October 3, 2025 • 28 mins
In part four of this week’s four-part case against generative AI, Ed Zitron walks you through how there’s literally not enough money in the world to pay for OpenAI’s estimated $1 trillion four-year burn, and why the media tell the truth about AI to save retail investors.
Latest Premium Newsletter: OpenAI Needs A Trillion Dollars In The Next Four Years: https://www.wheresyoured.at/openai-onetrillion/
YOU CAN NOW BUY BETTER OFFLINE MERCH! Go to https://cottonbureau.com/people/better-offline and use code FREE99 for free shipping on orders of $99 or more.
---
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
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):
Zoe Media.
Speaker 2 (00:04):
All Right, we're back.
Speaker 3 (00:05):
I'm ed Zetron and this is better offline. This is
the fourth and final episode of this four part series
where I dissect in excruciating comprehensive detail the AI bubble
(00:26):
and tell you why I think its implosion is inevitable.
We're at the end, which is fitting because in this
episode we're going to talk about why and how this
entire sham dies again. If you jumping in now, start
from the beginning. We talk about a lot of things
and it all kind of threads together. I know, I
know you want to hear my explanation of how open
ai is fucked with a capital fuck. But we've got
to lay the groundwork to get there. Okay, got up, good, Good,
(00:48):
We're good. We're god, You're ready, We're good, You're got yep, yep.
Speaker 2 (00:51):
Okay, we'll begin.
Speaker 3 (00:52):
One of the comfortable lies that people tell themselves is
that the AI bubble is similar to the fiber boom
or the dot com boom or uber, or that we're
in the growth stay, that this is what software companies do.
They spend a bunch of money, then pull the profit lever.
The thing is, this is nothing like anything you've ever
seen before, because this is the dumbest shit the tech
industry has ever ever done. AI data centers are nothing
(01:14):
like fiber because there are very few actual use cases
for these GPUs outside of AI, and none of them
are remotely hyper scale revenue drivers. As I discussed a
month or so ago, data center development accounted for more
of America's GDP growth than all consumers spending combined in
the first half of.
Speaker 2 (01:29):
Twenty twenty five.
Speaker 3 (01:30):
And there really isn't any demand for AI in general,
let alone at the scale that these hundreds of billions
of dollars are being sunk into. The conservative estimate of
capital expenditures related to data centers is around four hundred
billion dollars, but given the fifty billion dollar in a
quarter in private equity invested, I'm going to guess it
breaks half a trillion, all to build capacity for an
industry yet to prove itself. And this whole in video
(01:51):
OpenAI one hundred billion dollar funding news should only fill
you full of dread. But also it isn't fucking finalized.
Stop reporting as if it's done, it's fucking anyway good.
According to CNBC and I quote the initial ten billion
dollar tranch is locked in at a five hundred billion
dollar valuation and expected to close within a month or
so once the transaction has been finalized, with successive ten
(02:13):
billion dollar rounds planned, each to be priced at the
company's then current valuation. Is new capacity comes online, what
a horrible deal. They can just raise whatever they can,
just go back to video. But let me just be clear.
Open Ai has written a lot of checks that neither
it nor anyone else can cash, and for it to
survive it needs to raise more than, honestly about five
hundred billion dollars and maybe another four hundred billion dollars
(02:35):
and other people paying for stuff. It's astonishing, really, and
no point is anyone asking how exactly open ai builds
the data centers to fill for of the GPUs to
get the rest of Nvidia's funding. In fact, I'm genuinely
shocked and a little disgusted by how poorly this story
has been told.
Speaker 2 (02:50):
Let's go point by point.
Speaker 3 (02:52):
Ten billion dollars is not enough for open ai to
build a data center. The one point one to one
point two Gigawa Gabiline Texas data center being built by
Oracle and Cruiser for open Ai required fifteen billion dollars
in debt, and that's not including the forty billion dollars
of chips needed to power it. Though a question whether
all of those are going into Abilene or somewhere else,
which you get to later. Open Ai can also not
(03:13):
afford to pay for everything it's promised. Open Ai will
burn they say one hundred and fifteen billion dollars in
the next four years ago into the information, but I
believe the company intentionally leaked those numbers before the announcement
of their three hundred billion dollar deal with Oracle, as
when you take into account the numbers share but the
information which involved them burning forty seven billion dollars in
twenty twenty eight a year when it open ai is
meant to make seventy billion dollars in payments to compute
(03:36):
to Oracle, not for twenty eight billion dollars to Microsoft
or any other cost. There's no space for the three
hundred billion dollars to go, and there's actually another one
hundred billion dollars they're promising for backup compute as well.
On top of all of this, open ai is still
yet to convert to a for profit entity, and must
do so by the end of twenty twenty five, or
they'll lose twenty billion dollars in funding from SoftBank. A
(03:57):
few weeks ago, Microsoft and open Ai co published an
announcement that said they'd signed a non binding memorandum of
understanding for the next phase of their partnership, which the
media quickly took to mean the deal was done. No
deal has been done. This is an announcement of nothing.
On top of all of this, venture capital might run
out at the current rate of investment in around six quarters,
(04:17):
and open ai needs more investment than ever. The Information
recently published some worrying data from venture capitalist John Sokoda
that a Today's clip the industry would run out of
money in six quarters, adding that the money wouldn't run
out until the end of twenty twenty eight if it
wasn't for open.
Speaker 2 (04:31):
Ai and Anthropic.
Speaker 3 (04:33):
In a very real sense, open ai threatens the future
of available capital for the tech industry. Well, the good
news is that they're going to make lots of money, right.
I mean, open Ai leaked that they'll make two hundred
billion dollars in four years time. I mean, if you're concussed,
I can see how you believe that. Let's go, I'm
going to read you a bunch of numbers. I know
it's kind of going to be annoying to hear this.
(04:56):
For the next thirty second, I'm going to just throw
figures at you, but pay attention because it's important to
really us how insane this is. So. In twenty twenty five,
open ai projects to make thirteen billion dollars and have
free cash flow of negative nine billion dollars. In twenty
twenty six, open ai will make twenty nine billion dollars
or thirty billion dollars, but have free cash flow of
negative seventeen billion. In twenty twenty seven, open ai will
(05:17):
make sixty billion dollars but have negative cash flow of
negative thirty five billion dollars. And in twenty twenty eight,
open ai will make one hundred billion dollars but have
free cash flow of negative forty seven billion dollars. In
twenty twenty nine, open ai will make one hundred and
forty five billion dollars but a free cash flow of
negative eight billion dollars. Somehow, I just want to reiterate
there open AI's projection. See it reduce its negative cash
(05:39):
flow by thirty nine billion dollars or one and a
half times three m's revenue from the last financial year
in a single year, while also increasing their revenue by
a similar amount, give or take a few billion. Who's counting,
certainly not Sarah Fryer, the CFO of open Ai.
Speaker 2 (05:53):
It's also fucking stupid.
Speaker 3 (05:54):
But don't worry. I know, I know some of you
have been worried. You see big strong men with tears.
Now is going to be saying cirsa open Ai. They're
not gonna They're going to have enough money.
Speaker 2 (06:05):
They're gonna die. But don't worry.
Speaker 3 (06:06):
In twenty thirty, open ai will make two hundred billion
dollars and somehow have positive free cash flow of thirty
eight billion dollars. It's just that easy. I teach you
this in business school. It's just that easy.
Speaker 2 (06:17):
How they're gonna make it. They're gonna make it. They'll
be fine.
