeing.
Maybe it's somewhere in the middle? Well, I mean, you saw what happened last week. The word " capital expenditure" has become toxic, you know. And the problem is that history says: it often happens that excessive investment reduces profitability. Brett, ARC's Chief Futurist, thank you very much for your time this afternoon.
It's very nice to see you. So you're saying that AI is now a macroeconomic force. Tell us more about this. Give us your vision of this " leakage economy" that we're seeing through all this capital spending. I don't know if I would use the term " leakage economy," but you can understand it through disruptive technologies like AI.
You get the macroeconomic benefit four times over as they spread throughout the economy. The first time is an accelerated pace of investment. Look at the mega- tech companies that used to just sit on cash. Now they are using this money to build data centers. These data centers provide jobs for people who have to dig the ground and install all the equipment.
And we believe that the return on this invested capital is much higher than from other uses. I think , you know, you talked about the core quarter, and I think Wall Street is significantly underestimating the ROI potential of infrastructure as a service right now. So, you get an accelerated rate of return.
You get a higher return. Then you get paid again when you displace competitors or create market activity that did n't exist before. Think about driving. We are about to enter a world of robo-taxis, where robo-taxis will take us everywhere. Now we do 99% of our driving miles manually. This is a waste of our wonderful human brain.
So we will be willing to pay robotaxi services to get rid of this. And this will set us free. The fourth way we get paid is by being able to do other things. Maybe it's watching Netflix or TikTok, or maybe starting a new business. We believe that the initial stages of macroeconomic growth driven by disruptive technologies, which we see today in data center investments, are only the first of four stages in a process that will accelerate the pace of macroeconomic growth.
Oh, I hope, for the sake of our children's future, it won't be watching TikTok in this free time. But since we're talking about AI as a macroeconomic force right now, what's the one economic indicator you think is already changing under the influence of AI? Well, you can already see it in capital formation and capital expenditure.
At the micro level, we can see that firms that are more focused on using AI hire more employees. Therefore, at the level of an individual firm, I believe the narrative that AI will destroy jobs is false. In fact, the opposite is true: firms using AI are finding more ways to provide services to end customers, so they need more people to use these tools effectively.
So I think that over this decade we will see real GDP growth exceeding 7% annually until 2030. We will see a disruption and growth not seen since the advent of electrification, telephony, and the internal combustion engine in the early 1900s. This will be the transformation of macroeconomic activity that we will experience throughout this decade.
And now is the perfect time for this in the market, as people are very skeptical that even an edge data center will generate profits. So I think this is a great situation for equity investors. Brett, the firm took advantage of yesterday's short-term drop in Nvidia shares after that announcement and bought even more.
I'm curious what you think about today's news. Jensen Huang is outspoken about chips becoming a new asset class for investment, and there are new big players in town. Yes, I think there is a huge shortage of computing power in the world. We need much more of these things. Alternatively, people are skeptical about this cycle.
Now, just over 20% of smartphone users use AI chatbots like ChatGPT to ask questions, and we estimate that around 2-5% of knowledge workers use agent-based tools like Codex and Claude Code. So the AI boom is changing more than just the GPU market. One of the biggest changes is happening in a part of the semiconductor industry that investors once considered a highly specialized commodity business: memory, and Micron is making a huge bet that memory will become one of the most strategically important parts of AI infrastructure.
Micron is planning a $ 10 billion research initiative designed to bring together Micron scientists, customers, universities, startups and other ecosystem partners . The goal is not simply to create more memory. It's about developing the next generation of memory technologies for the age of artificial intelligence, and that's important as AI systems become increasingly demanding of memory.
Memory has moved from background to bottleneck . For years, memory was often considered the less attractive part of the semiconductor industry. All attention was focused on the GPU and CPU. Memory has often been treated as something that simply has to be present. Artificial intelligence changes this equation.
Modern AI systems generate huge amounts of data, and this information needs to be stored, transmitted, and processed. As AI models get larger and workloads become more complex, they require more memory, faster storage, higher bandwidth, lower power consumption, and larger storage capacities. That's why Micron's CEO called memory , not just another component.
This is not limited to traditional data centers. The next big wave could come from robotics, automated vehicles, and humanoid machines. A robot capable of interacting with the physical world requires significant amounts of data processing and memory. So the potential market for memory could go far beyond today's AI servers.
Micron is choosing a long-term approach strategy. The most interesting thing about Micron's new research initiative is the time horizon. This is not about next quarter's earnings. It's about developing technology that can define the future of the next generation of computing. The company plans to work in various areas, including materials, semiconductors, memory, computing, and the interaction between them .
This creates a much broader research ecosystem. Historically , some of the most important breakthroughs in semiconductors have come from large research organizations and institutions. Micron is actually trying to recreate this long-term research engine in memory and AI, and it's doing so in the United States.
Why is this important? Because Micron's roots in this industry go back decades. A company that has continued to invest in advanced research and manufacturing in the United States, even during challenging periods in the memory cycle — has become part of a national strategy. The United States wants more domestic semiconductor manufacturing capacity, and AI requires n turn requires a huge amount of memory.
So Micron's investment is not just a story of corporate expansion; it's also part of a broader push to boost the American semiconductor ecosystem. This is where investors should be careful. Micron's management argues that the memory market has fundamentally changed, and there is certainly evidence to support this view.
댓글 0