저자: ChamathPalihapitiya · 외국인
요약 보기
AI 모델은 과거 '우리는 이미 아이폰의 n번째 버전을 쓰는 단계'로 비유되며, 대부분 용도에서는 저가 모델로도 충분하다는 주장. 고가 모델(Claude)은 사이버보안 등 특정 고수익 용도에 한정 권장. 넷플릭스는 라이브 스포츠가 촉매라는 견해, 스트리밍 시장은 포화 상태. 엔비디아는 횡보 후 돌파 가능성과 상대적으로 매력적인 밸류에이션을 언급하나, AI 투자수익률 검증은 어렵다는 논지. 거시 지표(실업수당, 소매판매, CPI 등)는 연준에 긍정적이라는 평가.
다듬은 글
Instead, I think the analogy now is this: we are already seeing the nth version of the iPhone. We all keep updating it. But we still use it for roughly the same purposes. My elbow hurts. What should I do? And you ask yourself: why am I spending so much money? I think this is exactly the stage we are at now .
So if you're going to direct all this behavior, why not just spend 50 cents or a dollar? This is quite rational. But then the question arises: is the Meta model, for example, or Grok, good enough or close enough in capabilities for what you need to do? For most cases, the answer is a resounding "yes." For some very narrow tasks, no.
But in those narrow cases, I would say: spend that $50 per barrel of oil. Spend it. That makes sense. If you're a cybersecurity company like Palo Alto Networks and can generate billions of dollars in additional revenue by protecting the infrastructure of large companies— use Claude. But the panic over the weekend, even within Anthropic — we talked about this yesterday morning — Anthropic, as you know, was planning to end the free use of Claude for some of its paid users.
In fact, canceling this extension would force you to upgrade or pay more for access. We can't afford it and suddenly they say to themselves: we can't afford it. We can't do that because if we do, people will unsubscribe and go to OpenAI. That's right. And the OpenAI model is great. Claude is also great.
Yes. Anthropic models are great when you can use them. So, all these companies face the same problem. They are severely limited in energy resources. They are severely limited by data center capacity and therefore limited in use. Meanwhile, the guys who have all these resources —Google, xAI, and Meta—are finally getting back on their feet and starting to consistently release model after model that are 80-95% as good.
This is a very complex dynamic for leading laboratories. By the way, do you see any changes? You know, for a long time, OpenAI and Sam have been under a barrage of criticism. There were many questions about trust and so on. And, by the way, perhaps even more questions given the situation with Apple on Friday.
Um, but it looks like Anthropic's reputation is starting to suffer a bit too. And this may be about Alex Carpet, but people who think that either the costs or the math of what's happening doesn't make sense, or you look at the story with Figma, for example, about how they're getting into different businesses.
Really confusing. I mean, listen, let's be honest. There are three tech CEOs who have been in the spotlight for 20 years, as I said, they don't provide as much information as they used to , Jen. So, there's not much that could excite either side, Jen. I know. I think it was an interesting move, and maybe it will be beneficial in the long run, but to cut some of those subscriber metrics, because as we know, Wall Street doesn't like less information.
So it looks like we're still feeling that pain of adjusting to a different type of reporting structure from Netflix. And I completely agree. I was so skeptical about their live sports content. And then of course they've done NFL games on Christmas for the last two years, and it's been great. So ever since then, I've been a believer in live sports streaming for Netflix.
It's just that, as we know, they are expensive, but they give great rewards, and something like this, in my opinion, is the catalyst they are missing. In terms of content right now, it seems like it's an extremely saturated space. And unless you're showing a lot of growth or have a series that attracts a significant number of users, it's very difficult to stand out in that environment.
Is it really just that we, as consumers, are now experiencing subscription fatigue, and the US markets are, frankly, relatively saturated? The situation abroad is also not developing quite the way Netflix tried to present it. I think once you see live sports , and remember, linear television, which is declining rapidly, still has a foothold in sports, and Amazon is starting to take its first steps into Thursday Night Football and stuff.
You see Netflix and what they've done. I think this is the trend. But when will one of these streaming services make a big splash and shell out the necessary money? ESPN, all the big networks—they're all competing. And I think the next big thing for Netflix is live sports . Streaming is not going anywhere.
It may decrease. It may stabilize with the arrival of new players, but it will not disappear. Jen, regarding linear TV, I'm not sure. As you know, people in my age group are getting older, and linear TV, streaming, is becoming increasingly popular among younger viewers. I believe that streaming is the future, not linear television.
And when will they make the big leap to dominate live sports broadcasts? I think this is the next hurdle for streaming services. I know when we'll have to streamline our live sports viewing . I mean, I read something. It costs about $1,100 to stream an average NFL season if you want to watch every game on all platforms for a year.
Of course, a season is not a year, but it's just as pointless as watching live sporting events now . It's quite difficult if you want to watch the entire NFL season. But, despite everything, let's move on to some of the economic data we received this week. Obviously, we saw some improvement in unemployment claims yesterday.
Retail sales are a bit more modest, but the CPI and CPI are a good indicator of inflation. Now, housing starts and permits, as well as imports and exports. So, does this change the overall picture? Because I'll say this: If you were the Fed this week, I think you would be pretty pleased with some of the economic data that we got.
to analyze. But beneath the daily volatility, one company may be quietly preparing for a significant move, and that company is Nvidia. swings in the semiconductor industry. SK Hynix is falling sharply, and suddenly the entire chip sector is feeling the pressure. Memory maker stocks are turning around, while AI infrastructure companies are selling off.
Then another positive data appears, and money immediately starts flowing back into trading. This volatility has created an unusual market where the dynamics of a few semiconductor stocks can influence sentiment in the broader technology sector. But Nvidia may be starting to separate itself from this noise.
Some investors are now looking at these stocks from a completely different perspective. Not because Nvidia's fundamental story has suddenly changed. The demand for computing for AI has long been well understood. Instead, the argument becomes purely technical. After a long period of consolidation, Nvidia may try to break out of this range, and if momentum starts to build, the next move could happen quickly.
We've seen this model before with other mega-capitalization tech companies. Stocks can move sideways for months while investors become frustrated. Fundamentals remain strong , earnings continue to rise , but the share price refuses to respond. Eventually, momentum returns and the stock overcomes the resistance level.
Quantitative strategies are starting to catch on to this movement, and algorithmic buying is on the rise. Impulse investors return, and suddenly a stock that seemed stationary begins to grow significantly. Some investors believe that this is exactly the scenario that could be developing around Nvidia now.
Interestingly, this potential breakout comes as Nvidia's valuation has become significantly more attractive relative to its growth profile. This creates a powerful combination: strong fundamentals, massive opportunities for long-term growth, and improving technical momentum, although the overall story around AI is becoming more complex.
Investors are increasingly asking whether the hundreds of mill yield sufficient returns. This is a perfectly valid question, but assessing the return on investment in AI may not be as easy as investors expect. There probably won't be a separate line item on the income statement for " return on investment in AI"; Instead, the benefits of AI will be evident across the business: better productivity, higher advertising engagement and faster growth of cloud services, improved search results and more efficient software development, better recommendations, lower operating costs, and increased customer retention.
AI can improve dozens of different financial metrics, even if companies can't pinpoint exactly how much revenue came directly from AI.
Comments 0