Speaker 3 (06:20):
Open ai is current reported burn is one hundred and
fifteen billion dollars through twenty thirty, which means there's no
way that these projections that were leaked included three hundred
billion dollars in compute costs. Even when you factor in
the revenue, there's no space in the projections to absorb
the Oracle money, and from what I can tell, by
twenty twenty nine, open AI will burned up have burned
upwards of two hundred ninety billion dollars, assuming it suffves
(06:40):
that long, which I do not believe it will. Don't
worry though, open ai is about to make some crazy money.
I say in no way being totally sarcastic. I'm gonna
now read you some of the projections that CFO Sarah
Fryer signed off on. This is a professional chief financial officer. Again,
more numbers, but pay attention, Pay attention. This is I
(07:02):
want to give you, because these are numbers you just heard,
but I want to give you some comparison points. So
in twenty twenty six, open AI will make, according to
the projections, or these are the projections, thirty billion dollars
in revenue, or just under thirty the thirty four billion
dollars that Salesforce made in twenty twenty four. In twenty
twenty seven, open AI will allegedly make sixty billion dollars
or two point six billion, more than the fifty seven
(07:23):
point four billion dollars the Oracle made in its fiscal
year twenty twenty five, or eight point four billion dollars
more than Broadcom made in twenty twenty four. In twenty
twenty eight, this is crazy. Open ai will make one
hundred billion dollars two point three billion dollars more than
Tesla made in twenty twenty four, or eleven point eleven
point sixty six billion dollars more than TSMC, the single
largest chip manufacturer in the world.
Speaker 2 (07:44):
Made in twenty twenty four.
Speaker 3 (07:46):
In twenty twenty nine, things get crazier because open ai
will then make one hundred and fifty four billion dollars
or fourteen point five billion dollars more than the one
hundred and thirty point five billion dollars that in Video
made and its fiscal year twenty twenty five. But in
twenty thirty, this is when they really blow the doors off.
They will make two hundred billion dollars or thirty five
point five billion dollars in the one hundred and sixty
four point five billion that Meta made in twenty twenty four.
(08:09):
Just so we are clear, open ai intends to ten
exits revenue in the space of four years, selling software
and access to its models in an industry with about
sixty billion dollars of revenue. In twenty twenty five, How
will it do this? Open ai does not say. I
don't know open ai CFO Sarah Fryer, but I do
know that signing off on these numbers is at the
very least ethically questionable. But putting aside the ridiculousness of
(08:33):
open AI's deals or its funding requirements, Fryar has wilfully
allowed Samulman and open ai to state goals that defy
reality or good sense, all to take advantage of investors
and public markets that have completely lost the plot. I'm
actually you know what, I'm actually going to be blunter.
Open ai has signed multiple different deals and contracts for
amounts it cannot afford to pay that it cannot hope
(08:55):
to raise the money to pay for that defy the
amounts of venture capital and private capital available, or to
sustain a company that will burn half a trillion dollars
in the next four years based on their own expenditures,
and they have no path of the profitability of any kind.
But what about that Invidia deal? And one hundred billion
dollars is a lot of money, right, it is? But
(09:16):
the announcement is bullshit. I know you may have read otherwise,
but in Vidia didn't give open Ai a hundred billion dollars.
In fact, ninety billion dollars of that is contingent on
open Ai building roughly one hundred and twenty five billion
dollars worth of data centers with two hundred billion dollars
worth of GPUs inside them. And those data centers will
have to be built faster than anyone could possibly build
something of that scale. And there's no evidence that open
(09:37):
ai knows where or when these facilities will.
Speaker 2 (09:40):
Be built or who will build them.
Speaker 3 (09:42):
Important detail as well, in Vidia's data centers have nothing
to do with stargate. So when you read that open
ai has seven gigawatts of stargate whatever, it doesn't exist.
Who cares? That's nothing to do with Nvidia. Oracle is
not building these data centers. The ten gigawatts that they
need to build with video are something completely separate. But
(10:10):
you know what, let's take a step back with something
really simple. Data centers take forever to build. As I've
said previously, based on current reports, it's taking Oracle and
Cruiser around two point five.
Speaker 2 (10:21):
Years per giga.
Speaker 3 (10:22):
What of data circumpans city and nowhere in these reports
there's one reporter take a fucking second to say, hey,
what data centers are you talking about?
Speaker 2 (10:29):
Hey, didn't Sam alm.
Speaker 3 (10:30):
And say back in July he was building ten gigawats
of data center capacity with Oracle. But wait now Oracle
and open Ai have done another announcement that says they're
only doing seven gigawats, but they already had a schedule
on ten gigawatts. Like I said, totally different ten gigawats
to the one within video. But you know what, got
a few hundred billion dollars of data centers. Now you're
breaking real money anyway. I cannot be clear enough how
(10:52):
unlikely it is that the first giga what of in
video systems will be deployed in the second half of
twenty twenty six, as in Video has stated with their
deal with open Ai. And that's as if the land
has been bought and got permits and the construction has started.
None of this has happened.
Speaker 2 (11:06):
But you know, let's get really specific on costs.
Speaker 3 (11:10):
Kruso's one point two gigawatts of compute for open Ai
is a fifteen billion dollar joint venture, which means a
gigawat of compute runs about twelve point five billion dollars
worth of construction, Abilene's eight buildings are meant to hold
fifty thousand in video GB two hundred GPUs and their
associated networking infrastructure. So let's say a Gigawa is around
three hundred and thirty three thousand Blackwell GPUs, even though
(11:31):
the GB two hundreds are technically two of them put together.
Though this math is a little funky due to in
video promising to install its new Reuben GPUs in these
theoretical data centers. That means these data centers will require
a little under two hundred billion.
Speaker 2 (11:42):
Dollars worth of GPUs.
Speaker 3 (11:43):
By my maths, that's about three hundred and twenty five
billion dollars, but in videos saying it could be half
a trillion. But you know, who gives a shit, right,
numbers don't matter. You've got Jensen Wong telling CNBC that
one giga what costs fifty sixty billion dollars. You've got
Altman claiming the same thing to Bloomberg. But my mass
is something completely fucking different, But no one seems.
Speaker 2 (12:04):
To want to bother.
Speaker 3 (12:06):
You know. It's only like the places I'm quoting are
leading business outlets with teams of thousands of people, with
incredible infrastructure where they could sit down and say, well,
you know it costs this much to build stargate ABERI,
and why don't we work out actually what this costs? No, no, no,
just eat the info slop yum yum yum. You know
it doesn't exhaust me. It doesn't make me cynical and
(12:28):
frustrated with people who could do a good job. Indeed,
I've seen them do a good job, but they just
seem to not want to do one with this, and
it genuinely bothers me on a day to day basis. Okay,
I realize I'm just ranting at this point, but I'm
gonna be honest. I'm just I'm so tired of this.
As a reminder as well, Open Eyes agreed to spend
(12:50):
three hundred billion dollars on compute with Oracle. But at
this point, as you can tell, numbers basically don't have meaning.
Numbers have stopped meaning anything. If you believe these numbers,
you're a fantasist or a liar, or you just don't
want to think about them too much, because when you
think about them, you start to realize how unrealistic it is.
And even if you remain steadfast in your belief of
(13:11):
the transformative potential of large language models. Eventually, there comes
a point where reality must prevail and you accept that
we're in a bubble, whether you like it or not.
