00:00:00
All right, everybody. Welcome back to
00:00:01
the number one podcast in the world.
00:00:04
We've got the core four here and dare I
00:00:08
say,
00:00:09
>> the king of Adams.
00:00:10
>> The king of Adams. Yes. Captain Travis
00:00:14
Kalanick is here. How you doing,
00:00:16
brother?
00:00:16
>> I'm pretty good. Pretty good. I'm uh
00:00:19
sitting here doing the podcast just next
00:00:21
door to David.
00:00:22
>> Ah, yes. There you go.
00:00:24
>> Don't reveal our locations.
00:00:25
>> Please don't dock.
00:00:26
>> I didn't put my address out there, dude.
00:00:28
>> No, that's true. Don't worry, Mandami
00:00:30
did. He's outside your houses right now
00:00:33
asking them to
00:00:33
>> foresee trying to collect 3.9% or so.
00:00:36
>> Yeah, he's he they he decided there's
00:00:38
rich people left in New York. So, he's
00:00:40
looking for other
00:00:42
is it 3.9% a year. Is it per year or is
00:00:45
it
00:00:46
>> I don't know if the percentage has been
00:00:47
released yet, but the speculation I've
00:00:48
seen is 3.9% but I don't think that's
00:00:50
final.
00:00:51
>> But yeah, it's a pia ter tax. So, if you
00:00:54
have a second home Yeah. every year.
00:00:56
>> Wow. And by the way, it's for any home
00:00:59
over 5 million. There's no homes under 5
00:01:02
million in Manhattan. This is not a rich
00:01:04
person tax. This is within
00:01:07
15 miles of Midtown Manhattan, you're
00:01:10
paying an extra tax. Uh I don't
00:01:12
>> but only but Jal only if it's a pat. So
00:01:14
what it means is that the most
00:01:16
>> second homes.
00:01:16
>> Well, yeah, but the most elastic part of
00:01:18
the market is what they're targeting for
00:01:20
this tax. So in other words, people who
00:01:22
don't live in New York, who just have it
00:01:24
as a second or third home, who could buy
00:01:27
that property anywhere,
00:01:29
>> yeah,
00:01:29
>> are now being taxed the most. So what do
00:01:31
you think that's going to do? It's going
00:01:32
to have a massive impact on demand for
00:01:36
second homes in New York, which will
00:01:38
crash the whole market.
00:01:39
>> Yes, congratulations, Mummy.
00:01:41
>> But in a weird way, that'll be good for
00:01:42
housing affordability in New York.
00:01:44
>> Well, that that's sort of the the claim,
00:01:46
but I don't think it'll be good for it
00:01:48
because there'll be no incentive to
00:01:49
build more.
00:01:51
Yeah, all units matter. Every time you
00:01:53
add units, people upgrade. And it's not
00:01:55
like these are going to be lowincome
00:01:57
housing. Like penthouse on 57th Street
00:02:00
or, you know, in Grammarcy. That's not
00:02:03
lowincome housing. You'd have to break
00:02:05
it into seven units. Makes no sense. But
00:02:07
by the way, I don't know if you guys saw
00:02:09
the video, not to get too serious, but
00:02:11
he's doxing a certain billionaire who
00:02:13
owns a certain place and he's literally
00:02:15
pointing at his home.
00:02:16
>> No, I said that to you, Jason. He's not
00:02:17
doxing because everybody's known for
00:02:19
years that Ken Griffin bought that
00:02:21
place. Everybody knows that address.
00:02:23
Everybody knows that unit. We all knew
00:02:24
it. It was marketed widely. I don't
00:02:27
think that's really doxing. He doesn't
00:02:28
live there and everybody knew he owned
00:02:30
it. It would be very different if it was
00:02:32
a place where somebody was keeping their
00:02:34
primary residence and you didn't know
00:02:36
and they stood in front of the house.
00:02:37
That I would agree with you. I think
00:02:39
this one is a little bit more tenuous.
00:02:40
>> Okay, fair enough. But what I will tell
00:02:42
you is it's a dog whistle. crazy people
00:02:45
and this thing's been seen by 30, 40, 50
00:02:47
million people. Now,
00:02:48
>> it's a dog whistle to say that's the
00:02:51
next United Healthcare CEO and in the
00:02:53
week that a fire a Molotov cocktail and
00:02:56
a bullet gets shot into Sam Alman's
00:02:58
house, it's deadly serious. And if you
00:03:00
reverse it, I always reverse it. What if
00:03:02
a Republican sat outside of Bernie
00:03:04
Sanders second or third home and said,
00:03:06
"We should, this is his piano, this is
00:03:08
his summer home. We should add,
00:03:12
you know, taxes to And I'm sure you can
00:03:13
look up Bernie Sanders home pretty
00:03:15
easily. So it just before you point at
00:03:18
people's homes and say this is the
00:03:19
villain,
00:03:21
be careful folks because then if
00:03:22
something does happen to that person
00:03:24
like you know what happened to Sam Wman
00:03:26
this week and you feel however you want
00:03:27
about him, but nobody deserves to have
00:03:29
their house fire bombed or shot at.
00:03:32
Period. Full stop.
00:03:34
I just think that it's going to kill the
00:03:37
demand for, you know, maybe people who
00:03:39
already have a home in New York like Ken
00:03:42
Griffin, they probably are just going to
00:03:44
suck it up and pay the tax and keep
00:03:46
whatever they have. But if you were a
00:03:48
person who is thinking about buying a
00:03:51
pat in New York, there's no way you
00:03:53
would do it now.
00:03:54
>> Yeah.
00:03:54
>> Because you don't know what the tax rate
00:03:55
is going to be and it's going to keep
00:03:57
going up.
00:03:58
>> After a decade with interest and
00:03:59
inflation, you've effectively almost
00:04:01
doubled the price of your unit. So, if
00:04:03
you're going to buy a $10 million unit,
00:04:06
you're probably then looking at a $20
00:04:08
million purchase price just to break
00:04:11
even after about 10 or 11 years. That's
00:04:13
crazy. The math doesn't work anymore.
00:04:15
>> What is the downstream effect of these
00:04:17
individuals not coming to New York,
00:04:18
going to a Knicks game, going to a
00:04:20
restaurant? No, they spend money. They
00:04:22
spend money in New York.
00:04:24
>> They have their birthday party. They do
00:04:26
all kinds of things.
00:04:26
>> I'm not a big fan of the piet culture. I
00:04:28
I own one house. I don't like this whole
00:04:31
multiple houses you flitter flattering
00:04:32
everywhere. It ruined London because
00:04:35
when when you look at London
00:04:36
>> where do you vacation Shimoth? Where do
00:04:38
you vacation?
00:04:39
>> I go to a hotel.
00:04:40
>> Okay.
00:04:42
So my point is this that I think the way
00:04:44
that it works is it's not just pedair.
00:04:47
If you if you own a home that you fully
00:04:50
rent out but that is not the primary
00:04:52
residence of the person who is renting
00:04:55
it,
00:04:57
it's also a problem. So, this is any
00:05:00
even like a like a, you know, you had a
00:05:02
two-e rental or a oneweek rental or a
00:05:04
30-day rental or even a 5-year rental
00:05:07
and it was somebody else's second home,
00:05:10
the owner still gets gets hit. So, the
00:05:13
rental the sort of rentals of like folks
00:05:16
who want to come to that city or who
00:05:18
don't want to own in that city. Uh, but
00:05:22
it's a second place. The second place
00:05:23
thing goes away. So, maybe this is just
00:05:25
good for hotels. I think it's good for
00:05:27
hotels, but if you look at London, it's
00:05:29
probably the best example where there
00:05:31
was a lot of people who used London as a
00:05:33
place to store assets. Real estate
00:05:35
became the primary way in which they
00:05:38
would do that. It hollowed out parts of
00:05:40
London where it's not as if it was
00:05:43
unlivable. It's just that it was
00:05:44
unlived.
00:05:46
>> Nobody was there. You drive around
00:05:47
Chelsea, you drive around certain parts
00:05:50
of London and it was like a ghost town
00:05:52
on like a Thursday night or a Friday
00:05:54
night. that is an issue is the land
00:05:56
banking. It's like your bitcoin is
00:05:59
hollowing out a neighborhood and and
00:06:00
there is something there. But you know
00:06:02
what the whole uh housing thing is
00:06:05
complete cap and utter [ __ ] because
00:06:08
if you move to a place like Austin where
00:06:10
you are in Nevada or Florida, you see
00:06:12
what happens when you allow people to
00:06:14
build units in Austin three years in a
00:06:17
row rents and housing prices have gone
00:06:20
down while net migration has gone up.
00:06:23
But that's been good for Austin, you
00:06:24
guys would say. Isn't that good?
00:06:26
Incredible.
00:06:26
>> Yeah. Amazing.
00:06:28
>> Austin has like roughly doubles as the
00:06:30
city over the past decade. And yet the
00:06:32
rent for, you know, whatever a one or
00:06:34
two bedroom apartments gone down.
00:06:36
>> That's incredible.
00:06:37
>> So, in other words, if you let people
00:06:38
build to satisfy the demand, you won't
00:06:40
have this problem.
00:06:42
>> And then who's stopping the building?
00:06:45
>> It's Democratic cities. It's nimi
00:06:48
people. And then in a Republican town,
00:06:50
they're actually building units for an
00:06:54
affordability. So you have one group
00:06:55
saying they care about affordability and
00:06:57
they're doing nothing about it and
00:06:58
they're stopping it. And in another
00:07:00
place, they're like, "We're just going
00:07:01
to let you build because it's your right
00:07:03
to build because it's your land." That's
00:07:05
the approach in Texas. It's your land.
00:07:07
You have the right to develop it. Go.
00:07:09
>> By the way, the high end of the market
00:07:10
in London has basically turned over and
00:07:12
collapsed Sachs. to your point, like
00:07:14
they introduced the stamp tax, which I
00:07:16
think is equivalent to this tax that the
00:07:18
New York City mayor is proposing. And if
00:07:20
you look at London as a guide, the real
00:07:22
estate market just bid it at the high
00:07:25
end. And
00:07:26
>> I don't think that's going to be good
00:07:27
for that city or even the UK as a whole.
00:07:29
There's a handful of cities where people
00:07:31
do what you say, which is kind of park
00:07:33
money there. The reason why they do that
00:07:35
is because they believe that that city a
00:07:38
has the rule of law and b is a unique
00:07:40
worldass city that's going to keep
00:07:42
appreciating. And if you have new
00:07:44
management that doesn't really believe
00:07:45
in the rule of law that basically keeps
00:07:47
imposing all these arbitrary taxes. That
00:07:49
money is going to flee and find other
00:07:51
kinds of investments. People aren't
00:07:52
going to park their money there. That
00:07:54
has to be a bad thing for the city. It's
00:07:56
like, you know, who cares if the top
00:07:57
floor of that building where Ken Griffin
00:07:59
lives is owned by one guy who isn't
00:08:02
there that much? It's not going to
00:08:02
affect the city that much, but having
00:08:04
all these billionaires from all over the
00:08:06
world decide to park money in that
00:08:08
American city has to be good for the US,
00:08:11
just like it's good for the UK. And if
00:08:15
you give that up, then again, the money
00:08:18
will just go somewhere else.
00:08:19
>> Well, to your point,
00:08:20
>> the only thing is they're already paying
00:08:22
taxes on the property. And because
00:08:24
they're not there very often, they're
00:08:26
not using city services. So they're
00:08:29
paying taxes on the property. They're
00:08:31
not using city services.
00:08:33
>> They're profitable to the city. Yeah.
00:08:35
>> Think about a developer who's
00:08:36
underwriting some new project. You know,
00:08:39
the fact that a whale like Ken Griffin
00:08:42
is willing to overpay in the sense of
00:08:45
price per foot for that top floor might
00:08:48
make that whole project pencil. And this
00:08:50
is what contributes to the vitality of
00:08:52
New York is there's constantly
00:08:53
development going on. There's constantly
00:08:55
cranes. And so you take out that part of
00:08:58
the market that in effect was price
00:09:02
insensitive and was subsidizing all
00:09:04
these projects and I think development
00:09:06
is going to dry up to a large degree.
