Live data from Hacker News

Google to pay SpaceX $920M a month for compute capacity at xAI data centers

cnbc.com

951–960 of 1001 posts

Re: Google to pay SpaceX $920M a month for compute capacity at xAI data centers

#951

Earlier quoted context omitted.

Active management in general is a poor idea. You'll get better risk adjusted returns by investing in total world equities (like VT). Check out Bogleheads to learn the basics. If you want to get more advanced, you can learn about factor investing as well, but VT is enough for the vast majority. If you want to get intuition for why this works , this is a really fun and interesting video: https://youtu.be/TQuxVz52w2w

The boglehead approach has worked fantastically for ~40 years, but now that everyone is doing it, it may no longer be the case going forwards. Normally with these things when absolutely everyone is crowded on one side of the boat, you want to be on the other side

What exactly is the opposite side? Is it actively managing a portfolio? Because most people don't have the time to do that.

Re: Google to pay SpaceX $920M a month for compute capacity at xAI data centers

#952
post #887

Earlier quoted context omitted.

As others have pointed out, bonds are barely (or not) keeping up with inflation. I would like to suggest a third alternative to stock index and bonds: stable dividend stocks. They should increase in value along with inflation but still pay out a steady dividend as long as the company is strong.

Very different risk profiles. Bonds are about steady cash flow, not about total return. "stable" dividend stocks are almost never really stable when the financial world crashes.

    > "stable" dividend stocks are almost never really stable when the financial world crashes.
Completely agree. Also, many div stocks are just one industrial accident or scandal away from a huge drop in their stock price. People who tout preferred shares are in a similar camp in my opinion. As we discovered in 2008/2009, during a crash, there is no where safe except cash. Suddenly, all financial assets have a correlation of 1.0.

Re: Google to pay SpaceX $920M a month for compute capacity at xAI data centers

#953
post #888

Earlier quoted context omitted.

It is precisely what the video says. Ben has discussed this multiple times as well, not just in this video. If you have better behavioral tolerance for volatility (as in you're not the type to panic sell), total market equities will outperform and lead to less failure in retirement.

While partially true, that "If you have better behavioral tolerance for volatility" is HUGE. Most people can not do this. Once they see their net worth go from $x to $x/2 or worse, they panic sell. People are emotional beings and it's very very hard to not let your emotions dictate what's going on. If you haven't lived through a market panic and crash(last one in the US was 2008/2009), then chances are you shouldn't…

    > Bonds provide stable cash flow. Equities provide growth/return. Use both in the appropriate amounts for your situation.
This is sound advice. I want to add some nuance about "bonds": Consider some broad categories: (1) regular gov't bonds, (2) inflation protected gov't bonds, (3) investment grade corporate bonds, and (4) high yield corporate bonds. In category (4), it is possible to get both cash flow and capital appreciation. It is the bond-equivalent of "stock picking".

Re: Google to pay SpaceX $920M a month for compute capacity at xAI data centers

#954

Earlier quoted context omitted.

Bus Tesla isn’t in anything looking like a growth phase

2024 total deliveries: 1,789,226 2025 total deliveries: 1,636,129 8% decline YOY, Tesla shut down production of Model S & X. Eventually, they will become a pure speculation as a service stock, with zero production. But its cheaper to produce than bitcoin, no energy needs to be expended, runs on pure Musk energy!

I feel that the value of their supercharger network is overlooked. No one outside of China is even close to their global footprint. It would probably be a great business to spin out of Tesla if they pivot away from building cars. When I watch YouTubers talk about the trouble of getting their EVs charged, many complain that competitors are much worse on a variety of metrics. Time and time again, they favour Tesla. And these are not Tesla fanatics -- many don't even own a Tesla.

Re: Google to pay SpaceX $920M a month for compute capacity at xAI data centers

#955
post #818

Earlier quoted context omitted.

2024 total deliveries: 1,789,226 2025 total deliveries: 1,636,129 8% decline YOY, Tesla shut down production of Model S & X. Eventually, they will become a pure speculation as a service stock, with zero production. But its cheaper to produce than bitcoin, no energy needs to be expended, runs on pure Musk energy!

They shut down production of S & X to make more capability for cars that they want to focus on and which sell way more, AND Cybercab. Tesla grows in large steps. Next big step is Robotaxis, which is well on its way. After that, robots, for which they have the best real-world AI platform for. You could say Tesla is a speculation stock as well when they had released the Roadster. Tesla shorters always lose.

    > Tesla shorters always lose.
This is categorically untrue. Look at a chart of their stock from 2020 forward. It has massive spikes up and down. Plenty of shorters made good money in those falls using put options.

Re: Google to pay SpaceX $920M a month for compute capacity at xAI data centers

#956
post #614

Earlier quoted context omitted.

That’s the story, but it’s bullshit. The underlying intrinsic value of a stock can only be materialized if the company liquidates and you receive a share of the sell off of its assets. How many publicly traded companies abruptly decide they’re tired of the business, stop in their tracks, and liquidate their assets? This only really happens if the company is acquired or if it goes bankrupt. Acquisition is the closest…

> The underlying intrinsic value of a stock can only be materialized if the company liquidates and you receive a share of the sell off of its assets. This is wildly incorrect. A profitable company can decide to begin paying out dividends, which can eventually return > 100% of the investor's purchase price. A company can issue more stock or bonds to raise cash to pay investors. A company can spin off assets to raise c…

The grandparent I was responding to said:

>If a stock never intends to pay dividends, the value of the stock is simply the price the next shumck is willing to pay.