What we're witnessing is one of the most egregious wastes
of capital in history, sold by career charlatans, with their
reputations laundered by a tech and business media afraid to
criticize the powerful an analyst that don't seem to want
to tell their investors the truth. There are no historic
(13:34):
comparisons here. Even Britain's abominable eighteen hundreds railway bubble, which
absorbed half of the country's national income, created valuable infrastructure
for trains, a vehicle that can move people to and
from places. So the train doesn't randomly just go backwards
when you make it go forwards. It doesn't become a
bus or a dog at random. It doesn't hallucinate doors
that open and close at random. Trains work and GPUs
(13:57):
are not trains, nor are they cars or even CPUs,
not adaptable to many other kinds of work, nor are
they the infrastructure of the future of tech, because they're
already quite old and with everybody focused on buying them,
you'd absolutely see one other use case by now that
actually mattered. GPUs are expensive, power hungry, environmentally destructive, and
require their own kinds of cooling and server infrastructure, making
(14:17):
every GPU data center an environmental and fiscal bubble onto themselves.
And whereas the Victorian train and infrastructure still exists in
the UK, though it has been upgraded over the years,
a GPU has a limited useful lifespan. These are cards
that can and will break after a period of extended usage,
whether that period is five years or later, though I
hear it's one to three, and they'll inevitably be superseded
(14:39):
by something better and more powerful, meaning that the resale
value of that GPU will only go down with a
price depreciation. That's the kin to a new car, and
then a used car, and then a car that you
kick the door over every so often. I am telling you,
as I have been telling you for years, again and
again and again, that the demand is not there for
generative AI, and the demand is never ever arriving. The
(15:01):
only reason anyone humors any of this crap is the
endless hoarding of GPUs to build capacity for a revolution.
Speaker 2 (15:07):
That will never arrive.
Speaker 3 (15:08):
Well, that and open Ai, a company that's built and
sold on lies about chat GPT's capabilities. Chat GPT's popularity
in open hunger for endless amounts of compute have created
the illusion of demand due to the sheer amount of
capacity required to keep their services operational. Also that they
can burn around eight billion dollars or more in twenty
twenty five and hundreds of billions of dollars more by
(15:28):
twenty thirty if they make him, which they won't. The
Nvidia dealers are fast an obvious attempt by the largest
company on the American stock market to prop up the
one significant revenue generator in the entire industry, knowing that
time is running out for it to create new avenues
for eternal growth. I'd argue that Nvidia's deal also shows
the complete content that these companies have for the media.
(15:49):
There are no details about how this deal works beyond
the initial ten billion dollars. There's no land purchase, no
data center construction started, yet the media slurps it down
without a second thought. I am but one man, and
I'm fucking peculiar. I did not learn financial analysis in school,
but I appear to be one of the few people
doing even the most basic analysis of these deals. And
(16:09):
while I'm having a great time doing so, I'm are
so extremely frustrated about how little effort is being put
into praying apart these deals by other people.
Speaker 2 (16:17):
I realize how ridiculous all that this sounds. I get it.
Speaker 3 (16:19):
There's so much money being promised, that so many people
market rallies built off the back of these massive deals,
and I get that the assumption that this is that
this much money can't be wrong, that this many people
wouldn't just say stuff without intending to follow through, or
without considering whether their company could afford it. I know
it's hard to conceive that these hundreds of billions of
dollars could be invested in something for no apparent reason,
but it's happening, right goddamn now, in front of your eyes.
(16:42):
And I am going to be merciless on anyone who
attempts to write, how could we possibly have seen this coming?
Generator AI has never been reliable, has always been unprofitable,
and has always been unsustainable. And I've been saying so
since at least February twenty twenty four. These economics have
never made sense. Something I've said repeatedly since April twenty
twenty four, and when I wrote how does open Ai
(17:03):
Survive in July twenty twenty four, I had multiple people
suggest I was being an alarmist. Here's some alarmism for you.
The longer it takes for open ai to die, the
more damage it will cause once it does. As I
talked about earlier in this series, venture capital could run
out in six quarters, with investor and research of John
Socoda estimating that there will be only around one hundred
and sixty four billion dollars of dry powder, which is
(17:25):
available capital, by the way, in USVC firms, by the
end of twenty twenty five. In July, the French Tech
journal reported using Pitchbook DAYA, the global venture capital deal
activity reached its lowest first half total since twenty eighteen,
with one hundred and thirty nine point four billion dollars
in deal value in the first half of twenty twenty five,
down from one hundred and eighty three point four billion
in the first half of twenty twenty four, meaning that
(17:47):
any further expansion or demands for venge capital from open
ai will likely sap dwindling.
Speaker 2 (17:51):
Funds available for other startups.
Speaker 3 (17:54):
Things get worse when you narrow things to the US
venture capital In a piece from April Erstin Young report,
the VC backed in in US companies say eighty billion
dollars in Q one twenty twenty five, but one forty
billion dollar deal account for half of the investment.
Speaker 2 (18:06):
Open Ai is forty billion.
Speaker 3 (18:08):
Dollar deal, of which only ten billion dollar has actually closed,
and that didn't happen until fucking June. Without the imaginary
money from open Ai, US venture capital would have declined
by thirty six percent. I should also add another eight
point three billion dollars was added on the top of that,
but still that's less than forty dollars, folks. The longer
the open Ai survives, the longer it will sap the
(18:28):
remaining billions from the tech ecosystem, and I expect its
tendrils to extend to private credit soon too. The three
hundred and twenty five billion dollars it needs just to
fulfill its Nvidia contract, albeit over four years, is an
egregious sum that I believe exceeds the available private capital
in the world. And I went and actually looked, assuming
that all of this capital is currently available, the top
ten private equity films in the world have around four
(18:50):
hundred and seventy seven billion dollars of available capital. A reminder,
four hundred and seventy seven billion dollars is a lot
less than open AI's total commitments over the next few years.
In fact, I put a premium newsletter out just before
recording this, so I didn't totally update the script. But
open ai needs a trillion dollars. They need about five
hundred billion for operations and about four hundred and fifty
billion for all the data centers they promised and the
(19:12):
GPUs in them.
Speaker 2 (19:14):
It's not really good, it's not good at all.
Speaker 3 (19:16):
But if we include the cash at hand at the
biggest investment banks, this brings us up to By the way,
if we as seen that all of these funds are
available about five hundred and sixty two billion dollars in
capital and one hundred and sixty four billion dollars a
venture capital in the US to spend, and that's all
meant to go in more places than just open AI,
that's not enough. If we combind all the available private
(19:36):
equity firms in the top ten and all of the
available USVC, that's not enough to pay for all the
bills that the open aiy has said, it's we're in
a fucking bubble man. Come on, sure, I also get
that there's more than ten private equity firms and there's
more venture money in that they could raise more money
from lpiece, but what's the left exactly in our one
(19:57):
hundred and fifty two hundred and three hundred billion. They
need that, they need all of that. Hand that over
to Clammy Sami right now. Clammy needs to give that
money to Jensen. Jensen need money. Number go up, you asshole,
give money to Jensen Wong. Now you see open ai
(20:25):
needs to buy those GPUs, and it needs to build
those data centers, and it needs to pay it's thousands
of staff and marketing and sales costs too well. Open
Ai likely wouldn't be the ones raising the money for
the data centers, and honestly, I'm not sure who would.
