March 21, 2025 • 9 mins
In this week's monologue, Ed Zitron walks you through CoreWeave, the first IPO of the AI bubble, and how its terrible finances, burdensome loans and questionable partners are cause for concern.
Newsletter: https://www.wheresyoured.at/core-incompetency/
---
LINKS: https://www.tinyurl.com/betterofflinelinks
Newsletter: https://www.wheresyoured.at/
Reddit: https://www.reddit.com/r/BetterOffline/
Discord: chat.wheresyoured.at
Ed's Socials:
https://www.instagram.com/edzitron
https://bsky.app/profile/edzitron.com
https://www.threads.net/@edzitron
See omnystudio.com/listener for privacy information.
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Speaker 1 (00:02):
Zone Media burd Up. I'm at Zetron and this is
your weekly better offline monologue. So last week I put
out a newsletter about core Weave, an a cloud provider
that sells GPU compute to AI companies looking to run
(00:24):
or train their models, and they recently filed the paperwork
to go public. Now, the newsletter explained the shaky fundamentals
of the business, which is putting it mildly and ask
the question how the fuck does core Weave survive. This
is a company that's drowning in debt and relying on one
company for more than half of its revenue. And said company,
Microsoft is actively pulling away from building new data center
(00:46):
capacity and has reportedly pulled back from some contracts with
core Weave, though it's not clear what those contracts are. Now,
this is an important story and one that raises questions
not just about the viability of core Weave, but the
generative AI industry. At that said, the subject matter is
a little dense to turn into a hulking two part episode,
so I decided to condense it into a shorter monologue
(01:07):
for you. Now, Core We've had intended to go public
a week ago. They might go public and a week
from now, and they're apparently going with like a twenty
five to thirty billion dollar valuation. While it's hardly a
recognizable name like say open AI or Microsoft or in Vidia,
this company is definitely worth taking a look at, if
not for the fact that it's arguably the first major
(01:28):
IPO that we've seen from the current generative AI hYP bubble,
and it's undoubtedly the biggest. Moreover, corwave is a company
that deals in the infrastructure aspect of AI, where one
would naturally assume is where all the money really is
putting up service for hyperscals to run their hallucination pro
and nonprofitable models. If in Vidia, the company that makes
(01:48):
the GPUs for generative AI, is selling the pick axes
for the current gold rush, cor we've well, they own
the land, or maybe they're the shovels. I'm not going
to get too far into that metaphors side up. Until
a few years ago, Corweave was a different company, a
cryptocurrency mining data center company to be specific. Now twenty
tens were turbulent for crypto and the company then pivoted
(02:11):
to providing high performance computing for third parties that didn't
want to run their own infrastructure for things like three
D modeling. Around the time of the launch of chair GPT,
cor We've pivoted again, this time towards providing compute to
hyperscalers and then generative AI interest. It gambled, and it
gambled successfully. You'd imagine that such a company would be
a thriving, healthy business though, right they I mean they
(02:34):
sell the surfer architecture. The demand is there, the growth
is there. I mean everyone wants generative AI. Right, it
wouldn't be another thing where the reality Jesus Christ. Even
a cursory glance at Corwey's financial disclosure documents reveals a
business that's precarious at best, and in my most uncharitable opinion,
(02:55):
it's utterly rancid. If this company was in any other industry,
be seen as a big pile of shit, except it's
one of the standard bearers of the generative AI boom
and so exists within its own reality distortion field. Now
Callw've makes its money by renting out GPUs to companies
looking to run or train their models, which means that
it's incredibly indicative of whether there's a real business in
(03:18):
generative AI and you'll be shocked to hear that there
isn't call. We've made just under two billion dollars in
twenty twenty four, yet somehow managed to lose eight hundred
and sixty three million dollars. Worse still, seventy seven percent
of its revenue comes from two companies, Microsoft sixty two percent, and,
although not said explicitly, most likely in Video fifteen percent.
(03:38):
Now the information is reported Project Gosprey, which is basically
Nvidia's stake in the fact that they're finaling money and
it's actually not immediately obvious what it is that in
Video is paying Core week four, though Invidio is both
invested in the company and provides it with priority access
to its chips. With Core, we've getting some products before
even Amazon or Microsoft. It's bonkers. And again, yeah, I
(04:00):
think with all this preferential accents Core, we've just printing
money right versus burning it, burning it by the hundreds
of millions full. I wish I could get this much
money and just I don't know, I would be doing
a far weirder podcast. It would be exactly the same thing,
better offline would exist. I would just have hundreds of
millions of dollars we'd have just like Samuel L. Jackson
(04:21):
coming in for some episodes. Anyway, I don't have that money.