00:09:08
>> Mhm. He's paying he must be paying three
00:09:10
or four million in taxes every year and
00:09:12
getting no services to your point Travis
00:09:13
good point like all profit for the
00:09:15
>> London also did something else which is
00:09:16
that they essentially crippled what's
00:09:18
called non-dom status which is the big
00:09:20
tax orb if you moving or parking assets
00:09:23
in London and to your point Sax what did
00:09:26
all the rich people do they just
00:09:27
redirected themselves to Zurich to
00:09:29
Lugano to Milan and they they took
00:09:32
advantage of more hospitable tax policy
00:09:36
in other places now what the people
00:09:38
Britain would tell you, I actually met
00:09:39
with the UK government yesterday.
00:09:42
They don't see a meaningfully enough
00:09:44
measurable impact yet where they think
00:09:46
it's a five alarm fire. So the real
00:09:48
question is, is it more of a slow bleed
00:09:49
and a slow melt and at some point it's
00:09:52
just hollowed out and it's very hard to
00:09:54
reverse because there's not going to be
00:09:55
a cataclysmic acute moment where people
00:09:57
say, "Oh my god, we need to reverse
00:09:59
these policies." It doesn't seem like
00:10:01
that's in the offing. You know, LA did
00:10:03
something a little different but similar
00:10:05
is they introduced that 5% mansion tax.
00:10:08
This on like all the areas except for
00:10:10
like Beverly Hills and
00:10:11
>> San Francisco has that too.
00:10:13
>> Yeah, San Francisco got that too. I
00:10:14
think it's actually like 6% in San
00:10:16
Francisco.
00:10:16
>> San Francisco. It's 6% over 25 million
00:10:20
and it starts at 5 million. So there's a
00:10:22
transfer tax, an excess transfer tax
00:10:24
above $5 million properties that scales
00:10:27
up above 25 million to an extra 6% on
00:10:30
top of your brokerage commission. So
00:10:32
when you sell a house in San Francisco,
00:10:33
you're paying 13% now.
00:10:37
>> Which is why you look at the transaction
00:10:39
volume in the LA real estate market and
00:10:41
it's just completely dried up. You know,
00:10:43
people aren't doing the house flipping
00:10:44
anymore, like that kind of stuff,
00:10:46
because the transaction costs are too
00:10:47
high. But it just shows I mean again
00:10:50
this tax was imposed retroactively. For
00:10:52
example, I had my house in SF and then
00:10:54
they just take you know a couple of
00:10:55
rooms of it
00:10:58
and we talked about this on the show
00:10:59
when it happened.
00:11:01
But my my point is just your property is
00:11:03
not safe in blue states and wealthy
00:11:06
people who have a choice of where to
00:11:08
park their money are going to
00:11:09
increasingly realize that and they're
00:11:12
not going to buy I think real estate in
00:11:14
blue states is dangerous because the
00:11:16
political class thinks that they can
00:11:18
take a chunk of it and you know the
00:11:20
wealthy people are going to react to
00:11:21
that and they're going to move their
00:11:22
money elsewhere.
00:11:23
>> All right, let's go to topic number one.
00:11:25
Open AI is apparently suffering from a
00:11:28
bit of an identity crisis on Sunday.
00:11:31
Open AI's chief revenue officer Denise
00:11:33
Dresser sent a four-page memo to
00:11:35
employees. Obviously, it leaked
00:11:37
immediately, probably the point of it,
00:11:39
and she called out Enthropic. She said
00:11:41
their $30 billion run rate is cap
00:11:44
inflated by 8 billion due to a revenue
00:11:47
share and some accounting with AI model
00:11:50
providers. Chimati pointed that out in
00:11:51
the last couple weeks. Also, she said
00:11:54
anthropic stories built on quote fear
00:11:56
restriction and the idea that a small
00:11:58
group of elites should control AI.
00:12:01
Obviously, she's a fan of the pod. She
00:12:02
also laid out OpenAI's pivoting and she
00:12:06
said they are going hard after business
00:12:08
customers and they want to win the agent
00:12:12
platform layer. If you remember, they
00:12:14
hired the architect of the open source
00:12:16
project openclaw. They didn't acquire
00:12:18
openclaw. So Peter uh Steinberger is
00:12:21
working at Open AI. Cynical people said,
00:12:24
"Hey, maybe they want his next set of
00:12:25
innovations to go inside of OpenAI's
00:12:27
products as opposed to the open source
00:12:29
one." I kind of agree with that
00:12:30
directionally.
00:12:31
There's obviously Perplexity Computer is
00:12:33
doing really well. They quadrupled their
00:12:35
revenue. And uh I was over at XAI
00:12:38
earlier this week with Elon and I can
00:12:40
tell you he's got some very cool stuff
00:12:42
coming. And this new model, Spud, is
00:12:45
coming uh from Open AI. Here's your poly
00:12:48
market. 75% chance Spud is released next
00:12:52
week.
00:12:54
Additionally, and we'll go to the panel
00:12:56
in just a second. On Tuesday, 2 days
00:12:59
after this memo came out, obviously
00:13:00
people write these memos to go directly
00:13:02
to the press. So, they're obviously
00:13:04
trying to undercut Anthropics valuation
00:13:06
and they feel that's a threat. That's my
00:13:07
take. The FT cited anonymous OpenAI
00:13:10
investors who are frustrated with the
00:13:13
company's lack of focus. Here's the
00:13:14
anonymous quote. quote, "You have chat
00:13:17
GPT a 1 billion user business growing 50
00:13:20
to 100% a year. What are you doing
00:13:23
talking about enterprise and code? It's
00:13:24
a deeply unfocused company and uh we
00:13:28
talked about this chat GPT's market
00:13:29
share is going down as the number of
00:13:31
users is going up because Gemini and
00:13:34
Claude gaining significantly and Medage
00:13:37
just last week released their first
00:13:39
proprietary model and Apple doesn't have
00:13:42
a product in market yet, but they do
00:13:44
have a lot of users. So here's your Gen
00:13:46
AI website traffic.
00:13:50
Let's start uh Chimath with you. Your
00:13:52
thoughts on
00:13:54
OpenAI. Should they be pivoting straight
00:13:57
into business developers and getting
00:13:59
focused on that or should they stay
00:14:00
focused on the consumer where they are
00:14:02
the verb? as it relates to complex long
00:14:06
horizon coding tasks. What I can tell
00:14:08
you from my team at 8090 is codeex is
00:14:11
better
00:14:13
generally than anthropic. And so what
00:14:15
happens is for the more day-to-day work,
00:14:19
I think it's more reliable to use the
00:14:22
clawed ensemble of models, but when
00:14:25
you're dealing with something that's
00:14:26
very tricky and very complicated and a
00:14:28
little bit more long horizon,
00:14:31
codeex is really functional and really
00:14:33
good. So I think if you're looking at it
00:14:37
through the lens of open AI, what
00:14:39
they're probably saying is, hey, hold on
00:14:40
a second. If we allocate more resources
00:14:43
and just double down and crush consumer,
00:14:46
that's probably three or four trillion
00:14:49
of enterprise value.
00:14:51
And then if we slowly refocus the
00:14:53
company with the rest of the resources
00:14:54
and double down on codeex and do
00:14:56
something meaningful in enterprise, we
00:14:58
can probably capture two or three
00:15:00
trillion there. And now all of a sudden
00:15:02
you can paint a picture for a seven
00:15:05
eight n trillion dollar market cap in
00:15:06
the fullness of time. Not tomorrow
00:15:08
obviously, but Codeex is really good.
00:15:10
and there's a business to be had in
00:15:12
both. You have to separate the two
00:15:14
businesses. You can't have a lot of
00:15:15
overlap because there's too much context
00:15:17
switching. You got to let the consumer
00:15:18
team run and then you got to isolate the
00:15:21
enterprise team and let them do what
00:15:22
they think is right.
00:15:23
>> Travis, uh there's a big debate going on
00:15:25
amongst the investors. The FTPS also
00:15:28
questioned the $850 billion valuation of
00:15:31
OpenAI. Secondary markets have now
00:15:34
priced Anthropic higher than OpenAI for
00:15:36
the first time. This is the flipping
00:15:38
that people predicted. One investor said
00:15:40
OpenAI would need to IPO at a valuation
00:15:42
of 1.2 trillion for the last round to
00:15:44
make any sense, but there's no buyers
00:15:48
currently at the $850 billion valuation
00:15:51
that OpenAI just closed. According to
00:15:54
Bloomberg, Travis, how do you uh
00:15:57
handicap this race between these two
00:15:59
leading frontier models?
00:16:01
>> Growth is king right now in this in this
00:16:04
world uh in this segment. Growth is the
00:16:06
whole damn thing. And if Anthropic is
00:16:10
growing faster than OpenAI by a
00:16:12
significant clip, the investors right
00:16:15
now are going to play it forward. And
00:16:17
they're it, you know, you start to get
00:16:20
network effects around compute, network
00:16:22
effects around the number of tokens
00:16:24
you're pushing out for various
00:16:25
customers, enterprise or consumer. And
00:16:29
ultimately, you know, it's not great
00:16:30
today, but how that plays into
00:16:33
reinforcement learning and the things
00:16:34
getting smarter over time. There's so
00:16:37
many there's so much upside to volume
00:16:41
and scale that if they are growing
00:16:43
faster at the same size,
00:16:47
even if OpenAI is still growing, but if
00:16:49
they are half the growth rate, a third
00:16:51
the growth rate, a fifth the growth
00:16:53
rate,
00:16:55
I would be I'd be worried.
00:16:57
>> And you saw this at Uber specifically
00:17:00
when you accelerated away from the
00:17:02
competitors like Lyft and Door Dash.
00:17:04
Yeah.
00:17:04
>> Yeah. you had to network effects was the
00:17:07
whole thing at the end of the day and
00:17:09
network effects was based on scale. So
00:17:11
yes, I'm I'm sort of like coming from my
00:17:13
very specific experience. But if you
00:17:15
believe there's network effects with the
00:17:18
scale of data that you have and the
00:17:19
scale of customers and the revenue that
00:17:22
this is cash that's coming in that you
00:17:24
redeploy into compute and so say there's
00:17:27
network effects of large compute
00:17:30
I'd be very worried uh if I'm open AI
00:17:33
and seeing somebody growing faster at
00:17:36
the same size. So Freeberg, when you
00:17:38
look at this race between these two
00:17:40
giants, maybe your thoughts on the
00:17:42
flywheel as it relates, as Travis is
00:17:44
pointing out to advantages in compute,
00:17:48
reinforcement, learning, and then also
00:17:50
the ability to fund raise. One of the
00:17:52
great things that Travis and the team
00:17:54
did was as they were pulling away, they
00:17:57
just sucked all the oxygen out of the
00:17:59
room by using capital as a weapon. Uh,
00:18:02
so your thoughts, Dave, on this high
00:18:03
stakes game because there's also a point
00:18:05
at which, and I'll just end on this.
00:18:07
There's a point at which
00:18:08
>> you could run off the cliff. You raise
00:18:10
so much money
00:18:11
>> and you deploy it so fast and the
00:18:13
revenue doesn't catch up to it and then
00:18:14
you go public and the markets don't
00:18:16
believe the story. So, your thoughts,
00:18:17
Free, Bird,
00:18:18
>> I don't know the financials of the two
00:18:19
companies well enough. Obviously, Sam
00:18:21
has no problem raising money. The guy,
00:18:23
didn't he just close like $150 billion
00:18:25
round or something?
00:18:26
>> 122 billion.
00:18:28
>> 122 billion. largest round ever raised
00:18:30
in any market I think private or public.