At this point, somebody is going to need to build
twenty fucking gigawatts of data centers if we're going to
believe Oracle and Nvidia. You may argue that venture funds
(20:45):
and private equity funds can raise more, and you're right,
but at this point, there have been few meaningful acquisitions
of AI companies and zero exits from the billions of
dollars put into data centers. How are they meant to
justify doing this other than me need money?
Speaker 2 (20:58):
Please?
Speaker 3 (20:58):
Even open aims it's in its own announcement about new
stargate sites that this will be a four hundred billion
dollar investment over three years. Where the fuck's the money
coming from? Is open ai really going to absorb massive
junks of all available private credit and venture capital and
capex from major tech companies.
Speaker 2 (21:17):
For the next few years?
Speaker 3 (21:19):
And oh my fucking god, Oh my god, stop saying
the US government will bail this out. I'm tired of
hearing it. I get the things as scary right now.
I'm scared too, We're all fucking scared. But I get
jumping to doomerism.
Speaker 2 (21:33):
But stop. The US government actually really can't afford to
do this. Open AI needs about.
Speaker 3 (21:38):
A trillion dollars in the next four years. And while
the US government has spent equivalent sums in the past
to support private businesses, there's about four hundred and forty
billion dollars dispersed during the Great Recessions tart program, and
that was when the Treasury brought toxic assets from investment
banks to stop them from imploding.
Speaker 2 (21:54):
Like Layman Brothers that one.
Speaker 3 (21:56):
It's hard to imagine any case where open ai is
seen as vital to the global financial system and the
economic health of the US as the actual banking system.
And even if they were, that's more money than they
can easily mobilize, even for the US government. And sure,
we spent around a trillion dollars if we're being specific,
nine hundred and fifty three billion dollars on the paycheck
Protection program during the COVID era. That was to keep
(22:18):
people employed at a time when the economy outside of
Zoom and Walmart had, for all intents and purposes, ceased
to exist. There was an urgency that doesn't apply here.
If open ai goes tits up, soft Bank, Dragonnaer and
a few others lose a bit of money. Nothing new there,
and Sacha Nadella has to explain why he spent tens
of billions, actually hundreds of billions of dollars on data
centers fills with GPUs. But otherwise there's you can't really
(22:42):
bail this out. Are we going to nationalize in video?
Are we going to create a top fund just for
open ai? If you really believe that will happen. I
really must put a smile on your face.
Speaker 2 (22:52):
It can't.
Speaker 3 (22:53):
Even under the most crony capitalism scenario. This is more
money than actually really exists in the world. And yes,
while there will be and have been already disastrous economic consequences,
they won't be as systematically catastrophic as that of a
pandemic or a global financial crisis. To be clear, It'll
be bad, but not as bad. And there's also the
(23:14):
problem of a moral hazard. If the government steps in,
what's the stop big tech chasing its next frootless rainbow,
and of course the optics. If people resented bailing out
the banks after they acted like Gambler's and lost, how
will they feel bailing out fucking Sam Altman and Jensen Huang. Seriously,
no matter when you've got a conservative or Democrat government,
this is not going to be an easy thing to sell.
(23:35):
It's going to be hard to get the consensus, and
indeed every tech company will be kind of pissed. I
think that you miss you underestimate how upset companies that
have died would be, and indeed companies outside of tech
two this would be more than a scandal it would
be genuinely bad to do, even for the silliest government
(23:57):
in the world. I don't think it's happening, and even
if they try, there's not enough money.
Speaker 2 (24:03):
Breathe, they can't get you. You're doing a podcast.
Speaker 3 (24:06):
Look, this has been a slog and I appreciate you
sticking with me this long, but the significance of the
bubble requires this depth. There's little demand, little real money,
and little reason to continue, and the sheer lack of responsibility,
along with the willingness to kneel before the powerful, feels
me full of angry bio. I understand many journalists are
not in a position where they can just write this
shit sounds stupid, but we've entered a deeply stupid era,
(24:27):
and by continuing to perpetuate the myth of AI, the
media guarantees that retail investors and regular people's four O
one case will suffer. It is now inevitable that this
bubble bursts. Deutsche Bank has said that the AI boom
is unsustainable outside of tech spending remaining parabolic, which it
says is highly unlikely, and Bain Capital has said that
two trillion in new revenue is needed to fund AI scaling,
(24:48):
and even that math is completely fucked as it talks
about AI related savings, and I quote even if companies
in the US shifted all of their on premises, it
budgets the cloud and reinvested the savings from applying AI
in sales, marketing, customs support, and r D into capital
spending on new AI data centers, the amount would still
fall short of revenue, the revenue needed to fund the
full investment as AI's compute demands grow more than twice
(25:10):
the rate of Moore's law.
Speaker 2 (25:11):
Ban notes.
Speaker 3 (25:12):
Even when stared in the face by a ridiculous idea
two trillion dollars of new revenue in a global software
market and this is all software that's expected to be
around eight hundred and seventeen billion dollars in twenty twenty five,
Bain still ininks out some nonsense about the savings from
applying AI. Yet another myth perpetuated. I seem to plicate
the fucking moron sinking billions into this Every single vibe
(25:35):
Coding is the future, the power of AI and AI
job lot story written perpetuates a myth that will only
lead to more regular people getting hurt.
Speaker 2 (25:43):
When the bubble bursts.
Speaker 3 (25:44):
Every article written about open AI or in video or
oracle that does not explicitly state that the money doesn't exist,
that the revenues that are impossible, that one of the
companies involved burns billions of dollars and has no path
to profitability is an act of irresponsibility, and irresponsible make
belief and mythos. Look, I'm nobody, I'm not a financier,
(26:05):
I'm not anybody special. I write a lot, I read
a lot, and can do the most basic maths in
the world. I'm not trying to be anything other than myself,
nor do I have an agenda other than the fact
that I like doing this and I hate how.
Speaker 2 (26:16):
This story is being told.
Speaker 3 (26:18):
I never expected my newsletter to get big, nor did
I expect this podcast, nor did I expect to be
in the Financial Times. But here I am, and I'm
working my ass off, and I am pissed off at
where this has got to. I'm tired that things remain
honestly mythical, that we still see as things get more
and more ridiculous, that the headlines get more ridiculous too,
(26:42):
that there's questioning and people are saying, oh, the narrative
is shifting, sure, but they're still repeating increasingly dumb shit.
And I also believe that the way to stop this
happening again is to have a thorough and well sourced
explanation of everything that happens, ripping down the narratives as
they're spun, and making it clear who benefits from them
and how on why they're choosing to do so. When
(27:03):
things collapse, we need to be clear about how many
times people chose to look the other way or to
find good faith ways to interpret bad faith announcements and leaks.
Speaker 2 (27:12):
So how could we have seen this coming? I don't know,
did anybody.
Speaker 3 (27:15):
Try to fucking look?
Speaker 2 (27:25):
Thank you for listening to Better Offline.
Speaker 4 (27:27):
The editor and composer of the Better Offline theme song
is Matasowski. You can check out more of his music
and audio projects at Mattasowski dot com, m A T
T O.
Speaker 2 (27:36):
S O W s ki dot com.
Speaker 4 (27:40):
You can email me at easy at Better offline dot com,
or visit Better Offline dot com to find more podcast
links and of course my newsletter. I also really recommend
you go to chat dot Where's youreaed dot at to
visit the discord and go to our slash Better off
Line to check out our reddit thank you so much
for listening.
