Post New

Today’s WSJ has an AI story, nominally about CoreWeave, which details the reasons they think the investments in this sector is massively overdone: (Gift link below)

Spending on AI Is at Epic Levels. Will It Ever Pay Off?

Tech companies pour hundreds of billions into data centers, taking on heavy debt, but current revenue is relatively tiny; echoes of dot-com bubble


https://www.wsj.com/tech/ai/ai-bubble-building-spr...

Items:

The artificial-intelligence boom has ushered in one of the costliest building sprees in world history. Over the past three years, leading tech firms have committed more toward AI data centers like the one in Ellendale, plus chips and energy, than it cost to build the interstate highway system over four decades,

Flooded with money from Wall Street and private-equity investors, it has metamorphosed into a computing goliath with a market value bigger than General Motors or Target.

Chips in an AI center have a useful life of between 3-5 years before they are outdated.

[This is my favorite:] This week, consultants at Bain & Co. estimated the wave of AI infrastructure spending will require $2 trillion in annual AI revenue by 2030. By comparison, that is more than the combined 2024 revenue of Amazon, Apple, Alphabet, Microsoft, Meta and Nvidia, and more than five times the size of the entire global subscription software market.

[Much of the story is on the fragility of CoreWeave, which leases data centers for up to 10 years, but whose clients only sign use leases for 2-5 years. If/when there’s weakness, this could be a problem?] = “If the wave of building proves far more than needed, or if tech companies pivot away from third-party providers, the risk is that CoreWeave’s data centers could end up like the dormant fiber optic cables that snaked through the U.S. in the 2000s.”

Other issues:

An MIT report found 95% of organizations surveyed are getting no return on their AI product investments. A University of Chicago economics paper found AI chatbots had “no significant impact on workers’ earnings, recorded hours, or wages” at 7,000 Danish workplaces.

OpenAI’s release of ChatGPT-5 in August was widely viewed as an incremental improvement, not the game-changing moment many expected. Given the high cost of developing it, the release fanned concerns that generative AI models are improving at a slower pace than expected.

Each new AI model—ChatGPT-4, ChatGPT-5—costs significantly more than the last to train and release to the world, often three to five times the cost of the previous, say AI executives. That means the payback has to be even higher to justify the spending.

Another hurdle: The chips in the data centers won’t be useful forever. Unlike the dot-com boom’s fiber cables, the latest AI chips rapidly depreciate in value as technology improves, much like an older model car.

“This is bigger than all the other tech bubbles put together,” said Roger McNamee, co-founder of tech investor Silver Lake Partners, who has been critical of some tech giants.

So the question: how to play it?

"Chips in an AI center have a useful life of between 3-5 years before they are outdated."
"So the question: how to play it?"

Is NVDA the big-dog for data center chips ? But that is probably too obvious,
as the NVDA valuation ( P/e) will be high. Reading that there are high costs borne
by the local population wherever these data centers go in. Ground-water to help
with cooling all of those racks of equipment, electrical bills going up for every
consumer ( although that is happening everywhere ), heavy light and noise pollution
from the data center infrastructure ( but noise might go down after the heavy
construction phase is complete ). So there will be mounting costs for these
data centers. Bribing the local pol's seems to be relatively cheap, but gotta think
there are going to be a lot of costly lawsuits coming from grassroot opposition.
Data centers might become the modern nuclear power plant, fierce grassroot
opposition.

I think I'll just be content to participate via QQQ, SPY, and OEF. I have a "fun"
position in NVDA, just to track it and as a reminder to kick myself in the butt for having traded in and out 3 or 4 times,lol.

Spending on AI Is at Epic Levels. Will It Ever Pay Off?

Tech companies pour hundreds of billions into data centers, taking on heavy debt, but current revenue is relatively tiny; echoes of the dot-com bubble

Goofy,

I have just finished reading three stories on MSN, the portal site. AI-generated, horribly written, few-paragraph stories. Microsoft is hurting itself and paying for the honors.

Bribing the local pol's seems to be relatively cheap, but gotta think
there are going to be a lot of costly lawsuits coming from grassroot opposition.

My wife is on the data center construction team for a big tech company. Coincidently, one of her jobs is to bribe the local politicians. And yes, in the grand scheme it isn't that much. Drop a few grants on the local school system, some bike paths, maybe a conservation easement or two. But tech companies are very cognizant of the impact they have on these small towns and really want it to be a positive experience for the locals. After all, once they finish this one, they'll want to build five more right next to it and they need the town to be on board. And they want the next town over to hear good things and sign on as well.

So for example, they do things like take over road maintence for the town during construction. That way there is no contention about construction vehicles causing potholes or anything like that. They also bring in their own fueling stations and grocery stores because they don't locals to walk in the A&P and find empty shelves or be unable to buy gas. But they let the locals decide how much of that they do. They get a little sales boost but not so much it disrupts the town. Lots of stuff like that. They don't want to cause any problems.

The real pinch point though is lack of skilled workers. The number of workers they think they will need in the next few years is several multiples more than actually exist. Besides paying great wages, this has lead to an arms race between companies of who can provide the best job site. For example, the first thing they do is pave the site, so no one is parking in the mud or tripping over rocks. No porta-potties on site. They provide bathrooms with hot running water for everybody, not just the bosses. Some sites cater lunch for everybody on Fridays. That's just the tip of the iceberg. Because again, after this site is done, they have five more they need workers for.

While water can be an issue, they know exactly how much water they need for construction and operation so that all gets worked out ahead of time. The real problem is lack of power. There are lots of locations where they'd like to build, but they can't because there is no extra juice. They are desperate for power and can't wait for the utility companies to provide it. That's why you hear these crazy stories of tech companies investing in unproven tech like small modular reactors. I don't know how this ends but I expect it to get a lot wilder before it is over.

They are desperate for power and can't wait for the utility companies to provide it.

Let the utilities provide what is needed--BUT they are not allowed to raise rates on existing customers. That way, any increase in rates due to the data center is charged to the data center or absorbed by the utility.

Post New

Macroeconomic Trends and Risks FAQ

---

Contact Shrewd'm
Contact the developer of these message boards.