00:18:32
>> Yeah, probably. Yeah, I mean that's
00:18:33
crazy.
00:18:35
>> That doesn't seem to be an issue. What
00:18:36
I've noticed is just the the pace of uh
00:18:39
innovation at Anthropic is from my
00:18:42
experience unprecedented. I mean,
00:18:44
they're release cadence is
00:18:46
extraordinary. They've basically
00:18:48
supplanted Open Claw already with this
00:18:51
release they did a few days ago and then
00:18:52
today the new Opus model got dropped. So
00:18:55
there's something about the momentum,
00:18:58
not in necessarily just in user growth,
00:19:00
but in how they're operating this
00:19:02
business that just seems to be head and
00:19:04
shoulders above everyone else in the
00:19:06
cadence of of upgrades. If I look back 6
00:19:09
months ago, I think we were pretty heavy
00:19:10
on Cursor and Gemini. And now I think
00:19:14
we're probably 90% anthropic in just the
00:19:16
last 6 months at my my organization.
00:19:19
There's something very powerful about
00:19:20
the flywheel they have going on.
00:19:22
>> Yeah. Oh, here's where I'd go just real
00:19:24
quick, Dave, is
00:19:27
mad respect on a$1
00:19:30
120 or $130 billion raise. I mean,
00:19:33
>> this is obviously next level, but you
00:19:36
can use capital and investment to
00:19:39
acquire scale and network effects
00:19:41
associated with it. But if somebody is
00:19:44
getting that scale
00:19:46
with revenue and let's call it
00:19:48
contribution margin, contribution
00:19:50
profit,
00:19:52
efficiency will outstrip subsidy and it
00:19:55
will you can't just keep raising hundred
00:19:57
billion dollar things forever. That's
00:19:59
where the train will stop. And if
00:20:02
Anthropic is funding theirs through
00:20:04
revenue and other folks are funding it
00:20:07
through investment, there's like a
00:20:10
shortterm that's a short-term solve, but
00:20:13
it the long run is whoever is scaling
00:20:16
their actual usage and system and
00:20:19
ultimately with contribution profit that
00:20:22
then soaks up the need for investment.
00:20:25
That's a that's a that's a very scary
00:20:27
machine if you're competing against it,
00:20:30
>> which is exactly sachs what the um
00:20:33
legacy Mac7 are doing. you have massive
00:20:36
profits from uh Meta's core business,
00:20:40
Google's core business that are being
00:20:42
redeployed into infrastructure and even
00:20:44
uh Tesla which has a lot of profits and
00:20:47
SpaceX which has a lot of cash on hand
00:20:49
and they're building out Colossus and
00:20:51
other assets including uh Elon is
00:20:54
working on building a fab as folks have
00:20:56
been talking about. So, Sax, is there a
00:20:59
chance for the legacy companies, the
00:21:01
MAG7s to compete in this, or are we
00:21:04
looking at OpenAI and Anthropic are one
00:21:06
and two and then everybody else can uh
00:21:09
fight for third place and and the and
00:21:11
the and the bronze as it were?
00:21:14
>> Well, I think Google's clearly in the
00:21:15
mix. I mean, Deep Mind has an
00:21:17
outstanding team and I think Elon's
00:21:18
still in the mix with XAI and then
00:21:21
you've got Meta also has the resources
00:21:24
and they're seem to be further behind,
00:21:26
but they're going to compete. Look, let
00:21:28
me just go back to the central premise
00:21:29
here. I agree that there's some valid
00:21:33
criticism that OpenAI has been unfocused
00:21:38
and should be, you know, moving forward
00:21:41
more focused in what they do. I have no
00:21:44
idea, for example, what they're doing
00:21:46
buying a podcast that's not us. So, I
00:21:49
don't know what what that was about. Um,
00:21:52
if you were gonna buy a tech podcast for
00:21:54
a few hundred million dollars, I mean,
00:21:56
um, we were here.
00:21:57
>> We're here.
00:21:58
>> We're here for Dario, reach out. Small
00:22:01
potatoes for you guys though. Small
00:22:03
potatoes. They can afford you. They
00:22:05
can't afford you guys.
00:22:06
>> Yeah. They thought we were too
00:22:07
expensive.
00:22:08
>> Little did they know we would have sold
00:22:09
out. But, um,
00:22:11
>> punch the ticket, Daria. Yeah,
00:22:14
>> but look, this other part of the
00:22:16
criticism of Open AI that they shouldn't
00:22:19
do enterprise is totally misguided. One
00:22:22
of the reasons why they should have been
00:22:23
more focused is to do more enterprise
00:22:25
and get enterprise more correct. Now,
00:22:28
why do I say that? To Travis's point
00:22:30
about growth rates, it's true that
00:22:34
OpenAI and Anthropic as of the beginning
00:22:37
of Q2, so let's say two weeks ago, they
00:22:40
were both around 30 billion of revenue.
00:22:41
And that memo from that open AAI
00:22:43
employee was right that if you compare
00:22:46
them on an applesto apples basis that
00:22:48
anthropic is about 20% less because they
00:22:51
are including revenue made by their
00:22:53
channel partners. But that doesn't
00:22:55
matter. What matters is the growth rate.
00:22:57
You know again to Travis's point and let
00:23:00
me just put some numbers around this.
00:23:02
The open AI growth rates been around 3
00:23:04
to 4x a year. The anthropic growth rate
00:23:06
has been around 10x a year. So they went
00:23:09
from let's call it 1 to 10 billion of
00:23:12
ARR last year and by the end of Q1 this
00:23:16
year they were already at 30 billion of
00:23:18
again let's call it their revenue and
00:23:21
they're on their way you know like Brad
00:23:23
Gerson was saying on our podcast I think
00:23:25
in the last couple weeks they're going
00:23:26
to end this year at 80 to 100 billion at
00:23:29
least on the current trajectory and so
00:23:31
you can plot their revenue on a
00:23:33
logarithmic graph I mean again no one's
00:23:35
ever seen anything like before where
00:23:37
every unit on the y- axis is another
00:23:40
expon is 10x and it's a straight line
00:23:43
>> now it's crazy
00:23:44
>> it's crazy right so if it's taking
00:23:46
anthropic let's say one year to 10x and
00:23:48
it's taking open AAI 2 years to achieve
00:23:51
a 10x then it's obvious which one's
00:23:54
going to win now what is the reason for
00:23:56
this is because anthropic was very
00:23:59
focused on enterprise specifically
00:24:01
coding and what you're seeing is that
00:24:04
businesses are willing to pay for coding
00:24:06
code tokens
00:24:08
on a meter basis, let's call it like
00:24:10
electricity. The more they use, the more
00:24:12
they're willing to pay and their usage
00:24:15
just continues to scale and scale.
00:24:17
Consumer is completely different. I
00:24:18
mean, consumer is a thing that OpenAI
00:24:20
prioritized. Consumers have a lower
00:24:22
willingness to pay. Maybe only 3 or 4%
00:24:24
of them are willing to convert to
00:24:25
premium in the first place. And what
00:24:27
they want is a $20 a month all you can
00:24:30
eat subscription. So, the revenue simply
00:24:32
doesn't scale the same way that
00:24:33
enterprise does. And so if you want to
00:24:36
tap into the scalable revenue source in
00:24:38
the market right now, you have to go
00:24:39
after enterprise. So you know again
00:24:42
where I would agree with the criticism
00:24:44
of OpenAI is maybe they should have been
00:24:47
more focused but they need to be more
00:24:48
focused specifically to pursue coding
00:24:52
and enterprise and if they don't catch
00:24:54
up soon to Travis's point then you could
00:24:57
see Anthropic taking a lead here that
00:24:59
let's say over the next one or two years
00:25:01
could be insurmountable. Just by the
00:25:03
way, let me just say say one thing is
00:25:05
even though anthropics revenue has
00:25:08
followed this graph, this exponential
00:25:09
graph very predictably, it can't do that
00:25:12
forever, right? Like let's say it does
00:25:14
get to
00:25:15
>> 80 to 100 billion this year. Can it
00:25:17
really get to a trillion dollars in
00:25:19
revenue the year after that seems hard
00:25:21
to believe, right? And and and the
00:25:23
reason is because as you hit new levels
00:25:25
of scale, you encounter new problems. I
00:25:27
mean, you're simply going to run out of
00:25:29
compute or electricity, data centers,
00:25:32
you know, infrastructure. There are
00:25:34
physical limits or there's limits in the
00:25:35
physical world that you're going to hit.
00:25:37
And there's already some evidence that
00:25:39
Anthropic is hitting some of those
00:25:40
limits. Users were complaining, for
00:25:42
example, that Claude was thinking less.
00:25:45
Did you guys see this? That
00:25:47
>> that, you know, a typical Claude prompt,
00:25:49
they they seem to have cut down on the
00:25:50
thinking time by about 2/3. Now, I saw
00:25:53
someone tweeting today that they just
00:25:55
released Opus 4.7, replacing Opus 4.6,
00:25:58
and the thinking is back, but you know,
00:26:00
maybe they're charging more for that.
00:26:02
Hard to say, but they're going to hit
00:26:03
some sort of physical limits. And I do
00:26:05
wonder if over the next year Anthropic
00:26:08
will reconsider whether it's support for
00:26:11
all this like doomer nimiism was the
00:26:14
right call because it kind of made sense
00:26:17
for them from a business standpoint when
00:26:20
their competitors were building data
00:26:21
centers and they were just getting
00:26:22
compute from the hyperscalers. But now
00:26:25
that they're I think going to have to
00:26:27
move into the game of building their own
00:26:28
data centers, they might regret salting
00:26:30
the earth for data centers all over the
00:26:33
country. And I wonder if that'll be the
00:26:35
natural limit of their growth is they'll
00:26:38
be hoisted on their own petard of doomer
00:26:41
nimism.
00:26:42
>> And they're also using the uh coding
00:26:45
platform to build anthropic itself
00:26:47
better. So that in and of itself is a
00:26:49
reason to nail coding. You get the
00:26:51
double whammy. You can make a better
00:26:53
product and you can get paid for it.
00:26:55
>> Yes and no. Hold on. Because if you look
00:26:56
on Twitter, people are panning cloud
00:26:58
desktop and what they said is this is
00:27:00
all a bunch of vibecoded slop. I think
00:27:02
we have to remember if you keep these
00:27:03
agents on task and they're guard railed
00:27:06
properly, this stuff is a force
00:27:08
multiplier. The problem is nobody knows
00:27:11
how to do this really well yet. Nobody
00:27:13
has built real products of scale largely
00:27:16
using agents yet. Nobody knows how to
00:27:19
give an example of how an org structure
00:27:21
should be redefined. Nobody knows how to
00:27:24
budget properly. We had the CTO of Uber
00:27:26
say, "I give up. I've hit my token
00:27:28
budget." The problems I see it are
00:27:30
twofold.
00:27:32
The first is just to build on Sax's
00:27:35
point. All of these frontier labs have a
00:27:37
very serious issue which is both open
00:27:39
AAI and Anthropic are growing so fast
00:27:42
that they're at a point now where they
00:27:44
need their own infrastructure. It's kind
00:27:46
of like when you first start building
00:27:47
any kind of company, you're just much
00:27:49
easier renting capacity from the
00:27:52
hyperscalers
00:27:53
>> and it's a dependency. Yeah.
00:27:55
>> But then it becomes a dependency.
00:27:56
Exactly. And now when you're so big,
00:27:58
it's actually strategically a huge
00:28:01
mistake to not have your own comput.
00:28:04
Why? Because if you look at who's
00:28:06
leading, the frontier labs are leading.