Speaker 1 (27:57):
Better Offline is a production of Calls on Me. For
more from cool Zone Media, visit our website Coolzonemedia dot com,
or check us out on the iHeartRadio app, Apple Podcasts,
or wherever you get your podcasts.
Zoe Media.
Speaker 2 (00:04):
All Right, we're back.
Speaker 3 (00:05):
I'm ed Zetron and this is better offline. This is
the fourth and final episode of this four part series
where I dissect in excruciating comprehensive detail the AI bubble
(00:26):
and tell you why I think its implosion is inevitable.
We're at the end, which is fitting because in this
episode we're going to talk about why and how this
entire sham dies again. If you jumping in now, start
from the beginning. We talk about a lot of things
and it all kind of threads together. I know, I
know you want to hear my explanation of how open
ai is fucked with a capital fuck. But we've got
to lay the groundwork to get there. Okay, got up, good, Good,
(00:48):
We're good. We're god, You're ready, We're good, You're got yep, yep.
Speaker 2 (00:51):
Okay, we'll begin.
Speaker 3 (00:52):
One of the comfortable lies that people tell themselves is
that the AI bubble is similar to the fiber boom
or the dot com boom or uber, or that we're
in the growth stay, that this is what software companies do.
They spend a bunch of money, then pull the profit lever.
The thing is, this is nothing like anything you've ever
seen before, because this is the dumbest shit the tech
industry has ever ever done. AI data centers are nothing
(01:14):
like fiber because there are very few actual use cases
for these GPUs outside of AI, and none of them
are remotely hyper scale revenue drivers. As I discussed a
month or so ago, data center development accounted for more
of America's GDP growth than all consumers spending combined in
the first half of.
Speaker 2 (01:29):
Twenty twenty five.
Speaker 3 (01:30):
And there really isn't any demand for AI in general,
let alone at the scale that these hundreds of billions
of dollars are being sunk into. The conservative estimate of
capital expenditures related to data centers is around four hundred
billion dollars, but given the fifty billion dollar in a
quarter in private equity invested, I'm going to guess it
breaks half a trillion, all to build capacity for an
industry yet to prove itself. And this whole in video
(01:51):
OpenAI one hundred billion dollar funding news should only fill
you full of dread. But also it isn't fucking finalized.
Stop reporting as if it's done, it's fucking anyway good.
According to CNBC and I quote the initial ten billion
dollar tranch is locked in at a five hundred billion
dollar valuation and expected to close within a month or
so once the transaction has been finalized, with successive ten
(02:13):
billion dollar rounds planned, each to be priced at the
company's then current valuation. Is new capacity comes online, what
a horrible deal. They can just raise whatever they can,
just go back to video. But let me just be clear.
Open Ai has written a lot of checks that neither
it nor anyone else can cash, and for it to
survive it needs to raise more than, honestly about five
hundred billion dollars and maybe another four hundred billion dollars
(02:35):
and other people paying for stuff. It's astonishing, really, and
no point is anyone asking how exactly open ai builds
the data centers to fill for of the GPUs to
get the rest of Nvidia's funding. In fact, I'm genuinely
shocked and a little disgusted by how poorly this story
has been told.
Speaker 2 (02:50):
Let's go point by point.
Speaker 3 (02:52):
Ten billion dollars is not enough for open ai to
build a data center. The one point one to one
point two Gigawa Gabiline Texas data center being built by
Oracle and Cruiser for open Ai required fifteen billion dollars
in debt, and that's not including the forty billion dollars
of chips needed to power it. Though a question whether
all of those are going into Abilene or somewhere else,
which you get to later. Open Ai can also not
(03:13):
afford to pay for everything it's promised. Open Ai will
burn they say one hundred and fifteen billion dollars in
the next four years ago into the information, but I
believe the company intentionally leaked those numbers before the announcement
of their three hundred billion dollar deal with Oracle, as
when you take into account the numbers share but the
information which involved them burning forty seven billion dollars in
twenty twenty eight a year when it open ai is
meant to make seventy billion dollars in payments to compute
(03:36):
to Oracle, not for twenty eight billion dollars to Microsoft
or any other cost. There's no space for the three
hundred billion dollars to go, and there's actually another one
hundred billion dollars they're promising for backup compute as well.
On top of all of this, open ai is still
yet to convert to a for profit entity, and must
do so by the end of twenty twenty five, or
they'll lose twenty billion dollars in funding from SoftBank. A
(03:57):
few weeks ago, Microsoft and open Ai co published an
announcement that said they'd signed a non binding memorandum of
understanding for the next phase of their partnership, which the
media quickly took to mean the deal was done. No
deal has been done. This is an announcement of nothing.
On top of all of this, venture capital might run
out at the current rate of investment in around six quarters,
(04:17):
and open ai needs more investment than ever. The Information
recently published some worrying data from venture capitalist John Sokoda
that a Today's clip the industry would run out of
money in six quarters, adding that the money wouldn't run
out until the end of twenty twenty eight if it
wasn't for open.
Speaker 2 (04:31):
Ai and Anthropic.
Speaker 3 (04:33):
In a very real sense, open ai threatens the future
of available capital for the tech industry. Well, the good
news is that they're going to make lots of money, right.
I mean, open Ai leaked that they'll make two hundred
billion dollars in four years time. I mean, if you're concussed,
I can see how you believe that. Let's go, I'm
going to read you a bunch of numbers. I know
it's kind of going to be annoying to hear this.
(04:56):
For the next thirty second, I'm going to just throw
figures at you, but pay attention because it's important to
really us how insane this is. So. In twenty twenty five,
open ai projects to make thirteen billion dollars and have
free cash flow of negative nine billion dollars. In twenty
twenty six, open ai will make twenty nine billion dollars
or thirty billion dollars, but have free cash flow of
negative seventeen billion. In twenty twenty seven, open ai will
(05:17):
make sixty billion dollars but have negative cash flow of
negative thirty five billion dollars. And in twenty twenty eight,
open ai will make one hundred billion dollars but have
free cash flow of negative forty seven billion dollars. In
twenty twenty nine, open ai will make one hundred and
forty five billion dollars but a free cash flow of
negative eight billion dollars. Somehow, I just want to reiterate
there open AI's projection. See it reduce its negative cash
(05:39):
flow by thirty nine billion dollars or one and a
half times three m's revenue from the last financial year
in a single year, while also increasing their revenue by
a similar amount, give or take a few billion. Who's counting,
certainly not Sarah Fryer, the CFO of open Ai.
Speaker 2 (05:53):
It's also fucking stupid.
Speaker 3 (05:54):
But don't worry. I know, I know some of you
have been worried. You see big strong men with tears.
Now is going to be saying cirsa open Ai. They're
not gonna They're going to have enough money.
Speaker 2 (06:05):
They're gonna die. But don't worry.
Speaker 3 (06:06):
In twenty thirty, open ai will make two hundred billion
dollars and somehow have positive free cash flow of thirty
eight billion dollars. It's just that easy. I teach you
this in business school. It's just that easy.
Speaker 2 (06:17):
How they're gonna make it. They're gonna make it. They'll
be fine.
Speaker 3 (06:20):
Open ai is current reported burn is one hundred and
fifteen billion dollars through twenty thirty, which means there's no
way that these projections that were leaked included three hundred
billion dollars in compute costs. Even when you factor in
the revenue, there's no space in the projections to absorb
the Oracle money, and from what I can tell, by
twenty twenty nine, open AI will burned up have burned
upwards of two hundred ninety billion dollars, assuming it suffves
(06:40):
that long, which I do not believe it will. Don't
worry though, open ai is about to make some crazy money.
