Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Zitron: The Subprime Datacenter Crisis

wheresyoured.at

Post reply on HN
Print view View original

21–30 of 46 posts

Re: Zitron: The Subprime Datacenter Crisis

#21

Post by discodave »

ElProlactin wrote: post #13

Zitron is literally the worst person to raise alarms about the financials of the AI ecosystem because he's so hyberbolic and pollutes his own arguments with nonsense. Take: > When somebody decides to build an AI data center, they form a special purpose vehicle (much like a CDO), which then raises debt, in some cases slices it into tranches and, in most cases, sells them to institutional investors, asset managers or b…

CDOs and SPVs have two things in common:

They're acronyms and they are financial engineering.

Maybe this time it will be different?

Re: Zitron: The Subprime Datacenter Crisis

#22

Post by simianwords »

camillomiller wrote: post #10
babelfish wrote: post #4

I don't think Zitron will ever admit he's been wrong about LLMs

Because he‘s not?

There are in fact very few things he got right. I can’t point to 2 things he predicted and came out true.

He once said CoreWeave is going bust. Since then the stock tripled lol.

He said ChatGPT will no longer get more users in 2024.

He said that models haven’t improved significantly since. GPT 4.

He predicted in 2024 that OpenAI will collapse in few months to two years unless it creates a totally new form of AI.

Frankly at this point no one should trust his analysis.

Re: Zitron: The Subprime Datacenter Crisis

#23

Post by eru »

When the author leads with a part I know a few things about (the 2000s mortgage market) and gets it so wrong, that kind of makes it hard to put much credence in the point they actually want to make later.

Well, as an escape hatch you could say that the author merely says that the Big Short made these claims (which is true), and not that the claims themselves survive contact with reality. But that would be a lame cop out.

Re: Zitron: The Subprime Datacenter Crisis

#24

Post by simianwords »

camillomiller wrote: post #11
jsLavaGoat wrote: post #2

I thought the AI bubble was supposed to pop in two weeks a month ago?

Nobody with a minimum of credibility says that. Everyone says this is clearly a bubble and yet it could go on another couple of years. Nobody is timing it. But if you want to deny the obvious by numbers go ahead

Ed has timed it.

https://x.com/edzitron/status/1817955630784917548

> Newsletter: Based on estimates of their burn rate and historic analyses, I hypothesize that OpenAI will collapse in the next 12-24 months unless it raises more funding than in the history of the valley and creates an entirely new form of AI.

He posted this in 2024 June btw. He certainly doesn’t believe that OpenAI has a new form of AI. So where’s the collapse? Since then the revenue growth was exponential.

Re: Zitron: The Subprime Datacenter Crisis

#25

Post by eru »

alnxdrawr wrote: post #14

To me the main question is what happens after that.. We all move to frozen open source models running on 2nd hand Oracle/Coreweave GPUs? 10 years before someone dares make another training run? Culturally, do we all collectively sober up once money dries and hallucination are still here? Pendulum swing, AI consideredharmful moment? How to promote healthy use when cognitive surrender is so engrained in us? What happen…

I find that modern harnesses mostly get rid of hallucinations: ask your agent to provide quotes and citations and tests for everything, and have another agent cross check that.

Same way we used to do that as humans: I vaguely remember a case or an anecdote from history, but before I use it in a text, I go and look up whether my memory is playing tricks on me.

Re: Zitron: The Subprime Datacenter Crisis

#26

Post by simianwords »

uriahlight wrote: post #16

Ed Zitron is right about the need for a market correction, but he's fundamentally wrong on AI itself. He'll get his 15 minutes of "see I told you so" when the market corrects, but over the long term he'll be proven wrong. Ed's primary gripe is that he thinks the business models aren't viable for profitability. But Google's AI infrastructure buildout is already spending less then the depreciation value of the hardware…

You seem wildly confident about the market, so I ask.

What market correction are you expecting and by when?

Can you make a falsifiable prediction without moving goal posts and ambiguous timelines?

Re: Zitron: The Subprime Datacenter Crisis

#27

Post by eru »

N_Lens wrote: post #15
ElProlactin wrote: post #13

Zitron is literally the worst person to raise alarms about the financials of the AI ecosystem because he's so hyberbolic and pollutes his own arguments with nonsense. Take: > When somebody decides to build an AI data center, they form a special purpose vehicle (much like a CDO), which then raises debt, in some cases slices it into tranches and, in most cases, sells them to institutional investors, asset managers or b…

The problem is the current media landscape/environment rewards hyperbole and drama. Calm narratives of the facts doesn't find much traction.

Yes, but nothing 'current' about that.

Re: Zitron: The Subprime Datacenter Crisis

#28

Post by eru »

simianwords wrote: post #26
uriahlight wrote: post #16

Ed Zitron is right about the need for a market correction, but he's fundamentally wrong on AI itself. He'll get his 15 minutes of "see I told you so" when the market corrects, but over the long term he'll be proven wrong. Ed's primary gripe is that he thinks the business models aren't viable for profitability. But Google's AI infrastructure buildout is already spending less then the depreciation value of the hardware…

You seem wildly confident about the market, so I ask. What market correction are you expecting and by when? Can you make a falsifiable prediction without moving goal posts and ambiguous timelines?

Or just show us your short position. :)

Re: Zitron: The Subprime Datacenter Crisis

#29

Post by watwut »

simianwords wrote: post #17
Arn_Thor wrote: post #8

Earlier quoted context omitted.

the market can remain irrational longer than I can stay solvent

1. Make a prediction about bubble 2. Get it wrong 3. Excuse yourself by claiming market is irrational

That is the quote of famous financial advisor made in 1986. It keeps being repeated, because bubbles and their pops and irrational markets are frequent enough.

There is nothing controversial about claim that markets are irrational, bubbles pop impossible to predict while bubble being visible. Previous bubbles were the same.

Re: Zitron: The Subprime Datacenter Crisis

#30

Post by simianwords »

watwut wrote: post #29
simianwords wrote: post #17

Earlier quoted context omitted.

1. Make a prediction about bubble 2. Get it wrong 3. Excuse yourself by claiming market is irrational

That is the quote of famous financial advisor made in 1986. It keeps being repeated, because bubbles and their pops and irrational markets are frequent enough. There is nothing controversial about claim that markets are irrational, bubbles pop impossible to predict while bubble being visible. Previous bubbles were the same.

The problem with this type of thinking is that it deliberately ignores the times that bubbles were predicted incorrectly.

Imagine there was a guy. He kept predicting that a war may happen in 1 year. Keeps getting it wrong. In 59 a years a war does happen. He then retroactively justifies his wrong predictions by claiming “but people are irrational and it’s hard to predict”.

Post reply on HN