00:28:08
And Sax, to your point, you mentioned
00:28:10
this on X. You're like, there was all
00:28:11
this doomerism, but maybe it was tied to
00:28:13
compute capacity because when Bedrock
00:28:16
opened up more capacity for anthropic,
00:28:17
all the doomerism went away. You're left
00:28:19
wondering like, is it really tied to
00:28:21
just the fact that they were just trying
00:28:22
to throttle usage? So if you're a
00:28:24
frontier lab, you don't want to have to
00:28:26
go through Amazon and GCP and Azure and
00:28:28
Tin Cup for access and capacity. What
00:28:31
you'd much rather have is go straight to
00:28:33
your customer. On the other side, if
00:28:36
you're Gemini or
00:28:38
Microsoft or Meta and you have all this
00:28:41
compute, because I saw a stat this week,
00:28:43
the hyperscalers control 60% of all the
00:28:46
compute. So the game theory there is if
00:28:49
you kneecap the Frontier Labs, it'll
00:28:50
give you some chance to catch up and it
00:28:52
gives you time to catch up because no
00:28:54
matter what the demand is on the upside,
00:28:56
you remember you guys remember like in
00:28:57
social networking when Fster was the
00:28:59
cat's meow.
00:29:01
>> Yeah.
00:29:01
>> Cat's meow.
00:29:03
>> Remember what the biggest problem at
00:29:04
Fster was?
00:29:04
>> Cats meow.
00:29:06
>> Fster was slow as a dog.
00:29:07
>> Yes.
00:29:08
>> Yes. Yes.
00:29:09
>> And what happened? Then MySpace came in
00:29:11
and took all the share. Then we came and
00:29:12
Facebook and we took all their share. So
00:29:15
there is a way where you can handicap
00:29:16
and kneecap these companies by
00:29:18
throttling compute access to them. So a
00:29:22
they are forced to now go and get in the
00:29:24
game which is weird because look open
00:29:26
AAI has tried to displace some of the
00:29:29
Stargate spend. I don't see any path
00:29:32
except they're going to have to do it
00:29:33
themselves and anthropic will have to do
00:29:34
it themselves. But then separately, the
00:29:36
other problem is when you change the
00:29:38
subscription model in enterprise and you
00:29:41
say, "Hey, we're not going to subsidize
00:29:42
any more tokens."
00:29:44
What's going to happen is all these
00:29:46
token budgets are going to go crazy. And
00:29:47
what Freebrook said is going to happen
00:29:48
where he's like, "Hey guys, why are you
00:29:50
spending all this money? What are you
00:29:51
making?" And you inspect the code and
00:29:53
you're like, "What is this slop?"
00:29:55
And you're not going to add 30 40 50%
00:29:58
opex
00:30:00
to produce nothing. So I think that
00:30:02
that's an open question and that
00:30:03
question will become more amplified over
00:30:05
the next year as they push the cost off
00:30:09
of them. So as Travis said, no more
00:30:10
subsidy from the capital.
00:30:13
You have to grow into it, but you're not
00:30:14
going to support negative gross margins.
00:30:16
So you're going to pass through the
00:30:17
token costs. I So I think it's a very
00:30:19
dynamic moment right now for these. If
00:30:21
you look at the ranking of these
00:30:24
clusters and who has the most right now,
00:30:27
uh people have forgotten about Colossus,
00:30:28
which Elon's been building, he's
00:30:30
expanding to 555,000
00:30:33
GPUs across three buildings, 18 billion
00:30:36
in investment. He then if you look at
00:30:39
Prometheus Meta's planned 2026, that's
00:30:42
150,000 GPUs. So this is uh
00:30:45
>> well Elon just announced a deal. He
00:30:47
announced a deal this morning with
00:30:48
Kurser. So Elon's Elon's renting a bunch
00:30:51
of capacity. So he's now getting
00:30:52
effectively into the data center
00:30:53
business. He's going to be a hypers
00:30:55
skill.
00:30:56
>> So he's he's going to use as much as he
00:30:58
can for XAI and whatever is left over
00:30:59
he'll give to well in this case he's
00:31:01
giving to cursor to train their model.
00:31:03
He could
00:31:03
>> what you're saying is you might as well
00:31:04
overbuild capacity because that way your
00:31:08
own models would be in a privileged
00:31:09
position and you can sell the rest to to
00:31:11
your competitors. But to Jamath, to your
00:31:13
point about the thing I was saying on X,
00:31:15
I actually I think I was retweeting Mark
00:31:17
Andre who pointed out that one of the
00:31:20
reasons why Anthropic might have wanted
00:31:22
to hold back Mythos is they simply
00:31:24
didn't have the compute to serve it. The
00:31:26
model was huge and very expensive to
00:31:28
serve something like maybe even 10 or 20
00:31:31
times the token cost of of say Opus.
00:31:34
They knew Opus 4.7 was coming out,
00:31:36
right? So they hold it back knowing that
00:31:39
they don't have the compute to serve it
00:31:40
anyway and they save their compute for
00:31:43
the next iteration of Opus. And then by
00:31:45
holding it back they create this
00:31:47
impression of scarcity and altruism and
00:31:50
it turns into this gigantic marketing
00:31:51
event for their product because everyone
00:31:54
in the government's like, "Oh wow,
00:31:55
they're holding it back because it's so
00:31:57
amazing." Now look, I think it may have
00:32:00
been genuinely altruistic as well in the
00:32:03
sense that Mythos does reveal coding
00:32:05
vulnerabilities that people didn't know
00:32:08
about before. And you know, it does make
00:32:10
sense to give time to companies with
00:32:12
large code bases to patch these dormant
00:32:15
bugs and vulnerabilities, but it's
00:32:18
looking more and more like Anthropic
00:32:20
could not have offered that model
00:32:22
commercially anyway because it was just
00:32:24
too big and expensive and they need to
00:32:26
create space for Opus 4.7. So, it's an
00:32:30
interesting theory on what actually
00:32:32
happened there.
00:32:32
>> All right, guys. Uh before we go to our
00:32:34
next story, some breaking news here. Uh
00:32:36
here's your poly market, gentlemen. And
00:32:38
I don't this is I don't know if you're
00:32:39
placing some insider bets here uh
00:32:41
Freeberg but looks like the All-In
00:32:43
podcast uh anthropic now 37% chance of
00:32:47
buying the All-In podcast. This is live
00:32:48
>> by what time? By
00:32:50
>> poly market. Yeah, this is by the end of
00:32:51
the year.
00:32:52
>> By end of the year. Okay, got
00:32:53
>> by end of the year. So it's
00:32:55
>> this isn't real, is it?
00:32:56
>> Yeah, absolutely. They've been People
00:32:58
have been trading on this. This is
00:32:59
heavily traded.
00:33:00
>> It can't be. It says $92 million of
00:33:02
volume.
00:33:03
>> Yeah, this is the number three largest
00:33:06
market. There's no way this is real,
00:33:07
dude. There's no way.
00:33:08
>> It's breaking news, guys. I don't I
00:33:10
don't control the news flow.
00:33:12
>> This is Jake slump.
00:33:13
>> This is no way this is real.
00:33:15
>> Anybody can write a prediction market, I
00:33:17
guess. Hey, what's the what's the uh
00:33:19
volume of that thing?
00:33:20
>> 92 million.
00:33:21
>> It's not real.
00:33:22
>> Is there $2 behind it?
00:33:24
>> There's no way that's real, guys. Come
00:33:25
on,
00:33:26
>> guys. It's all good upgrade,
00:33:28
>> guys. Time to upgrade the planes. I just
00:33:31
Hey, guys. Just I don't need the PC24.
00:33:33
I'll take somebody's G650. Whoever's got
00:33:35
a G650, go up to the 800. It's going to
00:33:38
be trickle down to JCAL. Let's do it.
00:33:40
I'm I'm giving up my United Platinum
00:33:42
status.
00:33:43
>> Trickle down economics.
00:33:44
>> Trickle down avionics. All right,
00:33:45
listen. Story number two.
00:33:47
>> Triple triple down.
00:33:50
That's good. This is pretty good. Okay,
00:33:53
number two. All birds, speaking of data
00:33:56
centers, all birds just pivoted from
00:33:58
ugly sneakers to scratiad to AI and the
00:34:02
stock has ripped. talking about peak
00:34:05
bubble behavior podcast getting bought
00:34:07
by Frontier models and sneaker companies
00:34:10
uh pivoting to data centers. All birds
00:34:13
as you know is the ugliest sneakers on
00:34:15
the planet and uh this became a massive
00:34:20
delusion in our industry that this
00:34:21
company was worth billions of dollars.
00:34:24
They went public in 2021. This might be
00:34:27
one of the peak zer moments. Raised $350
00:34:29
million in their IPO. Sax, were you an
00:34:33
investor in this thing?
00:34:34
>> All birds.
00:34:35
>> All birds back in. No,
00:34:37
>> no, no, no. You're thinking about the
00:34:38
scooter company.
00:34:40
That's right.
00:34:41
>> Don't Don't remind me of the the
00:34:43
investments that didn't work.
00:34:44
>> But no,
00:34:46
>> Bird was crazy.
00:34:48
>> This reminds me of the late 90s where
00:34:51
all you had to do was change your name
00:34:52
to whatever.com.
00:34:54
>> Yeah. Yes.
00:34:55
>> And you get a huge pop in your valuation
00:34:57
>> and you could spit it out. Barnes &
00:34:58
Noble.com became then Barnes & Noble.
00:35:02
And so the stock crashed and never
00:35:05
looked back. There's your Stark chart.
00:35:06
Um they sold off all their brand assets
00:35:08
for 39 million, about 10% of what they
00:35:11
raised in the IPO. Congratulations to
00:35:13
whoever owns those ugly ass sneakers.
00:35:15
>> Why were they worth 4 billion? They sold
00:35:18
>> collective delusion. Collective
00:35:19
delusion. People in Silicon Valley liked
00:35:21
them and they just thought it was the
00:35:22
next Nike.
00:35:23
>> Tulips.
00:35:24
>> It was tulips. Well, I think I think
00:35:26
that was an era in Silicon Valley where
00:35:29
people rewarded rapid growth without
00:35:32
really looking at gross margins or cost
00:35:34
of goods sold. People didn't really make
00:35:36
the distinction between software and
00:35:37
everything else, right? Software, you
00:35:40
never really had to worry about COGS or
00:35:43
gross margin because the incremental
00:35:45
cost of serving a customer with software
00:35:48
is like almost zero. So people in
00:35:50
Silicon Valley weren't really trained to
00:35:51
look at gross margin and there was this
00:35:53
rash of physical world companies that
00:35:56
all of a sudden started getting crazy
00:35:58
valuations. Travis, quite frankly, you
00:36:00
might have the success of Uber might
00:36:02
have ushered in this era of, you know,
00:36:04
these physical world companies that
00:36:06
started getting valued like software
00:36:07
companies even though obviously they
00:36:09
didn't deserve it.
00:36:10
>> David, thank you so much for that shout
00:36:12
out. I really appreciate it.
00:36:13
>> You and Airbnb your fault. Anyway, let's
00:36:16
get back to
00:36:16
>> No, no, no, no. Hold on, hold on. The
00:36:18
era of 21 is super interesting though
00:36:20
guys. I don't believe it was a physical
00:36:22
versus digital thing. I believe it was a
00:36:26
moment in time in the COVID zer massive
00:36:30
money going in. We hadn't seen inflation
00:36:32
yet. It was all of that happening and
00:36:35
the investor class was basically
00:36:37
deciding we're going to look two and
00:36:39
three years forward on your current
00:36:41
growth.
00:36:43
And it wasn't one year forward. It was
00:36:45
like two or three years forward. And
00:36:47
that's where these crazy valuations got
00:36:48
weird. So if you went from 0 to, you
00:36:51
know, if you went from a hundred to $300
00:36:53
million in the last year, they'll play
00:36:56
that two, three years forward and go,
00:36:57
"Oh yeah, you're totally 20 times
00:37:00
bigger. We'll pay you for that now."
00:37:02
>> Yeah,
00:37:02
>> that's where it got weird.
00:37:04
>> It got weird. Hey, Bird Scooters was
00:37:06
valued at two or three billion. They
00:37:07
were doing micromobility.
