저자: TerrySmith · 외국인
요약 보기
Terry Smith은 AI가 생산성을 높인다는 단순한 결론에 회의적이며, 현 시점에서 승자를 확신할 수 없다고 주장한다. 엔비디아 보유자는 잘하고 있지만 엔비디아가 AI GPU의 확실한 승자인지는 모른다고 본다. 과거 기술 붐에서도 초기 선두주자들이 실패한 사례(인텔, AOL, 야후, 블랙베리, 마이스페이스, 노키아)를 들며, AI 수혜가 경쟁우위로 이어질지 불확실하다는 입장이다. 어도비도 경쟁우위 가능성은 있으나 사실이라 보기 어렵고, 모델 보유사가 이익을 가져갈 수 있다고 본다. 전반적으로 AI 관련 확신에 찬 투자 논리를 경계한다.
다듬은 글
You have access to AI through Microsoft, access to weight loss drugs through a stake in Novo Nordisk. I would like to know your thoughts on these topics. Of course, last year you had a huge tailwind for all stocks in this industry. This year, the situation is still evolving. This is one of those situations where the wind is at your back and you say, can we go get some tea while you enjoy the ride?
Um, no, I think those two things are quite different . I don't think they have much in common, except that both events happened last year and are large-scale. Let's take AI. Of course, we have seen a huge surge in interest in AI. This has prompted some companies, both within the "magnificent seven" and outside it—sometimes to our benefit, sometimes to our shame, and so on.
I wrote about this in my letter, as you probably know. I think it's too early and too simplistic to say, "Oh, yeah, AI is going to be something big. " This will increase productivity, so you should own it, and the best options for investment are Nvidia, Microsoft, etc., Adobe. If you look at the development of technology over the last 30 years and the early beneficiaries in terms of stock market performance, and you would say, "Okay, I'll just pick this first winner." God, you would buy some bad stocks: Intel in semiconductors, America Online in internet services.
Yahoo in search engines— they were really early leaders. Research in Motion with Blackberry in smartphones, MySpace was there before Facebook. Oho. Nokia in portable devices. So I don't think we know anything for sure right now, although, listen, those who own Nvidia are doing great, right? I'm happy for them.
Do I think they know that Nvidia is the clear winner in the field of GPUs for AI? No, I don't think so. That's right . I don't believe it. I also don't think we know yet how AI will work, because we know that we have big language models , generative AI, and we know that there are certain companies like OpenAI that are building it, and we know that there are people like Microsoft that are using it in their operating system, making it available within their OS, and so on, and that's great.
What we don't know right now is whether this will provide any competitive advantage , right? Because for something to lead to overall growth and other things. It should provide some competitive advantage to those who have it, compared to those who do n't. Well, I'm not entirely sure what that is . The analogy I brought up again in the annual letter is borrowed from Warren Buffett: artificial intelligence could be like the way we are all in a football stadium .
We all sit until the attacker runs into the penalty area with the ball, and the second row stands up to see better. Well, in 30 seconds, every row stands up, because the front, the second row, and we all have the same view as before we stood up, only we became less comfortable than before. And sometimes some developments are just like that.
The idea that this will bring victory, in my opinion, is not something guaranteed. One of the people Julian and I really respected when we worked together as banking analysts in the 80s was Sir Brian Pitman, the CEO of Lloyds, who created enormous value. Well, the numbers are impressive even by today's standards, considering how early he ran the bank.
And he steadfastly refused to participate in the "big bang," to buy a broker or a job or a bank that was growing , and to get involved in big banking matters. And he explained it to us by saying that some markets don't create winners. So, thirdly, even if it's ... even if it's not, even if you think you have great foresight, you can identify winners, and that will create value through productivity and so on.
What I don't understand is that part of the hype was around things like Adobe. Adobe doesn't have generative AI, a big language model, right? And they don't actually provide it through infrastructure like Microsoft does. They're just going to use it. Well, maybe this will give Adobe an advantage, although I think this is far from a fact.
But wouldn't those who I just mentioned, well, the owner of the big Open AI language model and infrastructure, wouldn't they charge them for it? So, whatever advantage there is, maybe there will be, but the idea that people know what it is at this stage is just wow .
Comments 0