If you'd like to give it to me, please please
do now. A few weeks ago, The Financial Times reported
that Microsoft had pulled out some of their core Weave contracts,
though Core We've denied the allegations, and mysteriously, a week later,
open Ai announced that they had an eleven point nine
billion dollar contract with Corewave to provide compute services. Now,
(04:44):
I can't say for certain. Can you think of another
reason Microsoft needed a whole bunch of GPU compute other
than to host open AI's models. Hmm, Now, I'm just
guessing here, but one has to wonder if it wasn't
so much open Ai se a new deal and more
at taking over the future compute than Microsoft was formally handling.
(05:05):
And yes, you're remembering correctly. Microsoft has recently canceled more
than the gigawater future compute capacity, though from what I've heard,
its core Weave contracts future or otherwise weren't included in
that number. Anyway, For core Weave to make that eleven
point nine billion dollars in revenue from open Ai, they're
going to need compute, which will cost them tens of
billions of dollars to build, which will be difficult as
(05:26):
the company has burdened with about eight billion dollars of
debt with horrifying interest rates, which made lead it to
pay upwards of one point five billion dollars a year
in loan payments, with one of them requiring Core Weave
to repay the loan with any other debt they raise
in the future. Core Weave, by my estimates, needs at
least thirty billion dollars to expand. It has about one
point three billion dollars in the bank and just under
(05:47):
four billion dollars left on its loans that they can draw.
One of them is a term loan where you can
draw more up until about June twenty twenty five. Although
Core We've will raise capital in its IPO, the amount
will be anywhere near any to meet its needs. And
when I say its needs, I mean like the ability
to service the revenue that they need to spend more
money really does not make a lick of sense when
(06:09):
you think about it, and you may wonder if things
can get worse, and the answer is that they can.
Coreweave has planned one point three gigawats of expansion and
its partner to build it is a public company called
Core Scientific. Different company, similar name, genuinely different. Here are
some facts about Core Scientific. They went public in twenty
(06:30):
twenty two in a disastrous spack merger and then filed
for bankruptcy in the same year. Their entire business has
been focused on mining bitcoin, which requires specialized acic computer
chips that are entirely different both in their construction and
their maintenance to GPUs, and they can only really be
used for mining crypto. They're not easily repurposed for other tasks,
(06:51):
and indeed their data centers are not just a plug
and play thing. You basically have to demolish them. And
another thing about Core Scientific they only made twenty four
million dollars twenty twenty four from selling AI compute related
services to their customers. That's around five percent of its
total revenue, with the rest coming from mining and selling crypto.
Also really dumb ass mistaken. Just make that I said customers,
(07:12):
I meant customer. Course Scientific has one customer and that
customer is it's core Weave. As of right now, Corewave
has approximately three hundred and sixty megawatts of compute power. Somehow,
it intends to build another one point three gigawatts worth
of compute using a partner that has never built an
AI data center and does not be able to have
any meaningful compute right now, this is all so good.
(07:35):
I love this. Let's fucking go. This is the big IPO,
this is the big aiaper. I am going insane every
time I read about these goddamn companies. But let's summarize.
Core Weave is burdened by interest payments that may belloon
to as much as two billion dollars a year, and
lost eight hundred and sixty three million dollars on two
(07:56):
billion dollars of revenue in twenty twenty four. Sixty two
percent of that revenue is from Microsoft, which has materially
pulled back on data center buildouts and may have dropped
some core Weave contracts, though core We've denies this is
the case. Of course, Corby's expansion, which is critical to
servicing future revenue and growth, requires it to invest tens
of billion dollars the core Weave does not have. Corwy's
data center expansion is dependent on what is primarily a
(08:18):
bitcoin mining company called Core Scientific. It appears to have
no AI capacity building, and they're meant to build over
a gigawak capacity a time where they do not appear
to have done so. And as a reminder, converting cryptocurrency
mining data centers to HPC data centers is effectively starting
from scratch. It's not good. It's not good at all.
(08:40):
Corby should have been a positive signal for generative AI,
or at least the way for generative AI to kind
of go further than where it is today, or at
least a way for AI boosters to shut me up.
If GENERATIVEAI had this incredible demand, both from companies looking
to integrate it and us looking to use it, Corweve
would be making fat stacks of cash. I have a
far more diverse customer base and if I'm honest, not
(09:03):
have to take up more than five times its revenue
and burdens and loans with loan shark level interest rates
just to survive. But it is what it is, I guess.