00:37:09
>> You're just trolling them, Jay. I'm No,
00:37:11
we're giving everybody their flowers for
00:37:14
incredible investments and even even the
00:37:17
mighty sachs could trip up. Did you sell
00:37:19
IPO?
00:37:20
>> Yeah.
00:37:20
>> No. No, we don't do that.
00:37:22
>> Uh it was a series A we did. I mean,
00:37:24
look, we we got it right in the sense
00:37:26
that it was a total phenomenon,
00:37:28
>> but then the cities just cracked down on
00:37:30
it and killed it. Look, if the cities
00:37:31
had leaned into it Yeah. Look, if the
00:37:33
cities had reacted differently, if they
00:37:34
had leaned into it, if they had created,
00:37:36
let's say, a scooter or like a small EV
00:37:40
lane, they could have transformed
00:37:42
cities. I mean, it would have been a lot
00:37:43
easier to get around. But instead, they
00:37:44
banned it. They limited it. They didn't
00:37:46
create designated areas for it. And then
00:37:50
the the cuda grass was basically when
00:37:52
they would take a city where Bird
00:37:54
already had like 80% dominant market
00:37:56
share and then they would say that well
00:37:58
we're going to choose four operators and
00:38:00
give them each 25% of say a thousand
00:38:03
scooter allocation. That just killed the
00:38:05
economics for everybody.
00:38:06
>> Yeah. There's no market anymore.
00:38:07
>> There's no market. Yeah.
00:38:08
>> Yeah.
00:38:09
>> There's no marketplace network effect,
00:38:11
right? When when they're basically just
00:38:12
picking the winners and deciding the
00:38:13
market.
00:38:14
>> Yeah. You can't you can't compete
00:38:16
>> and splitting it even. So then
00:38:17
>> yeah, there's no there's no competition.
00:38:19
You can't drive value.
00:38:20
>> Regulatory capture at its worst,
00:38:22
>> guys. Like we should watch for this in
00:38:24
the in the autonomous car space too.
00:38:26
Cities may get cute and start doing
00:38:28
things like that. And and what they did
00:38:30
on scooters, they could do on cars.
00:38:33
>> That's a really good point. Yeah, you're
00:38:34
right. Instead of just letting the
00:38:36
market
00:38:37
>> play out, they say, "Well, we're only
00:38:38
going to have x number a thousand
00:38:41
autonomous cars.
00:38:42
>> New York is literally doing this right
00:38:44
now.
00:38:46
New York and Boston
00:38:46
>> and then the prices prices go up. The
00:38:49
innovation doesn't get realized.
00:38:50
Consumers don't benefit. No one
00:38:52
benefits. Basically, you've deleted the
00:38:54
market.
00:38:55
>> Yes. And anyway, just to wrap up this
00:38:58
story on
00:39:00
the the shoe company, they're now Newird
00:39:01
AI. They bought eight H100s, I think,
00:39:05
with $50 million in a convertible note.
00:39:07
And the stock has gone up.
00:39:10
>> Not eight. Stop. Eight. Really? No, it's
00:39:12
not.
00:39:12
>> No, it's a joke. Obviously, the stock's
00:39:14
now at $14 a share. It's up 450% in the
00:39:17
last week. Shout out to Wall Street Bets
00:39:19
for staying Can I say something more
00:39:22
than the shorts can stay solvent.
00:39:25
>> Can I say something serious?
00:39:27
>> Yes, please.
00:39:27
>> There are a handful of transactions that
00:39:29
have happened in the last few days that
00:39:32
if you look far away are head
00:39:34
scratchers. So, this is one. I don't
00:39:38
know if you guys saw recently, but Jane
00:39:40
Street did a a billion dollar investment
00:39:42
in essentially a neo neocaler and then
00:39:46
also did a $6 billion compute deal with
00:39:49
them. That was like a little
00:39:51
interesting. So what would I like to say
00:39:54
about this? I think the thing that the
00:39:55
capital markets are getting right is
00:39:58
that we are massively compute
00:40:00
constrained. Massively.
00:40:02
And those are two problems. One is the
00:40:05
power. So if you look at companies like
00:40:07
Bloom Energy, it has gone absolutely
00:40:10
straight up vertical nuclear. And the
00:40:12
reason is because Bloom has a solution
00:40:15
that allows you to useNAT gas that
00:40:18
allows you to do something on site and
00:40:21
critically allows you to get your clean
00:40:22
air permits very quickly because it has
00:40:25
very very little emissions and it's been
00:40:28
proven as such. And so instead of
00:40:30
waiting for years to get on the grid, if
00:40:32
you wanted to build a data center, you
00:40:34
can now use their services. The other
00:40:37
part that's going absolutely nuclear is
00:40:38
the actual land and the shell because
00:40:41
it's turning out it's impossible to get
00:40:43
these approvals. Now, why is that? And I
00:40:45
sent Nick an image. The reason is
00:40:48
because underneath at the core of it all
00:40:50
is there's a tide that is shifting on
00:40:53
AI. the American population is
00:40:56
incrementally
00:40:58
getting more and more negative on the
00:41:00
whole subject matter rit large
00:41:03
and it's not clear exactly why they're
00:41:07
doing that. Maybe it's the doomerism
00:41:08
which we talked about last week. Maybe
00:41:10
it's the fear of the job loss that Jason
00:41:11
has been talking about. Maybe it's just
00:41:14
this idea of yet another wave of
00:41:16
innovation that's only going to benefit
00:41:17
a few in an extreme way. minting
00:41:21
trillionaires all over the place while
00:41:23
everybody else stands still. I don't
00:41:24
exactly know what's causing it, but the
00:41:27
sentiment is shifting. And as the
00:41:29
sentiment shifts, the most scaled action
00:41:32
that they can take, they're taking,
00:41:33
which is then they are going and voting
00:41:35
down data centers. Here's an example
00:41:36
which was insane. A town approves a $6
00:41:39
billion data center build and then half
00:41:42
the the the board gets ousted. just
00:41:46
voted out overnight so that they could
00:41:48
put in new people to undo the decision.
00:41:52
So if you look at this all around the
00:41:54
country, the answer is not, oh, we're
00:41:57
only going to build in Texas. That
00:41:58
doesn't work. There's not enough power.
00:42:01
There's not enough grid capacity.
00:42:02
There's not enough net gas that allows
00:42:04
that to happen. Maine just passed a bill
00:42:08
that bans all data center buildings. So
00:42:10
I think the reason why all birds went
00:42:12
crazy
00:42:14
is a very very small canary in a very
00:42:17
important coal mine which is we are
00:42:19
absolutely compute constrained.
00:42:22
I think if you play this out the real
00:42:24
problem again goes back to anthropic and
00:42:25
open AI. If I were them it is a five
00:42:28
alarm fire for them. They more than
00:42:30
anybody else needs to get their hands on
00:42:32
compute. They need to have land power
00:42:34
shell, but otherwise that revenue could
00:42:37
either slow down or hit a wall and it
00:42:40
will not be because of product quality
00:42:42
and adoption. It will entirely be
00:42:44
because of the Fster effect. You just
00:42:46
couldn't keep the site up and I think
00:42:48
that that that would be a huge huge
00:42:50
problem for everybody.
00:42:51
>> Saxs, final thoughts on data centers and
00:42:53
the buildout.
00:42:54
>> Well, I think I think Jimoth is right
00:42:55
that the the data centers become very
00:42:57
unpopular. probably in 30 states they're
00:42:59
just going to ban them outright and then
00:43:02
it's very hard to get projects approved.
00:43:03
Look, I think there's a few reasons for
00:43:05
this. One is that there are a lot of
00:43:08
let's call them real estate developers
00:43:09
who are kind of wildcatterers who are
00:43:11
out there trying to get entitlements and
00:43:13
they did bring a lot of projects up for
00:43:16
local permits where they didn't have a a
00:43:17
power solution. And there's no question
00:43:20
that local communities do not want the
00:43:22
data centers drawing off the grid,
00:43:25
thereby increasing residential prices if
00:43:28
that data center is not bringing its own
00:43:30
power generation. And the administration
00:43:33
agrees with this. This is why the
00:43:34
president did the rateayer protection
00:43:36
pledge where we got all the major users
00:43:38
of the data centers, you know, all all
00:43:40
the hyperscalers to agree that they
00:43:43
would not build new data centers without
00:43:45
bringing their own power. So again, it
00:43:47
was designed to be power neutral to the
00:43:50
grid or in fact it would increase the
00:43:52
amount of energy available to the grid
00:43:53
because these data centers would give
00:43:56
back when they're not at peak usage. So
00:43:58
that's sort of like category number one
00:44:00
is there's this fear of electrical rates
00:44:03
going up. But there's a couple other
00:44:04
categories I think of groups. The second
00:44:06
one was future of life and a lot of
00:44:09
these like doomer groups saw that data
00:44:11
centers were a way to stop AI progress
00:44:15
and there are interviews with some of
00:44:17
these doomer folks who say things like
00:44:19
we have to meet people where they are
00:44:21
meaning that they've been unable to
00:44:23
convince people that AI is going to lead
00:44:24
to the terminator but they can convince
00:44:26
them that AI data centers are going to
00:44:29
use up their water for example which
00:44:30
isn't true and so a lot of the nimism
00:44:33
has been kind of astrourfed by a lot of
00:44:36
the doomer groups which have a lot of
00:44:38
money thanks to contributions from a few
00:44:41
tech billionaires like we've talked
00:44:43
about in the past. So that's category
00:44:44
number two. Category number three
00:44:47
ironically again is anthropic itself. It
00:44:50
has allied itself politically with a lot
00:44:53
of the doomer groups, a lot of the nimi
00:44:55
groups. It didn't seem to matter in the
00:44:57
first couple of years because Anthropic
00:44:59
had made the strategic decision not to
00:45:01
build its own data centers. So they
00:45:03
probably thought that they were just
00:45:04
throwing sand in the gears of OpenAI or
00:45:08
XAI, their competitors, and they would
00:45:10
just rely on hyperscalers to get their
00:45:13
compute. And I think that that strategy
00:45:16
has now backfired in the sense that they
00:45:18
apparently have reached the limits of
00:45:20
the compute that's available to them by
00:45:23
buying it from a third party and they
00:45:24
need to build their own data centers.
00:45:25
It's going to be very interesting to see
00:45:27
how they adapt to that and how the
00:45:30
message around data centers changes over
00:45:32
the next year as let's call it the
00:45:34
effective altruists decide that all of a
00:45:36
sudden data centers or certain kinds of
00:45:38
data centers might be a good thing
00:45:40
because they serve their mission. So I
00:45:42
think that's going to be a very
00:45:43
interesting thing to watch. I'll tell
00:45:44
you the one thing I think you're
00:45:45
missing, which is that most people in
00:45:48
America really are starting to really
00:45:51
hate rich people. And there's no
00:45:54
physical space that better represents
00:45:56
the wealth in America, the wealth
00:45:58
creation that's happened that a lot of
00:46:00
people feel left behind from than the
00:46:02
data center. What other physical space
00:46:04
is there to go to? It is the temple of
00:46:06
the wealthy. It is the mechanism, the
00:46:09
tool, the machinery of the wealthy. It
00:46:12
is the way that the rich elite tech kind
00:46:16
of political connected billionaires that
00:46:19
we're obviously all attached to are
00:46:22
taking from the poor, getting themselves
00:46:26
ahead, shooting themselves to space,
00:46:29
leaving everyone else behind. And the
00:46:31
data center, I think, is the
00:46:32
representation of their progress. And it
00:46:35
is a representation of the progress that
00:46:37
others don't feel. So that's why I think
00:46:39
it is physically like the manifestation
00:46:41
that people want to attack and destroy.
00:46:44
Uh there's very little rationale about
00:46:46
oh let's stop AI robots from killing us.