So, by construction, we're talking about the value of shares in a hypothetical company that admits it will _never_ pay dividends. And we're asking what value that stock has BESIDES selling it to another shmuck, so for the purposes of the exercise, it's clearest to just imagine we are not allowed to sell to someone else. Most people will tell you that the stock nevertheless still has value because you own a share of the company itself, which entitles you to a share of its liquidation value. However, the argument I've been making here and in other posts are that:

a. A company tends to be "greater than the sum of its parts". The techno-social arrangement of people and business flows is part of what allows the company to be profitable, so disassembling it, selling off the machinery and returning whatever cash assets it had to the investors is unlikely to cover the market cap (at least, as they are priced today in current climate)

b. Even looking at whatever value IS leftover, the circumstances that lead to you realizing that value are extremely fraught / carry other baggage. It usually doesn't lead to common investors getting value back out, and cannot realistically be a justification for the current valuation of most big non-dividend stocks. For instance, consider how valuable it was to own a share of the underlying capital assets of Bed Bath and Beyond when it declared bankruptcy. It was far worse than just point 'a' ("oh no, we sold all the inventory and real estate it still didn't cover the market cap"). No, if you were a common investor, you essentially got $0 because there were lenders and preferred investors ahead of you in line that consumed those assets and left you crumbs.

c. Acquisitions are the best chance of turning your "ownership of the company itself" into dollars... but this is also slightly cheating, because you're appealing to sale of the shares to another entity again. Now, in real life, if a single entity owned the entire company, it would probably be able to extract some of the business's cash flows (a power which common investors lacked). So it's not quite fair to call the acquiring entity "the next shmuck", since they may be able to realize actual $ value in a way that the common investor couldn't. But technically, if we're playing along with the thought exercise, the premise is that the company continually reinvests in itself and refuses to pay out to the owners. If somebody buys out the company, takes it private, and redirects the profits to their own coffers, the new owning entity is essentially getting dividends by another name.

Re: Google to pay SpaceX $920M a month for compute capacity at xAI data centers

#957

Earlier quoted context omitted.

It might be. It might not be. That's where the money is or isn't. I think the mature part of their business is 3/Y, energy storage, and maybe cybertruck, although I also think it's too early to call it because it depends on lower cost cell in house cell production and they only started that recently. In the near term, the growth part is Semi, cybercab, FSD, lithium cell production, and maybe cybertruck. In the long t…

Semi is DoA, FSD has been 1 year away for 10 years now give or take, cybercab is flailing, cybertruck same, and China is eating everyone's lunch on lithium cells.

Why do you think "Semi is DoA"? The current offering for heavy haul electric trucks is tiny (very few competitors), but the addressable market is huge. I think there is a good chance we will be surprised by its success. Even if you dislike the wild hype around Elon Musk (I don't care for it), it is hard to disagree that he has built an incredible EV company. The products they produce are excellent (minus the Cybertruck, too early to say for Cybercab), both from a hardware and software perspective. I think they can do the same for heavy haul electric trucks. The economics of diesel vs electric for heavy haul trucks is a no-brainer. Diesel is much more expensive per kilometer compared to electricity. And maintenance is much cheaper for electric vehicles.

Before finishing this reply, I checked for recent news about the Tesla Semi. I learned that they have a new separate factory (1.7m sq feet!) that has started production and has capacity to produce 50,000 Semis annually. It is next door to the original Gigafactory.

Re: Google to pay SpaceX $920M a month for compute capacity at xAI data centers

#958
post #818

Earlier quoted context omitted.

2024 total deliveries: 1,789,226 2025 total deliveries: 1,636,129 8% decline YOY, Tesla shut down production of Model S & X. Eventually, they will become a pure speculation as a service stock, with zero production. But its cheaper to produce than bitcoin, no energy needs to be expended, runs on pure Musk energy!

They shut down production of S & X to make more capability for cars that they want to focus on and which sell way more, AND Cybercab. Tesla grows in large steps. Next big step is Robotaxis, which is well on its way. After that, robots, for which they have the best real-world AI platform for. You could say Tesla is a speculation stock as well when they had released the Roadster. Tesla shorters always lose.

> They shut down production of S & X to make more capability for cars that they want to focus on and which sell way more, AND Cybercab.

Which Tesla models sell more than the S and X?

Re: Google to pay SpaceX $920M a month for compute capacity at xAI data centers

#960
post #917

Earlier quoted context omitted.

You're acting like the landscape of law hasn't changed and been affected drastically by the purchased Supreme Court justices. That has literally happened. It's not a hypothetical or speculation.

I’m not. But this particular example is so beyond the pale for everyone, that it simply isn’t going to happen.

If I had a few dollars for every time I heard “that’s so beyond the pale, it couldn’t possibly happen”, I would be buying some more RAM for my homelab right now.
Post reply on HN