And another note Jim Kramer of CNBC. He said that
Corweve was going to be one of the biggest IPOs
of the year. That should tell you about everything
Zone Media burd Up. I'm at Zetron and this is
your weekly better offline monologue. So last week I put
out a newsletter about core Weave, an a cloud provider
that sells GPU compute to AI companies looking to run
(00:24):
or train their models, and they recently filed the paperwork
to go public. Now, the newsletter explained the shaky fundamentals
of the business, which is putting it mildly and ask
the question how the fuck does core Weave survive. This
is a company that's drowning in debt and relying on one
company for more than half of its revenue. And said company,
Microsoft is actively pulling away from building new data center
(00:46):
capacity and has reportedly pulled back from some contracts with
core Weave, though it's not clear what those contracts are. Now,
this is an important story and one that raises questions
not just about the viability of core Weave, but the
generative AI industry. At that said, the subject matter is
a little dense to turn into a hulking two part episode,
so I decided to condense it into a shorter monologue
(01:07):
for you. Now, Core We've had intended to go public
a week ago. They might go public and a week
from now, and they're apparently going with like a twenty
five to thirty billion dollar valuation. While it's hardly a
recognizable name like say open AI or Microsoft or in Vidia,
this company is definitely worth taking a look at, if
not for the fact that it's arguably the first major
(01:28):
IPO that we've seen from the current generative AI hYP bubble,
and it's undoubtedly the biggest. Moreover, corwave is a company
that deals in the infrastructure aspect of AI, where one
would naturally assume is where all the money really is
putting up service for hyperscals to run their hallucination pro
and nonprofitable models. If in Vidia, the company that makes
(01:48):
the GPUs for generative AI, is selling the pick axes
for the current gold rush, cor we've well, they own
the land, or maybe they're the shovels. I'm not going
to get too far into that metaphors side up. Until
a few years ago, Corweave was a different company, a
cryptocurrency mining data center company to be specific. Now twenty
tens were turbulent for crypto and the company then pivoted
(02:11):
to providing high performance computing for third parties that didn't
want to run their own infrastructure for things like three
D modeling. Around the time of the launch of chair GPT,
cor We've pivoted again, this time towards providing compute to
hyperscalers and then generative AI interest. It gambled, and it
gambled successfully. You'd imagine that such a company would be
a thriving, healthy business though, right they I mean they
(02:34):
sell the surfer architecture. The demand is there, the growth
is there. I mean everyone wants generative AI. Right, it
wouldn't be another thing where the reality Jesus Christ. Even
a cursory glance at Corwey's financial disclosure documents reveals a
business that's precarious at best, and in my most uncharitable opinion,
(02:55):
it's utterly rancid. If this company was in any other industry,
be seen as a big pile of shit, except it's
one of the standard bearers of the generative AI boom
and so exists within its own reality distortion field. Now
Callw've makes its money by renting out GPUs to companies
looking to run or train their models, which means that
it's incredibly indicative of whether there's a real business in
(03:18):
generative AI and you'll be shocked to hear that there
isn't call. We've made just under two billion dollars in
twenty twenty four, yet somehow managed to lose eight hundred
and sixty three million dollars. Worse still, seventy seven percent
of its revenue comes from two companies, Microsoft sixty two percent, and,
although not said explicitly, most likely in Video fifteen percent.
(03:38):
Now the information is reported Project Gosprey, which is basically
Nvidia's stake in the fact that they're finaling money and
it's actually not immediately obvious what it is that in
Video is paying Core week four, though Invidio is both
invested in the company and provides it with priority access
to its chips. With Core, we've getting some products before
even Amazon or Microsoft. It's bonkers. And again, yeah, I
(04:00):
think with all this preferential accents Core, we've just printing
money right versus burning it, burning it by the hundreds
of millions full. I wish I could get this much
money and just I don't know, I would be doing
a far weirder podcast. It would be exactly the same thing,
better offline would exist. I would just have hundreds of
millions of dollars we'd have just like Samuel L. Jackson
(04:21):
coming in for some episodes. Anyway, I don't have that money.
If you'd like to give it to me, please please
do now. A few weeks ago, The Financial Times reported
that Microsoft had pulled out some of their core Weave contracts,
though Core We've denied the allegations, and mysteriously, a week later,
open Ai announced that they had an eleven point nine
billion dollar contract with Corewave to provide compute services. Now,
(04:44):
I can't say for certain. Can you think of another
reason Microsoft needed a whole bunch of GPU compute other
than to host open AI's models. Hmm, Now, I'm just
guessing here, but one has to wonder if it wasn't
so much open Ai se a new deal and more
at taking over the future compute than Microsoft was formally handling.