00:46:48
I think people just don't see the value
00:46:50
in AI. The average person doesn't see
00:46:52
the value in AI today yet. Like we just
00:46:54
talked about so much of the value of AI
00:46:56
is showing up in the enterprise and in
00:46:57
the rebuilding of enterprises. But for a
00:46:59
consumer's life to actually be altered
00:47:01
in a meaningfully positive way, most
00:47:03
people don't feel that yet. the best
00:47:04
thing they see is some medical advice
00:47:06
they're getting on chat GPT or something
00:47:08
and that's kind of the end of it for
00:47:10
them. So I think there's a lot of this
00:47:12
populism that's swollen and that's taken
00:47:14
over not just the US but probably a good
00:47:16
chunk of the west and and the the data
00:47:18
center is the target. It is the pier
00:47:20
deter of this space in a way. This is
00:47:23
their attack vector.
00:47:24
>> To David's point, they'll probably ban
00:47:26
data centers in 30 states. Data centers
00:47:29
can output and input at the speed of
00:47:31
light. So they can be anywhere. I mean,
00:47:32
you know, the reason you put data
00:47:33
centers in different states is cheap
00:47:35
power, low latency. So maybe you get a
00:47:37
couple milliseconds of latency. So the
00:47:39
data centers can go anywhere. So as soon
00:47:41
as all the states start banning the data
00:47:43
centers, the data centers will just go
00:47:45
to space to Iceland to Texas to Florida
00:47:49
and then and then the data center and
00:47:51
then everyone will be forced to move on
00:47:53
Travis. They'll basically find the next
00:47:55
target for populism which will be
00:47:57
something else. The peerto- tear, the
00:47:58
data center, what's next? You know, this
00:48:00
is this is part of what's going on right
00:48:02
now
00:48:03
>> where we've got this kind of level of
00:48:05
debt from the I always bring it back to
00:48:07
this because I do think this is at the
00:48:08
root cause of populism is the fact that
00:48:11
the government promised so much is so
00:48:13
inefficient and as a result it's
00:48:14
destroyed the value of the dollar not
00:48:15
given anyone anything that they thought
00:48:17
they were getting.
00:48:17
>> And we do have the rateayer pledge sachs
00:48:20
to address the energy issue that was a
00:48:22
big win I think and you worked on that I
00:48:25
believe.
00:48:25
>> Yeah.
00:48:26
>> But hold on. Can I just say something? I
00:48:28
think that rate payer pledge is
00:48:29
important. It doesn't change the
00:48:31
business model of the utility. Meaning
00:48:35
>> what the utility is allowed to do is
00:48:37
every year build a budget. And the way
00:48:40
that they decide how to charge you is
00:48:43
that they are allowed to make
00:48:44
investments, right? And then they're
00:48:47
allowed to earn 10% roughly on the
00:48:51
amount of investment that they make.
00:48:53
What do you think their incentive is?
00:48:55
their incentive is to find ways to keep
00:48:57
investing and upgrading the
00:48:59
infrastructure. So I suspect what you'll
00:49:01
see is independent of what the data
00:49:04
centers do on site, which I think is
00:49:06
smart and good, and I'm glad you guys
00:49:08
did that. It doesn't slow down the
00:49:11
actual fundamental business model of the
00:49:13
utility because those are local monopoly
00:49:16
licenses that are granted at the state
00:49:18
and county level. And so when you go
00:49:20
inside and you look at those utilities,
00:49:22
what they're allowed to do is say,
00:49:23
"Well, I'm going to bury the lines
00:49:25
underground because there's a wildfire
00:49:27
threat that's going to cost 10 billion."
00:49:29
And they present that. The PUC has to
00:49:31
say yes. And they're allowed to make 10%
00:49:32
on 10 billion. So the business model of
00:49:35
the utility has to be looked at because
00:49:37
it is independent of all of these other
00:49:39
things. And they have an incentive like
00:49:40
insurance companies to just walk prices
00:49:43
up over time.
00:49:44
>> It's true.
00:49:44
>> That's why I think behind the meter is
00:49:45
so so important there for sure. But
00:49:48
look, I mean, this idea that that data
00:49:50
centers don't create jobs is just wrong.
00:49:51
I mean, it's been a huge boon for
00:49:53
bluecollar jobs and the construction
00:49:55
industry. I mean,
00:49:56
>> while they're building the data center.
00:49:58
>> Yeah. But that that construction could
00:49:59
be a wave that goes on for a decade or
00:50:01
two decades. I mean, the capex is not
00:50:04
going down. It's increasing. Listen to,
00:50:06
for example, what Jensen says about his
00:50:07
projections. I mean, it's not like the
00:50:09
capex is a one-year thing. We're seeing
00:50:11
tens of thousands of new construction
00:50:13
jobs being created and 25 to 30% higher
00:50:16
wages for electricians, carpenters, the
00:50:19
guys who hang drywall or pour concrete,
00:50:21
create roads, install equipment. I mean,
00:50:24
these are blue collar jobs and it's
00:50:26
creating again not just new jobs, but
00:50:28
also wage increases. It's a good thing.
00:50:31
It's not a bad thing. I think the issue
00:50:33
is
00:50:33
>> it's not a bad thing, but it's not the
00:50:35
same as a fab where the jobs are
00:50:37
permanent and they're there. Uh we can
00:50:39
debate it all we want but you can
00:50:41
there's about a hundred data centers
00:50:42
right now that are being contested. The
00:50:44
data is for every hundred that are
00:50:47
contested about 40 get cancelled. That
00:50:49
number is increasing for this year. It's
00:50:51
more than doubled already from what the
00:50:52
number was last year. And the total
00:50:55
economic value of those 100 data centers
00:50:57
right now is about $162 billion.
00:50:59
>> Jason, the the thing is is that there
00:51:01
are permanent jobs because what where's
00:51:04
the energy that's powering that? Where's
00:51:06
the semiconductors that's powering that?
00:51:08
You know, there's a lot of
00:51:11
>> we we actually had um Chase from Crusoe
00:51:14
and Michael from Cororeweave on the
00:51:16
All-In interview show and they said
00:51:18
they're bringing energy with them.
00:51:20
That's their big thing. Bring BYO, bring
00:51:22
your own energy to the space. They're
00:51:24
bringing in Knackas. They're bringing in
00:51:26
diesel fuel. They're bringing in solar.
00:51:29
BYOE is the model. Well, c can I also
00:51:31
say listen if if you were of the opinion
00:51:34
that you didn't want more data centers
00:51:36
in America because it was using up say
00:51:38
scarce energy resources or something
00:51:40
like that but you weren't just
00:51:42
anti-progress altogether then what you
00:51:45
would want to do is at least see more
00:51:47
data center construction among US allies
00:51:49
right energyrich US allies well a year
00:51:53
ago that's one of the things I worked on
00:51:54
was allowing the Gulf states to build
00:51:56
data centers with American technology
00:51:59
for American companies and remember the
00:52:02
controversy that caused. Everyone
00:52:03
accused us that somehow this was like
00:52:05
serving China. Well, those data centers
00:52:07
are getting bombed and I don't think
00:52:08
they'd be getting bombed right now by
00:52:10
Iran if they were serving China. These
00:52:13
were
00:52:13
>> they're very strategic. Yeah.
00:52:15
>> The IRGC has put these data centers on a
00:52:17
list of assets that they've threatened
00:52:19
to destroy because they're American
00:52:22
assets or they're assets of our Gulf
00:52:24
State partners in partnership with
00:52:27
America and American companies. This
00:52:29
whole idea that somehow this was a
00:52:32
threat to American national security was
00:52:33
a total hoax just like the data center
00:52:36
uses too much water. And what I'm saying
00:52:38
is that a lot of the same forces were
00:52:40
behind this hoax as are behind these
00:52:42
data center hoaxes. Now Freeberg is
00:52:44
right about the resentments, but those
00:52:46
resentments get whipped up and marshaled
00:52:49
by people who have an agenda and that
00:52:52
agenda is wellunded and strategic. And I
00:52:55
can tell you then the case of the GCC
00:52:57
data centers, Anthropic was adamantly
00:53:00
opposed to that and they were lobbying
00:53:02
against it and they were again salting
00:53:04
the earth against those projects. I
00:53:06
think the whole issue is kind of moot
00:53:07
now anyway because I think the data
00:53:08
centers have been blown up but
00:53:10
>> uh or they've been threatened to but it
00:53:12
just shows how ridiculous
00:53:14
>> and and what a scop some of the
00:53:16
opposition to these ideas are. and it's
00:53:20
a cell phone in many ways. If you think
00:53:22
about who are the top two spokespersons
00:53:24
for our industry and what they're
00:53:27
communicating to Americans versus say
00:53:30
how the Chinese view AI, which is
00:53:32
incredibly positive, we've got Daario
00:53:34
who says it's the end of days.
00:53:36
Everything's going to be hacked. All
00:53:38
your files are going to be hacked. All
00:53:40
your accounts are going to be hacked.
00:53:41
And then on the other and everybody's
00:53:43
losing their job. That's our top
00:53:45
spokesperson in Dario at Anthropic. The
00:53:47
other top spokesperson just had a
00:53:49
70,000word piece written about him where
00:53:52
Ronan Farah said, "I had a dozen people
00:53:55
tell me unprompted, Sam Wolman's a
00:53:57
sociopath." Those are the top two
00:53:59
spokespeople. Those people cannot be the
00:54:02
spokespeople for this industry. And this
00:54:04
industry's got to get focused on fixing
00:54:05
the big three. Healthcare is going to be
00:54:08
dramatically improved by AI. Housing
00:54:10
could be dramatically improved. I'm not
00:54:12
sure who's working on that. And
00:54:14
obviously education and the cost of
00:54:15
education. We need somebody out there, a
00:54:18
Michael Dell, you know, Jensen, Elon,
00:54:21
just explaining how world positive this
00:54:23
could be because right now AI is as
00:54:26
popular as is less popular than ICE, the
00:54:29
Democrats, and the government of Iran.
00:54:32
>> Joe Leand at Alpha School. On the
00:54:34
education side, I think you just have to
00:54:35
look at Alpha School. It's it's working.
00:54:37
>> All right. So, that's your story, uh,
00:54:40
Chimath. That's the story of all birds,
00:54:43
the most overvalued startup. But uh
00:54:45
we're going to play a little uh cue the
00:54:47
music here. We're going to play a little
00:54:48
game show for everybody. This is the
00:54:50
price uh David Saxs is wrong. Yes, this
00:54:53
is the price is wrong. The price of the
00:54:56
startup Travis is wrong.
00:55:00
>> And this is the game show where we guess
00:55:03
which overvalued disaster are we talking
00:55:05
about. First up, hell from Austin,
00:55:07
Texas, an enforcer in the PayPal mafia.
00:55:10
It's Mr. David Saxs. Welcome to the
00:55:11
program, David Saxs. Are you ready to
00:55:13
play the price?
00:55:14
>> I'm ready to play. Let's do it.
00:55:15
>> You're ready to play? Okay.
00:55:16
>> All right.
00:55:17
>> Here we go. The price is wrong. This
00:55:19
startup was once valued at over $13
00:55:21
billion. Their main business model
00:55:23
selling JPEGs for fake internet money.
00:55:26
People lost their minds and started
00:55:28
spending millions of dollars on monkey
00:55:30
images. Play the thinking music, please.
00:55:33
David Sachs, can you name that start at
00:55:37
one? Billion. 13 billion for is that is
00:55:41
that like the board apes thing?
00:55:43
>> Okay, you're closing in. Remember, form
00:55:45
it in the name of a question.
00:55:47
>> I know. I know. I know.
00:55:48
>> It's selling JPEGs for fake internet
00:55:50
money.
00:55:50
>> Me, me, me. Pick me. Pick me. Pick me.
00:55:52
>> People, you'll get your chance. Chim.
00:55:55
>> Oh, I think I know what you're talking
00:55:56
about.
00:55:58
I can't remember.
