저자: 미상 · 외국인
요약 보기
AI 설비투자는 앞으로도 강하게 지속될 것이며, 구글(알파벳)은 AI 주도권을 쥐고 있어 자금조달도 공세적 행보로 해석된다. AI 도입에 따른 고용 영향은 아직 거의 없고(상관관계 0 수준), 일부 기업의 감원은 AI 핑계로 보인다. 엔비디아는 강한 실적에도 주가 우려가 있으나, 추론·엔터프라이즈 수요 확대로 AI 팩토리 수요가 계속 늘어 성장 여력이 크다는 주장이다.
다듬은 글
That's a smart way to do it. And obviously, you know, you you see Berkshire as another, you know, sort of backer of it. Look, this I mean, we're talking about a trillion dollars is going to be spent in big tech next year in terms of cap backs. So, you know, if this is going to be If you look at how tech companies are doing this partial debt, partial stock sales, I actually think the market's shrugging this off pretty well relative.
I think it just shows right now you know, you go back a year ago, New York City cab drivers bearish on Alphabet. Look today. I mean, they are in a phenomenal position, you know, when it comes down the clouds and when it comes to Gemini. And I think it's a smart time to do this. You want to see them go on the offensive, not defensive.
All right. So, your point about the timing. Uh earlier we heard from David Solomon or Leslie Picker doing a great interview. He's saying right now there's more greed than there is fear in the market. So, actually I I I emailed with Alphabet. They say they 40 billion of this is going to start after July 1st as opposed to right now.
I mean, are you concerned that maybe sentiment could change? Solomon also talked about the lens that people are viewing these AI investments, that could potentially change. I know it's only a month. I know everybody's bullish, but these days a month seems like a very long time. Are you concerned at all that sentiment may change when it comes to the AI trade and the AI build-out over that time?
Look, and I get like obviously some of the the concerns out there, but it's my view we're still third inning of the AI revolution, right? Relative to where this nine-inning game's going and you know, we we could definitely have sort of ebbs and flows in this market, but when you look where Alphabet's position, I mean, they are They're kind of front and center.
I mean, they're they're top of the mountain right now. Amazon clearly has made up, you know, sort of narrowed the gap. Microsoft, I think, you know, obviously is doing a lot of great things in terms of Azure. But Look, Alphabet can't think about weeks or months. We're talking about years. We're talking about fourth industrial revolution in terms of what they're going after.
I think you take the burning comments. Well, but but let me also frame it this way with Google's equity raise and everything. Everyone is really wants to know what impact is AI actually having on the workforce party. What can you tell us? So, one, the amount of spend is staggering. I think it'll persist.
You know, it's adding close to a percentage point. I mean, if you imported adjusted, it's a little less than that. But whatever, even if it's a half a percent, it's still adding very meaningfully. When we look at the actual impact on jobs, the impact right now is like it's just it's it's non-existent at all.
And so, if we look at like AI usage by sector and then we look at the change in jobs by sector, and you put that like on a scatter chart, as I've been fond of saying, you know, it looks like I sneezed on the page, right? Like so the dots are everywhere. No, the R is basically zero. may have been cutting jobs.
Some may be adding, but it Sam Altman yesterday was on this program and he said the companies that are saying they're cutting jobs because of AI, it's a slight of hand. Yeah, so there's no question that there are some companies, you know, in a company-specific way that are probably seeing that effect, but in the aggregate, again, at the sector level, it's just not showing up in the data not at this point.
I assume you went to college. No. Of course I did. Yes. I assume I know you went to college. Anything is possible. I have, you know, three two two and a half three college degrees, so I may go back for a fourth, who knows? My point is this. You got to think about the value add of that. Well, we're being we're being a little bit snobs up here, and I don't mean here on the show, I mean in the media in general because what I think we're missing, and you've written about this, is this idea that yeah, maybe the mid-level office job is going to get disrupted by AI.
Let's hope it's not, but it might. But guess who's winning? The trades. Trades people making can make 150, 200, 250,000 dollars a year without a college education because of all the capital spending going into the cycle, and you don't think that spending is going to slow down anytime soon. No, I don't I don't think it does at all.
In fact, even speaking with our you know, the equity folks that look after tech, you know, they expect that in next year your your your starting point for next year is this dollar amount today. Right? Like it's it's not going to move all that much. Um if if I think if you want to make an argument for where you're starting to see maybe some of this show up, um you know, you can look at the spread between college entrance into the labor force versus everyone else.