(05:05):
And yes, you're remembering correctly. Microsoft has recently canceled more
than the gigawater future compute capacity, though from what I've heard,
its core Weave contracts future or otherwise weren't included in
that number. Anyway, For core Weave to make that eleven
point nine billion dollars in revenue from open Ai, they're
going to need compute, which will cost them tens of
billions of dollars to build, which will be difficult as
(05:26):
the company has burdened with about eight billion dollars of
debt with horrifying interest rates, which made lead it to
pay upwards of one point five billion dollars a year
in loan payments, with one of them requiring Core Weave
to repay the loan with any other debt they raise
in the future. Core Weave, by my estimates, needs at
least thirty billion dollars to expand. It has about one
point three billion dollars in the bank and just under
(05:47):
four billion dollars left on its loans that they can draw.
One of them is a term loan where you can
draw more up until about June twenty twenty five. Although
Core We've will raise capital in its IPO, the amount
will be anywhere near any to meet its needs. And
when I say its needs, I mean like the ability
to service the revenue that they need to spend more
money really does not make a lick of sense when
(06:09):
you think about it, and you may wonder if things
can get worse, and the answer is that they can.
Coreweave has planned one point three gigawats of expansion and
its partner to build it is a public company called
Core Scientific. Different company, similar name, genuinely different. Here are
some facts about Core Scientific. They went public in twenty
(06:30):
twenty two in a disastrous spack merger and then filed
for bankruptcy in the same year. Their entire business has
been focused on mining bitcoin, which requires specialized acic computer
chips that are entirely different both in their construction and
their maintenance to GPUs, and they can only really be
used for mining crypto. They're not easily repurposed for other tasks,
(06:51):
and indeed their data centers are not just a plug
and play thing. You basically have to demolish them. And
another thing about Core Scientific they only made twenty four
million dollars twenty twenty four from selling AI compute related
services to their customers. That's around five percent of its
total revenue, with the rest coming from mining and selling crypto.
Also really dumb ass mistaken. Just make that I said customers,
(07:12):
I meant customer. Course Scientific has one customer and that
customer is it's core Weave. As of right now, Corewave
has approximately three hundred and sixty megawatts of compute power. Somehow,
it intends to build another one point three gigawatts worth
of compute using a partner that has never built an
AI data center and does not be able to have
any meaningful compute right now, this is all so good.
(07:35):
I love this. Let's fucking go. This is the big IPO,
this is the big aiaper. I am going insane every
time I read about these goddamn companies. But let's summarize.
Core Weave is burdened by interest payments that may belloon
to as much as two billion dollars a year, and
lost eight hundred and sixty three million dollars on two
(07:56):
billion dollars of revenue in twenty twenty four. Sixty two
percent of that revenue is from Microsoft, which has materially
pulled back on data center buildouts and may have dropped
some core Weave contracts, though core We've denies this is
the case. Of course, Corby's expansion, which is critical to
servicing future revenue and growth, requires it to invest tens
of billion dollars the core Weave does not have. Corwy's
data center expansion is dependent on what is primarily a
(08:18):
bitcoin mining company called Core Scientific. It appears to have
no AI capacity building, and they're meant to build over
a gigawak capacity a time where they do not appear
to have done so. And as a reminder, converting cryptocurrency
mining data centers to HPC data centers is effectively starting
from scratch. It's not good. It's not good at all.
(08:40):
Corby should have been a positive signal for generative AI,
or at least the way for generative AI to kind
of go further than where it is today, or at
least a way for AI boosters to shut me up.
If GENERATIVEAI had this incredible demand, both from companies looking
to integrate it and us looking to use it, Corweve
would be making fat stacks of cash. I have a
far more diverse customer base and if I'm honest, not
(09:03):
have to take up more than five times its revenue
and burdens and loans with loan shark level interest rates
just to survive. But it is what it is, I guess.
And another note Jim Kramer of CNBC. He said that
Corweve was going to be one of the biggest IPOs
of the year. That should tell you about everything
Host
Ed Zitron
Popular Podcasts
40s and Free Agents: NFL Draft Season
Daniel Jeremiah of Move the Sticks and Gregg Rosenthal of NFL Daily join forces to break down every team's needs this offseason.
Dateline NBC
Current and classic episodes, featuring compelling true-crime mysteries, powerful documentaries and in-depth investigations. Follow now to get the latest episodes of Dateline NBC completely free, or subscribe to Dateline Premium for ad-free listening and exclusive bonus content: DatelinePremium.com
The Bobby Bones Show
Listen to 'The Bobby Bones Show' by downloading the daily full replay.