00:55:59
>> They're spending millions of dollars.
00:56:01
Can I say Can I say it's your turn to
00:56:04
steal the steel?
00:56:05
>> Open.
00:56:06
>> Correct. 100 points for Chimoth
00:56:08
Polyhapatia.
00:56:10
>> Okay. An amazing steal. An amazing
00:56:12
steal. And next up,
00:56:14
>> okay, here we go. He puts the dick in
00:56:16
dictator hailing from Palo Alto,
00:56:18
California. It's your favorite Door
00:56:21
Dasher Jamal Poly Hot. How long you been
00:56:23
door dashing there, Jamal?
00:56:24
>> Nine years.
00:56:25
>> Nine years. You like that? What do you
00:56:27
like about that? Is it the interaction
00:56:28
with the people? Is it the tips? Is it
00:56:30
stealing a couple of French fries? What
00:56:31
do you like about being a door dasher
00:56:32
there?
00:56:33
>> I typically take four fries. I lick the
00:56:35
burrito.
00:56:36
>> Okay. Licking burritos. Did he just say
00:56:38
licking burritos?
00:56:39
>> This is why we need robot delivery.
00:56:41
Robot delivery drivers. I mean, it's
00:56:43
just
00:56:44
>> coming.
00:56:44
>> It's coming. Okay.
00:56:46
>> We don't want our French fries being
00:56:47
licked.
00:56:47
>> Let's get the music going here. This
00:56:49
startup once valued at 4 billion.
00:56:52
>> Allowed you to talk to other people on
00:56:55
your mobile device during co and you
00:56:57
could listen to mid VC.
00:56:59
>> Yes. Yes. I got it. I got it. It's
00:57:00
called
00:57:00
>> generic startup device. Can you name
00:57:02
that overvalued startup?
00:57:03
>> Oh my god. It's called Clubhouse.
00:57:05
>> Oh my god. He's on a heater. That's 200
00:57:07
points for Chim Polyatilla. Hey, that's
00:57:10
a lot of Door Dashes. He's going to do
00:57:12
very well here in the final round.
00:57:13
>> That's an amazing memory. How How soon
00:57:15
we forget? I can't I couldn't replace
00:57:17
any of these names.
00:57:18
>> Okay, finally streaming in from a potato
00:57:21
field at an undisclosed location in
00:57:23
Idaho. It's It's Dr. David
00:57:25
Freriededberg. Uh you're you're in the
00:57:27
potato sciences, correct,
00:57:29
>> Mr. Freeberg? That's right. Potato.
00:57:31
>> Okay. You like the potatoes? You have a
00:57:33
favorite potato dish? Is it the
00:57:34
scallops? What do you like to do with
00:57:35
your potatoes?
00:57:36
>> What is your name, Mr. What is the
00:57:37
host's name? What is your name, host?
00:57:39
>> It is me, Jocular Jal. Let's go. Here we
00:57:42
go. It's Jocular Jal here at the helm.
00:57:45
>> You like a certain type of potatoes
00:57:47
there?
00:57:49
>> It's calacanis. And what do you like
00:57:51
there on the potatoes? What's your
00:57:52
favorite there? You like a creamy
00:57:54
mashen? Little little little little
00:57:56
cheesy.
00:57:56
>> You're liking You like the cheese? Okay,
00:57:58
very good. Okay, David, this startup was
00:58:00
once valued at $270 million. This is
00:58:03
your favorite. I understand you're a
00:58:04
vegan. They sold juice. These juice
00:58:07
packages went into a juice machine. I
00:58:09
know. I know. I know. I know. I know. I
00:58:10
know. I know. I know. Branded as the
00:58:12
next iPhone. The iPhone of juice. Can
00:58:14
you name
00:58:15
>> Oh, I know it.
00:58:15
>> I know it.
00:58:18
>> I know it. I know it. Pick me.
00:58:19
>> You'll have a chance to zero. Juice.
00:58:21
That's 100 points.
00:58:23
>> Wait, are you searching? It looks like
00:58:25
he's searching.
00:58:25
>> Did he? Oh my god. Are you on your
00:58:27
device?
00:58:28
>> He looked like he was on his device.
00:58:29
>> Internet free. Internet free here.
00:58:33
All right, folks. There you have it. A
00:58:34
clear winner. Chamal Poly Hapatia. Tell
00:58:36
him what he's won. He's won a trip to
00:58:40
Temptations 2 in Cabo. You're going
00:58:42
directly to Temptations to an adult
00:58:44
resort. Uh that's for you and two of
00:58:46
your friends. Okay. Enjoy Temptations 2.
00:58:48
Shalom. Polyapatia.
00:58:50
>> Two of your friends. What a fun game
00:58:52
this is.
00:58:53
>> You can bring your throppple.
00:58:54
>> I think this is going to need to become
00:58:56
a regular feature of the pod.
00:58:58
>> Okay, landed. Everyone Everyone forgets
00:59:01
the unicorns that don't work. They just
00:59:03
like disappear into the ether.
00:59:05
>> All right. All right. There you go. You
00:59:07
won a pier tear on a uh on a
00:59:12
>> tax bill in New York.
00:59:13
>> You got a pier tear available to you as
00:59:16
your first place winner. Chimoth Poly
00:59:18
Hapia Door Dasher for Ela Pear. All
00:59:22
right. Listen. Uh there's been a lot of
00:59:24
shenanigans going down in DC, the most
00:59:26
boring city in the world, where
00:59:27
apparently everybody's uh getting a
00:59:30
little frisky after hours. TMZ launched
00:59:33
a news bureau.
00:59:35
Eric Slaw is out of the governor's race.
00:59:38
There's a lot of dark stuff that's been
00:59:41
released. He is innocent until proven
00:59:44
guilty, but it's not looking good.
00:59:46
>> He also resigned from Congress, Jal.
00:59:49
>> And he resigned from Congress as well.
00:59:51
And Freedberg
00:59:53
are investigative journalist now working
00:59:56
as a stringer for TMZ Freeberg. What
00:59:59
have you learned? What's in Freeberg's
01:00:02
anything? I my um
01:00:04
>> Okay, you have an analysis then? Maybe
01:00:06
>> my anecdote on this is back in December
01:00:11
when
01:00:13
it was first rumored that Swallwell was
01:00:15
going to run for governor. I started
01:00:19
making some calls to various folks to be
01:00:22
like, "Hey, you know, what do we think
01:00:25
of this guy? Is he going to be a good
01:00:27
candidate?" You know, obviously I and a
01:00:29
lot of other people before they evacuate
01:00:32
the state care a lot about the future of
01:00:34
California. So, I started checking
01:00:36
around. I spoke to several people who
01:00:39
independently told me that there's
01:00:42
knowledge about this guy sending, you
01:00:44
know, pics to employees and that this
01:00:47
guy has a bunch of stuff that's going to
01:00:50
come out about him. So, I heard all of
01:00:53
this not from one person, but from
01:00:54
several different sources. This was back
01:00:56
in December going into January. And I
01:01:00
largely kind of dismissed it cuz I was
01:01:01
like, if this is true, this would have
01:01:03
all come out already. I'm like, there's
01:01:04
no way this is true. If it was true,
01:01:06
like people would have talked about it.
01:01:07
They would have made a thing about it.
01:01:09
If multiple people are telling me about
01:01:10
this, then I've got to assume that it's
01:01:14
a rumor that's being used to block him
01:01:16
from running for governor versus it
01:01:18
being a real thing because multiple
01:01:20
people have this knowledge and this
01:01:21
information. So, this was December,
01:01:23
January where I had these conversations
01:01:25
and at that point, nothing had come out.
01:01:27
So, I was like, "Okay, it doesn't seem
01:01:28
like this is real." And then everything
01:01:30
that I had been told started to come out
01:01:33
in the last week. So the striking aspect
01:01:35
of all of this for me was how much
01:01:37
knowledge there was about these various
01:01:40
incidents with the guy. How so many
01:01:43
people had this knowledge and how no one
01:01:45
had actually brought the knowledge to
01:01:47
bear. Which begs the question,
01:01:52
why did they not do what was right by
01:01:54
the victims? Or why did the victims sit
01:01:58
on the sidelines waiting for the right
01:02:00
moment to all come out together? because
01:02:03
this was broad knowledge within a
01:02:05
community of people and they made the
01:02:07
choice not to bring it forward with that
01:02:10
knowledge. And that's what was so
01:02:11
striking to me about this whole thing. I
01:02:13
had honestly dismissed the whole thing
01:02:14
as just being rumoring to try and
01:02:16
bismerch the guy. And it turned out that
01:02:18
these were all being held back
01:02:20
purposefully and deliberately for a very
01:02:22
particular moment in time when they were
01:02:24
all brought forward to be used.
01:02:26
>> And let me just be clear, these are all
01:02:28
allegations. Nothing's been proven in
01:02:29
court. He does get his day in court.
01:02:33
etc. We just want to make sure there's a
01:02:35
bunch of alleged here.
01:02:36
>> Yeah. No matter how bad and all that I'm
01:02:39
saying is that people had told me about
01:02:41
these supposed claims 5 months ago, four
01:02:45
months ago, multiple people, and had
01:02:46
chosen not to bring it forward. And that
01:02:48
the victims had not come forward
01:02:49
publicly with these claims. And then
01:02:51
there was this coordinated effort to
01:02:53
bring everything forward at the same
01:02:54
moment. And that's what was so striking
01:02:56
to me is just how coordinated all of
01:02:58
this happened.
01:02:58
>> Who's controlling it? Who do you think
01:03:00
like is there like a meeting? Is there
01:03:02
like a a counseling
01:03:04
>> Nancy Pelosi?
01:03:06
>> Jesus Christ. It's so
01:03:07
>> I don't know if that's true.
01:03:08
>> Well, I don't not just her.
01:03:10
>> I'll tell you my sense of it. My sense
01:03:12
of it is that there are certain insiders
01:03:13
and he's not an insider. And those
01:03:16
insiders are like the Katie Porters of
01:03:18
the world. Katie Porter, I think, is who
01:03:20
the Democratic establishment wants to be
01:03:22
governor. She's an insider to national
01:03:26
Democrats. She's an insider to
01:03:28
California Democrats. And I think that
01:03:31
she's the preferred candidate. Tommy
01:03:33
Styer, she's a spousal abuser.
01:03:35
>> I don't know that Katie Porter is like
01:03:37
this ultimate insider, but I think there
01:03:38
are, look, there are clearly insiders.
01:03:41
The Democratic Party is a machine that
01:03:43
exists to siphon off as much money as
01:03:45
possible from the public till the
01:03:47
interests that support the party and
01:03:50
it's their gravy train. And they're not
01:03:52
going to let anyone stop.
01:03:53
>> The Dems Sachs, I don't know if it's
01:03:54
just the Dem.
01:03:55
>> Hold on a second. They're not going to
01:03:56
let that gravy train stop for one
01:03:57
second. Now, the Democratic party had a
01:03:59
huge problem in this California
01:04:01
governor's race, which is that the
01:04:03
Democratic field was very fragmented.
01:04:05
And so, the two Republican candidates
01:04:07
actually were polling the highest. And
01:04:09
so, that just so the viewers have
01:04:11
context, California has this weird
01:04:13
jungle primary system where the top two
01:04:15
go to a runoff. They don't have a
01:04:17
Democrat lane and a Republican lane.
01:04:19
They just take the top two jungle
01:04:20
primary and then they go to the runoff.