I say in no way being totally sarcastic. I'm gonna
now read you some of the projections that CFO Sarah
Fryer signed off on. This is a professional chief financial officer. Again,
more numbers, but pay attention, Pay attention. This is I
(07:02):
want to give you, because these are numbers you just heard,
but I want to give you some comparison points. So
in twenty twenty six, open AI will make, according to
the projections, or these are the projections, thirty billion dollars
in revenue, or just under thirty the thirty four billion
dollars that Salesforce made in twenty twenty four. In twenty
twenty seven, open AI will allegedly make sixty billion dollars
or two point six billion, more than the fifty seven
(07:23):
point four billion dollars the Oracle made in its fiscal
year twenty twenty five, or eight point four billion dollars
more than Broadcom made in twenty twenty four. In twenty
twenty eight, this is crazy. Open ai will make one
hundred billion dollars two point three billion dollars more than
Tesla made in twenty twenty four, or eleven point eleven
point sixty six billion dollars more than TSMC, the single
largest chip manufacturer in the world.
Speaker 2 (07:44):
Made in twenty twenty four.
Speaker 3 (07:46):
In twenty twenty nine, things get crazier because open ai
will then make one hundred and fifty four billion dollars
or fourteen point five billion dollars more than the one
hundred and thirty point five billion dollars that in Video
made and its fiscal year twenty twenty five. But in
twenty thirty, this is when they really blow the doors off.
They will make two hundred billion dollars or thirty five
point five billion dollars in the one hundred and sixty
four point five billion that Meta made in twenty twenty four.
(08:09):
Just so we are clear, open ai intends to ten
exits revenue in the space of four years, selling software
and access to its models in an industry with about
sixty billion dollars of revenue. In twenty twenty five, How
will it do this? Open ai does not say. I
don't know open ai CFO Sarah Fryer, but I do
know that signing off on these numbers is at the
very least ethically questionable. But putting aside the ridiculousness of
(08:33):
open AI's deals or its funding requirements, Fryar has wilfully
allowed Samulman and open ai to state goals that defy
reality or good sense, all to take advantage of investors
and public markets that have completely lost the plot. I'm
actually you know what, I'm actually going to be blunter.
Open ai has signed multiple different deals and contracts for
amounts it cannot afford to pay that it cannot hope
(08:55):
to raise the money to pay for that defy the
amounts of venture capital and private capital available, or to
sustain a company that will burn half a trillion dollars
in the next four years based on their own expenditures,
and they have no path of the profitability of any kind.
But what about that Invidia deal? And one hundred billion
dollars is a lot of money, right, it is? But
(09:16):
the announcement is bullshit. I know you may have read otherwise,
but in Vidia didn't give open Ai a hundred billion dollars.
In fact, ninety billion dollars of that is contingent on
open Ai building roughly one hundred and twenty five billion
dollars worth of data centers with two hundred billion dollars
worth of GPUs inside them. And those data centers will
have to be built faster than anyone could possibly build
something of that scale. And there's no evidence that open
(09:37):
ai knows where or when these facilities will.
Speaker 2 (09:40):
Be built or who will build them.
Speaker 3 (09:42):
Important detail as well, in Vidia's data centers have nothing
to do with stargate. So when you read that open
ai has seven gigawatts of stargate whatever, it doesn't exist.
Who cares? That's nothing to do with Nvidia. Oracle is
not building these data centers. The ten gigawatts that they
need to build with video are something completely separate. But
(10:10):
you know what, let's take a step back with something
really simple. Data centers take forever to build. As I've
said previously, based on current reports, it's taking Oracle and
Cruiser around two point five.
Speaker 2 (10:21):
Years per giga.
Speaker 3 (10:22):
What of data circumpans city and nowhere in these reports
there's one reporter take a fucking second to say, hey,
what data centers are you talking about?
Speaker 2 (10:29):
Hey, didn't Sam alm.
Speaker 3 (10:30):
And say back in July he was building ten gigawats
of data center capacity with Oracle. But wait now Oracle
and open Ai have done another announcement that says they're
only doing seven gigawats, but they already had a schedule
on ten gigawatts. Like I said, totally different ten gigawats
to the one within video. But you know what, got
a few hundred billion dollars of data centers. Now you're
breaking real money anyway. I cannot be clear enough how
(10:52):
unlikely it is that the first giga what of in
video systems will be deployed in the second half of
twenty twenty six, as in Video has stated with their
deal with open Ai. And that's as if the land
has been bought and got permits and the construction has started.
None of this has happened.
Speaker 2 (11:06):
But you know, let's get really specific on costs.
Speaker 3 (11:10):
Kruso's one point two gigawatts of compute for open Ai
is a fifteen billion dollar joint venture, which means a
gigawat of compute runs about twelve point five billion dollars
worth of construction, Abilene's eight buildings are meant to hold
fifty thousand in video GB two hundred GPUs and their
associated networking infrastructure. So let's say a Gigawa is around
three hundred and thirty three thousand Blackwell GPUs, even though
(11:31):
the GB two hundreds are technically two of them put together.
Though this math is a little funky due to in
video promising to install its new Reuben GPUs in these
theoretical data centers. That means these data centers will require
a little under two hundred billion.
Speaker 2 (11:42):
Dollars worth of GPUs.
Speaker 3 (11:43):
By my maths, that's about three hundred and twenty five
billion dollars, but in videos saying it could be half
a trillion. But you know, who gives a shit, right,
numbers don't matter. You've got Jensen Wong telling CNBC that
one giga what costs fifty sixty billion dollars. You've got
Altman claiming the same thing to Bloomberg. But my mass
is something completely fucking different, But no one seems.
Speaker 2 (12:04):
To want to bother.
Speaker 3 (12:06):
You know. It's only like the places I'm quoting are
leading business outlets with teams of thousands of people, with
incredible infrastructure where they could sit down and say, well,
you know it costs this much to build stargate ABERI,
and why don't we work out actually what this costs? No, no, no,
just eat the info slop yum yum yum. You know
it doesn't exhaust me. It doesn't make me cynical and
(12:28):
frustrated with people who could do a good job. Indeed,
I've seen them do a good job, but they just
seem to not want to do one with this, and
it genuinely bothers me on a day to day basis. Okay,
I realize I'm just ranting at this point, but I'm
gonna be honest. I'm just I'm so tired of this.
As a reminder as well, Open Eyes agreed to spend
(12:50):
three hundred billion dollars on compute with Oracle. But at
this point, as you can tell, numbers basically don't have meaning.
Numbers have stopped meaning anything. If you believe these numbers,
you're a fantasist or a liar, or you just don't
want to think about them too much, because when you
think about them, you start to realize how unrealistic it is.
And even if you remain steadfast in your belief of
(13:11):
the transformative potential of large language models. Eventually, there comes
a point where reality must prevail and you accept that
we're in a bubble, whether you like it or not.