And that spread is actually starting to widen, meaning college entrance are having a harder time finding new jobs. So So you can make an argument that there's some of that happening now, but it's way too early. Like people that are saying, "Hey, it's it's the productivity that we've seen over the last, you know, sort of sort of year or so, it's all productivity." It's not.
Um it's Companies basically scaling back on hours worked because they had bloated hours, you know, in the sort of the post-COVID But there there was and I'm sorry to jump in on this, but it hits home. Um because you and I are both went to rural Virginia high schools. The majority of my high school, probably yours too, did not go to college.
In fact, about 90% did not. Most joined the military. God bless them, by the way. And about 20 years ago, 15 years ago, you remember this quote, somebody was like, "Well, truck drivers should just learn how to code or something like this." Guess what? Maybe the coders should now learn how to drive trucks.
No, that's totally reasonable. It's totally reasonable. I think it I think in the end, you know, I always think I I always use the example of my grandmother, right? My grandmother was a telephone operator, right? Like, you know, you hit zero and someone would pick up the phone. Yeah, there you go. Um you know, back at its peak in like the '50s, '60s, there were 350,000 telephone operators.
Um today they're effectively zero. Um but the labor force has done nothing but grow. So I think it's a I I think that you're not appreciating the dynamism of the economy to say that, you know, old jobs are basically going to face The reason this matters is because we are going to have these competing narratives, especially in an election, a midterm year, and all of that.
So if if Bert Basically, Bernie Sanders is positing that AI itself is not going to benefit people, therefore they need direct payments from AI companies. And so if you're telling us, "No, AI will benefit people." This is the argument Dave Cody was making this week as well. He said it's going to increase productivity, raise living standards.
In that case, it's harder to make the case that you need to create an additional vehicle by which to to sort of have these direct payouts. Yeah, I think that's totally fair. And by the way, I think you also have to throw the inflation part of that conversation into the mix. Um because I think about someone like Kevin Warsh, um, who keeps on saying, well, you know, it's going to be disinflationary.
Um, you know, all evidence to the contrary right now. I mean, AI has actually been more inflationary, which again is just a sort of another tax. Why is that? Because it's just raising the cost of components and Hey hy the market is still doesn't fully understand Nvidia. One of the most interesting thing happening right now is that Nvidia continued delivering extraordinary results.
Yet, many investors remain skeptical. The skepticism is understandable because the stock has already appreciated significantly. The company is the one of the largest in the world. And many investors naturally assume that growth must slow down. But, when I look at the fundamentals, I see something very different.
I see a company operating at the center of the most important technology transition of our lifetime. The key question is simple. What's happened if AI adoption accelerate faster than expected? Most valuation models assume the same level of the moderation, some level of normalization, and some level of slowing down.
But, what if demand remain exceptionally strong for a much longer? That's the scenario many investors many industry participant appear to be preparing for. The hyperscalers are not slowing spending. The leading AI labs are not slowing spending, and sovereign AI projects are not slowing spending. The enterprises customers are the only beginning their adoption journey, and that's an important part.
Enterprises AI adoption is still in the early inning. Many large organization are still experimenting, testing workflow, and evaluating evaluating the returns, deploying the pilot programs. The majority of the enterprises have not fully integrated AI into their operations. When that eventually happen, enterprise demand could increase dramatically.
Every employee become a compute consumer, and every workflow become a compute consumer, every software product, and every AI agent become a compute consumer. The scale of demand is a difficult to fully appreciate because we are not talking about millions of users. We are talking about billions. This is why Jensen Huang repeatedly discussed the concept of AI factory.
Traditional factories produce physical products. AI factories produce intelligence. Tokens become the output and just like industrial production expanded throughout the 20th century, intelligence production may be expand throughout the 21st century. This is an incredible powerful framework. The more useful AI become, the more intelligent society consume.
The more intelligent society consume, the more computer is required. And the more computer required, the greater the demand for AI infrastructure. Everything ultimately lead back to the computer. This is why so many discussion around the industry eventually converges on the same conclusion. Supply demand constant and demand remain strong.
Capacity remain valuable. The market often focuses on competition. Custom assets, internal chips, alternative architecture, and those technology absolutely have a role to play. But investors sometimes forget that Nvidia does not need to own 100% of the market. The market itself is expanding rapidly.
Even if competitors succeed, the Nvidia can still grow sustainably. If the total addressable market expand from hundreds of billions to multiple trillions of dollars annually, then multiple companies can succeed simultaneously. That's exactly what we are seeing. The industry is not fighting over the fixed pie. The pie itself is growing and Nvidia remain one of the largest beneficiaries.
Comments 0