What we're witnessing is one of the most egregious wastes
of capital in history, sold by career charlatans, with their
reputations laundered by a tech and business media afraid to
criticize the powerful an analyst that don't seem to want
to tell their investors the truth. There are no historic
(13:34):
comparisons here. Even Britain's abominable eighteen hundreds railway bubble, which
absorbed half of the country's national income, created valuable infrastructure
for trains, a vehicle that can move people to and
from places. So the train doesn't randomly just go backwards
when you make it go forwards. It doesn't become a
bus or a dog at random. It doesn't hallucinate doors
that open and close at random. Trains work and GPUs
(13:57):
are not trains, nor are they cars or even CPUs,
not adaptable to many other kinds of work, nor are
they the infrastructure of the future of tech, because they're
already quite old and with everybody focused on buying them,
you'd absolutely see one other use case by now that
actually mattered. GPUs are expensive, power hungry, environmentally destructive, and
require their own kinds of cooling and server infrastructure, making
(14:17):
every GPU data center an environmental and fiscal bubble onto themselves.
And whereas the Victorian train and infrastructure still exists in
the UK, though it has been upgraded over the years,
a GPU has a limited useful lifespan. These are cards
that can and will break after a period of extended usage,
whether that period is five years or later, though I
hear it's one to three, and they'll inevitably be superseded
(14:39):
by something better and more powerful, meaning that the resale
value of that GPU will only go down with a
price depreciation. That's the kin to a new car, and
then a used car, and then a car that you
kick the door over every so often. I am telling you,
as I have been telling you for years, again and
again and again, that the demand is not there for
generative AI, and the demand is never ever arriving. The
(15:01):
only reason anyone humors any of this crap is the
endless hoarding of GPUs to build capacity for a revolution.
Speaker 2 (15:07):
That will never arrive.
Speaker 3 (15:08):
Well, that and open Ai, a company that's built and
sold on lies about chat GPT's capabilities. Chat GPT's popularity
in open hunger for endless amounts of compute have created
the illusion of demand due to the sheer amount of
capacity required to keep their services operational. Also that they
can burn around eight billion dollars or more in twenty
twenty five and hundreds of billions of dollars more by
(15:28):
twenty thirty if they make him, which they won't. The
Nvidia dealers are fast an obvious attempt by the largest
company on the American stock market to prop up the
one significant revenue generator in the entire industry, knowing that
time is running out for it to create new avenues
for eternal growth. I'd argue that Nvidia's deal also shows
the complete content that these companies have for the media.
(15:49):
There are no details about how this deal works beyond
the initial ten billion dollars. There's no land purchase, no
data center construction started, yet the media slurps it down
without a second thought. I am but one man, and
I'm fucking peculiar. I did not learn financial analysis in school,
but I appear to be one of the few people
doing even the most basic analysis of these deals. And
(16:09):
while I'm having a great time doing so, I'm are
so extremely frustrated about how little effort is being put
into praying apart these deals by other people.
Speaker 2 (16:17):
I realize how ridiculous all that this sounds. I get it.
Speaker 3 (16:19):
There's so much money being promised, that so many people
market rallies built off the back of these massive deals,
and I get that the assumption that this is that
this much money can't be wrong, that this many people
wouldn't just say stuff without intending to follow through, or
without considering whether their company could afford it. I know
it's hard to conceive that these hundreds of billions of
dollars could be invested in something for no apparent reason,
but it's happening, right goddamn now, in front of your eyes.
(16:42):
And I am going to be merciless on anyone who
attempts to write, how could we possibly have seen this coming?
Generator AI has never been reliable, has always been unprofitable,
and has always been unsustainable. And I've been saying so
since at least February twenty twenty four. These economics have
never made sense. Something I've said repeatedly since April twenty
twenty four, and when I wrote how does open Ai
(17:03):
Survive in July twenty twenty four, I had multiple people
suggest I was being an alarmist. Here's some alarmism for you.
The longer it takes for open ai to die, the
more damage it will cause once it does. As I
talked about earlier in this series, venture capital could run
out in six quarters, with investor and research of John
Socoda estimating that there will be only around one hundred
and sixty four billion dollars of dry powder, which is
(17:25):
available capital, by the way, in USVC firms, by the
end of twenty twenty five. In July, the French Tech
journal reported using Pitchbook DAYA, the global venture capital deal
activity reached its lowest first half total since twenty eighteen,
with one hundred and thirty nine point four billion dollars
in deal value in the first half of twenty twenty five,
down from one hundred and eighty three point four billion
in the first half of twenty twenty four, meaning that
(17:47):
any further expansion or demands for venge capital from open
ai will likely sap dwindling.
Speaker 2 (17:51):
Funds available for other startups.
Speaker 3 (17:54):
Things get worse when you narrow things to the US
venture capital In a piece from April Erstin Young report,
the VC backed in in US companies say eighty billion
dollars in Q one twenty twenty five, but one forty
billion dollar deal account for half of the investment.
Speaker 2 (18:06):
Open Ai is forty billion.
Speaker 3 (18:08):
Dollar deal, of which only ten billion dollar has actually closed,
and that didn't happen until fucking June. Without the imaginary
money from open Ai, US venture capital would have declined
by thirty six percent. I should also add another eight
point three billion dollars was added on the top of that,
but still that's less than forty dollars, folks. The longer
the open Ai survives, the longer it will sap the
(18:28):
remaining billions from the tech ecosystem, and I expect its
tendrils to extend to private credit soon too. The three
hundred and twenty five billion dollars it needs just to
fulfill its Nvidia contract, albeit over four years, is an
egregious sum that I believe exceeds the available private capital
in the world. And I went and actually looked, assuming
that all of this capital is currently available, the top
ten private equity films in the world have around four
(18:50):
hundred and seventy seven billion dollars of available capital. A reminder,
four hundred and seventy seven billion dollars is a lot
less than open AI's total commitments over the next few years.
In fact, I put a premium newsletter out just before
recording this, so I didn't totally update the script. But
open ai needs a trillion dollars. They need about five
hundred billion for operations and about four hundred and fifty
billion for all the data centers they promised and the
(19:12):
GPUs in them.
Speaker 2 (19:14):
It's not really good, it's not good at all.
Speaker 3 (19:16):
But if we include the cash at hand at the
biggest investment banks, this brings us up to By the way,
if we as seen that all of these funds are
available about five hundred and sixty two billion dollars in
capital and one hundred and sixty four billion dollars a
venture capital in the US to spend, and that's all
meant to go in more places than just open AI,
that's not enough. If we combind all the available private
(19:36):
equity firms in the top ten and all of the
available USVC, that's not enough to pay for all the
bills that the open aiy has said, it's we're in
a fucking bubble man. Come on, sure, I also get
that there's more than ten private equity firms and there's
more venture money in that they could raise more money
from lpiece, but what's the left exactly in our one
(19:57):
hundred and fifty two hundred and three hundred billion. They
need that, they need all of that. Hand that over
to Clammy Sami right now. Clammy needs to give that
money to Jensen. Jensen need money. Number go up, you asshole,
give money to Jensen Wong. Now you see open ai
(20:25):
needs to buy those GPUs, and it needs to build
those data centers, and it needs to pay it's thousands
of staff and marketing and sales costs too well. Open
Ai likely wouldn't be the ones raising the money for
the data centers, and honestly, I'm not sure who would.
At this point, somebody is going to need to build
twenty fucking gigawatts of data centers if we're going to
believe Oracle and Nvidia. You may argue that venture funds
(20:45):
and private equity funds can raise more, and you're right,
but at this point, there have been few meaningful acquisitions
of AI companies and zero exits from the billions of
dollars put into data centers. How are they meant to
justify doing this other than me need money?
Speaker 2 (20:58):
Please?
Speaker 3 (20:58):
Even open aims it's in its own announcement about new
stargate sites that this will be a four hundred billion
dollar investment over three years. Where the fuck's the money
coming from? Is open ai really going to absorb massive
junks of all available private credit and venture capital and
capex from major tech companies.
Speaker 2 (21:17):
For the next few years?
Speaker 3 (21:19):
And oh my fucking god, Oh my god, stop saying
the US government will bail this out. I'm tired of
hearing it. I get the things as scary right now.
I'm scared too, We're all fucking scared. But I get
jumping to doomerism.
Speaker 2 (21:33):
But stop. The US government actually really can't afford to
do this. Open AI needs about.
Speaker 3 (21:38):
A trillion dollars in the next four years. And while
the US government has spent equivalent sums in the past
to support private businesses, there's about four hundred and forty
billion dollars dispersed during the Great Recessions tart program, and
that was when the Treasury brought toxic assets from investment
banks to stop them from imploding.
Speaker 2 (21:54):
Like Layman Brothers that one.
Speaker 3 (21:56):
It's hard to imagine any case where open ai is
seen as vital to the global financial system and the
economic health of the US as the actual banking system.
And even if they were, that's more money than they
can easily mobilize, even for the US government. And sure,
we spent around a trillion dollars if we're being specific,
nine hundred and fifty three billion dollars on the paycheck
Protection program during the COVID era. That was to keep
(22:18):
people employed at a time when the economy outside of
Zoom and Walmart had, for all intents and purposes, ceased
to exist. There was an urgency that doesn't apply here.
If open ai goes tits up, soft Bank, Dragonnaer and
a few others lose a bit of money. Nothing new there,
and Sacha Nadella has to explain why he spent tens
of billions, actually hundreds of billions of dollars on data
centers fills with GPUs. But otherwise there's you can't really
(22:42):
bail this out. Are we going to nationalize in video?
Are we going to create a top fund just for
open ai? If you really believe that will happen. I
really must put a smile on your face.
Speaker 2 (22:52):
It can't.
Speaker 3 (22:53):
Even under the most crony capitalism scenario. This is more
money than actually really exists in the world. And yes,
while there will be and have been already disastrous economic consequences,
they won't be as systematically catastrophic as that of a
pandemic or a global financial crisis. To be clear, It'll
be bad, but not as bad. And there's also the
(23:14):
problem of a moral hazard. If the government steps in,
what's the stop big tech chasing its next frootless rainbow,
and of course the optics. If people resented bailing out
the banks after they acted like Gambler's and lost, how
will they feel bailing out fucking Sam Altman and Jensen Huang. Seriously,
no matter when you've got a conservative or Democrat government,
this is not going to be an easy thing to sell.
(23:35):
It's going to be hard to get the consensus, and
indeed every tech company will be kind of pissed. I
think that you miss you underestimate how upset companies that
have died would be, and indeed companies outside of tech
two this would be more than a scandal it would
be genuinely bad to do, even for the silliest government
(23:57):
in the world. I don't think it's happening, and even
if they try, there's not enough money.
Speaker 2 (24:03):
Breathe, they can't get you. You're doing a podcast.
Speaker 3 (24:06):
Look, this has been a slog and I appreciate you
sticking with me this long, but the significance of the
bubble requires this depth. There's little demand, little real money,
and little reason to continue, and the sheer lack of responsibility,
along with the willingness to kneel before the powerful, feels
me full of angry bio. I understand many journalists are
not in a position where they can just write this
shit sounds stupid, but we've entered a deeply stupid era,
(24:27):
and by continuing to perpetuate the myth of AI, the
media guarantees that retail investors and regular people's four O
one case will suffer. It is now inevitable that this
bubble bursts. Deutsche Bank has said that the AI boom
is unsustainable outside of tech spending remaining parabolic, which it
says is highly unlikely, and Bain Capital has said that
two trillion in new revenue is needed to fund AI scaling,
(24:48):
and even that math is completely fucked as it talks
about AI related savings, and I quote even if companies
in the US shifted all of their on premises, it
budgets the cloud and reinvested the savings from applying AI
in sales, marketing, customs support, and r D into capital
spending on new AI data centers, the amount would still
fall short of revenue, the revenue needed to fund the
full investment as AI's compute demands grow more than twice
(25:10):
the rate of Moore's law.
Speaker 2 (25:11):
Ban notes.
Speaker 3 (25:12):
Even when stared in the face by a ridiculous idea
two trillion dollars of new revenue in a global software
market and this is all software that's expected to be
around eight hundred and seventeen billion dollars in twenty twenty five,
Bain still ininks out some nonsense about the savings from
applying AI. Yet another myth perpetuated. I seem to plicate
the fucking moron sinking billions into this Every single vibe
(25:35):
Coding is the future, the power of AI and AI
job lot story written perpetuates a myth that will only
lead to more regular people getting hurt.
Speaker 2 (25:43):
When the bubble bursts.
Speaker 3 (25:44):
Every article written about open AI or in video or
oracle that does not explicitly state that the money doesn't exist,
that the revenues that are impossible, that one of the
companies involved burns billions of dollars and has no path
to profitability is an act of irresponsibility, and irresponsible make
belief and mythos. Look, I'm nobody, I'm not a financier,
(26:05):
I'm not anybody special. I write a lot, I read
a lot, and can do the most basic maths in
the world. I'm not trying to be anything other than myself,
nor do I have an agenda other than the fact
that I like doing this and I hate how.
Speaker 2 (26:16):
This story is being told.
Speaker 3 (26:18):
I never expected my newsletter to get big, nor did
I expect this podcast, nor did I expect to be
in the Financial Times. But here I am, and I'm
working my ass off, and I am pissed off at
where this has got to. I'm tired that things remain
honestly mythical, that we still see as things get more
and more ridiculous, that the headlines get more ridiculous too,
(26:42):
that there's questioning and people are saying, oh, the narrative
is shifting, sure, but they're still repeating increasingly dumb shit.
And I also believe that the way to stop this
happening again is to have a thorough and well sourced
explanation of everything that happens, ripping down the narratives as
they're spun, and making it clear who benefits from them
and how on why they're choosing to do so. When
(27:03):
things collapse, we need to be clear about how many
times people chose to look the other way or to
find good faith ways to interpret bad faith announcements and leaks.
Speaker 2 (27:12):
So how could we have seen this coming? I don't know,
did anybody.
Speaker 3 (27:15):
Try to fucking look?
Speaker 2 (27:25):
Thank you for listening to Better Offline.
Speaker 4 (27:27):
The editor and composer of the Better Offline theme song
is Matasowski. You can check out more of his music
and audio projects at Mattasowski dot com, m A T
T O.
Speaker 2 (27:36):
S O W s ki dot com.
Speaker 4 (27:40):
You can email me at easy at Better offline dot com,
or visit Better Offline dot com to find more podcast
links and of course my newsletter. I also really recommend
you go to chat dot Where's youreaed dot at to
visit the discord and go to our slash Better off
Line to check out our reddit thank you so much
for listening.
Speaker 1 (27:57):
Better Offline is a production of Calls on Me. For
more from cool Zone Media, visit our website Coolzonemedia dot com,
or check us out on the iHeartRadio app, Apple Podcasts,
or wherever you get your podcasts.
Host
Ed Zitron
Popular Podcasts
Stuff You Should Know
If you've ever wanted to know about champagne, satanism, the Stonewall Uprising, chaos theory, LSD, El Nino, true crime and Rosa Parks, then look no further. Josh and Chuck have you covered.
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 Joe Rogan Experience
The official podcast of comedian Joe Rogan.