01:04:22
Even today still Hilton and was it
01:04:25
Bianca or whatever they are polling at
01:04:28
both around like 14 or 15% and if the
01:04:30
election were held today you'd have two
01:04:32
Republicans in the runoff. So the
01:04:33
Democrats needed to winnow the field
01:04:36
down and have fewer candidates. In
01:04:39
addition to that, they must have been
01:04:41
concerned that all this oppo on sua
01:04:44
would come out once it was him versus
01:04:47
say Steve Hilton and they didn't want it
01:04:50
to come out later when they could lose
01:04:51
the election. So the powers that be made
01:04:54
the decision to lance the boil probably
01:04:57
there was a conversation with him to
01:04:59
tell him to get out of the race. He
01:05:00
didn't listen. And by the way, remember
01:05:02
this feels a lot like what happened with
01:05:04
Joe Biden when he had to drop out of the
01:05:06
presidential race. That sounds
01:05:08
surprisingly like that.
01:05:09
>> The whole Democratic establishment and
01:05:11
all the mainstream media were saying
01:05:12
that Biden was sharp as attack. Okay.
01:05:15
And then he had that disaster debate
01:05:17
performance with Trump and it became
01:05:20
clear. Yeah. And you could just see the
01:05:23
text messages were flying during the
01:05:24
debate between the Democratic Party
01:05:26
insiders. And by the time that debate
01:05:28
was over, they had congealed on a new
01:05:31
position, which is that Biden had to
01:05:32
step aside. And then Nancy Pelosi was
01:05:35
reported as having gone to the president
01:05:36
and said, "We can do things the hard way
01:05:39
or the easy way." Imagine that. Like
01:05:42
telling the president of the United
01:05:43
States, "We can do things the hard way
01:05:45
or the easy way." And then Biden
01:05:47
disappeared for a week and magically he
01:05:50
stepped aside by tweets. Remember that
01:05:53
he published a statement that appeared
01:05:55
to be done by an autopin. People were
01:05:58
speculating whether he was even behind
01:05:59
this or whether the staff pushed him to
01:06:00
do it. The whole thing was extremely
01:06:02
weird. But he, you know, was clearly
01:06:05
muscled out of it. Just a few days
01:06:07
before he had said, "I'm not leaving the
01:06:09
race no matter what." I mean, it was
01:06:11
like straight out of that Wolf of Wall
01:06:12
Street meme. They're not getting me out
01:06:15
of here. And then a few days later, he's
01:06:17
resigning by tweet. And again, Nancy
01:06:20
Pelosi appears to be the figure at the
01:06:21
center of both these things. She was
01:06:24
hold
01:06:25
on a second. Pelosi is reported as
01:06:27
having been Swallwell's mentor. I guess
01:06:30
she found him roughly 20 years ago to
01:06:33
run for Congress in the first place when
01:06:35
the Republicans wanted to kick Swell off
01:06:38
of the intelligence committee, the House
01:06:40
Intelligence Committee for allegedly
01:06:42
being involved with that Chinese spy
01:06:44
Fangfang. It was Pelosi who protected
01:06:48
him. So, she's sort of been a central
01:06:50
figure in his career. I'm not saying she
01:06:54
approved of anything he did. But
01:06:55
>> listen, politics is
01:06:56
>> But there's no there's no way that that
01:06:59
button gets pushed without going to
01:07:01
Pelosi for the sign off, right? I mean,
01:07:04
she's like the boss of this operation.
01:07:06
And I think that the same thing happened
01:07:08
to Swallow that happened to Biden is
01:07:10
they went to him and said, "We can do
01:07:11
things the hard way, the easy way." He
01:07:13
was too dumb to listen and they did
01:07:15
things the hard way.
01:07:16
>> Yeah. Shout out Nancy Pelosi, incredible
01:07:19
day trader. And uh shout out to a friend
01:07:22
of the pod Ro uh Row who's now beaten
01:07:25
Nancy Pelosi. I think his trades this
01:07:26
year he's uh he's actually even beaten
01:07:29
>> Nancy Pelosi. That was the Roana has
01:07:33
traded $600 million of stock.
01:07:35
>> He trades more frequently than Citadel
01:07:37
Securities.
01:07:38
>> I mean it's incredible. Yeah. Where's
01:07:40
his piter?
01:07:41
>> I want to get in on it.
01:07:42
>> The congressman you've been supporting
01:07:44
for years is a great stock trader. He's
01:07:47
apparently we should have been talking
01:07:48
to him not about taxes. We should have
01:07:50
been talking to him about trades.
01:07:52
>> There were things I liked about Roana. I
01:07:54
mean he he did support freedom of speech
01:07:56
with the whole Twitter thing.
01:07:58
>> You know what it was is that the Overton
01:08:00
window shifted so much.
01:08:01
>> He supported him.
01:08:02
>> Yeah. When I supported Roana, freedom of
01:08:04
speech was a big issue and he was one of
01:08:06
the only Democrats to support that. He
01:08:08
was also, I think, the only member of
01:08:10
the progressive caucus to support a
01:08:11
diplomatic track for Ukraine, which I
01:08:13
supported and I gave him credit for
01:08:15
that. And yeah, I knew that he was in
01:08:17
favor of a wealth tax, but I thought
01:08:18
that was just a very unserious proposal
01:08:20
that, you know, it wasn't in play in any
01:08:22
way. Well, the Overton window shifted so
01:08:25
much that now the wealth tax really is a
01:08:27
possibility. So, you know, things have
01:08:29
changed.
01:08:30
>> Here it is. Shout out to our boy Ro
01:08:33
Kana. Look at this. Right about now,
01:08:35
Chimat's thinking about making a
01:08:36
managing director offer here. I mean,
01:08:38
Ro, if it doesn't if you get booted out
01:08:40
of office, you could become a managing
01:08:42
director
01:08:43
>> at Alt Catholic. Statistically, like
01:08:45
even if you had insider information,
01:08:47
making that much money is like very
01:08:49
hard.
01:08:50
>> Holy cow. Nancy Pelosi is going to
01:08:53
shiver Roana next for not giving her the
01:08:55
uh giving her the inside track on
01:08:58
whatever he's betting on. What did he
01:09:00
bet on?
01:09:00
>> Well, look, any any, as we all know, any
01:09:02
investor can have a good quarter or a
01:09:04
year. But to put up the kind of returns
01:09:06
that Nancy Pelosi has done over decades
01:09:08
is nothing short of miraculous.
01:09:10
>> It's generational. It's a generational
01:09:12
run. Give her her flowers.
01:09:14
>> She's substantially better than Warren
01:09:16
Buffett.
01:09:19
>> I mean, Warren Buffett, Stan, Dr.
01:09:21
Miller, and Nancy Pelosi are three of
01:09:24
the most accomplished investors of all
01:09:26
time.
01:09:27
>> Whatever you did to the poor Buffett
01:09:29
fans, Chimath, they are incredibly angry
01:09:33
at you for desecrating his legacy.
01:09:37
>> I didn't desecrate his legacy. I pointed
01:09:38
out one unavoidable fact which is his
01:09:42
returns are biodally distributed pre and
01:09:44
postreg.
01:09:46
>> When you have to follow the rules of
01:09:47
disclosure, everybody's returns got
01:09:50
kneecapped. When there was no disclosure
01:09:52
rules,
01:09:54
his returns were off the charts. That's
01:09:56
just a mathematical truism. Now, what's
01:09:58
interesting to note is the reason why
01:10:00
Nancy Pelosy's returns are so
01:10:01
consistently good is reggga FD does not
01:10:03
apply to people in Congress.
01:10:06
That should be the takeaway. They can
01:10:08
learn things.
01:10:09
>> We got to stop them.
01:10:10
>> They can learn things in their committee
01:10:12
meetings. In fact, there are situations
01:10:14
where things are disclosed and then they
01:10:17
are trading in real time.
01:10:18
>> Shout out to the skiff. Yeah, get out of
01:10:20
the skiff quick and get that trade in.
01:10:22
>> So, I have enormous respect for Warren
01:10:24
Buffett and what he's done. It's it's
01:10:25
he's he's the goat of goats,
01:10:27
>> but the returns postreg are they are
01:10:29
just materially worse than they were
01:10:31
preregg. And that's
01:10:32
>> what's he going to do with that cash
01:10:33
position? And Jamat speaking of Buffett
01:10:35
they
01:10:36
>> the market the market
01:10:37
>> 300 billion sitting there
01:10:38
>> the market is in a very complicated
01:10:40
moment right now if you look at
01:10:42
historical indicators of value so if you
01:10:45
look at Schiller as an indication of
01:10:47
value it's peaking
01:10:50
if you look at the Buffett index it's
01:10:53
peaking so there are things that when
01:10:56
you look at it look like all-time highs
01:10:59
and the problem with that is you would
01:11:01
say oh man But there's this weird
01:11:04
dispersion happening in the market.
01:11:06
Dispersion means literally a few
01:11:08
companies are hitting all-time highs. I
01:11:10
think it's like eight or nine and
01:11:12
everybody else is not. So, it's a really
01:11:15
complicated moment. It's hard to
01:11:17
understand what's going on, but he's got
01:11:18
a lot of cash. If he's sticking to his
01:11:20
knitting, he's looking at the chiller
01:11:22
index and he's looking at his own
01:11:24
indicator which shows all-time highs and
01:11:26
he's waiting for a correction.
01:11:27
>> He's not really in charge anymore. Like
01:11:29
Berkshire Hathaway is not Warren Buffett
01:11:30
anymore. Even though he's a big owner,
01:11:32
he's not really doing it.
01:11:34
>> But I mean, the fact that they're
01:11:35
sitting on that massive pile of cash
01:11:37
says something that they're not putting
01:11:38
it to work in the market says they don't
01:11:41
see an opportunity yet. Saxs, Travis,
01:11:43
I'm curious your takes. Uh either one of
01:11:45
you can go in whichever order you want
01:11:47
on how is the market crushing it while
01:11:51
we're in week seven of a war and we put
01:11:54
a hundred billion dollars or something
01:11:56
into this military activity in Iran and
01:12:01
the market is pricing it in shrugging it
01:12:03
off and we're hitting all-time high sex.
01:12:05
I don't know if you're feel first.
01:12:07
>> Yeah. I mean, look, I think it's pretty
01:12:09
straightforward, which is that in the
01:12:12
wake of the meeting in Islamabad that
01:12:14
the market is feeling confident and
01:12:17
pricing in that the war is going to get
01:12:21
resolved. The president also recently
01:12:24
said that it's very close to being
01:12:25
wrapped up. The military objectives are
01:12:28
close to being achieved and he's made it
01:12:31
sound like it's going to be resolved.
01:12:34
Yes. a deal was not signed in Islamabad.
01:12:36
I always thought that was an unrealistic
01:12:38
expectation that these two countries
01:12:40
which are at war with each other and in
01:12:41
fact have had hostile relations for
01:12:43
almost 50 years are going to resolve all
01:12:46
their differences in 24 hours is not
01:12:47
realistic. But the impression that the
01:12:50
market I think has and clearly is
01:12:52
trading on this is that that war is
01:12:54
going to be what Donald Trump said which
01:12:55
is an excursion and something that is on
01:12:59
its way to being resolved that they've
01:13:01
made progress. And Jal, you're right.
01:13:04
The whole week has just been in
01:13:05
incredibly strong. I think by Tuesday,
01:13:08
the market had recovered all of its
01:13:11
losses since the start of the war. I
01:13:13
think it made a new high yesterday on
01:13:15
Wednesday and is making new highs, fresh
01:13:17
highs today on Thursday. So, you just
01:13:20
have to say that at the present time,
01:13:23
the market thinks, and I I would
01:13:24
consider the stock market to be the
01:13:26
ultimate prediction market, that this
01:13:28
war is on its way to being resolved.
01:13:31
Just to qualify, I'm not speaking as a
01:13:33
member of the administration. I'm not
01:13:35
saying that I know something. Okay? I'm
01:13:38
just interpreting what the market is.
01:13:40
>> No clipping.
01:13:42
>> I'm not representing anyone and I don't
01:13:45
know anything different than what any of
01:13:47
you know. I'm just saying that I think
01:13:50
this is what the market is clearly
01:13:52
pricing in and I'm paying attention to
01:13:54
what the statements are of the president
01:13:57
and vice president. And this Travis, you
01:13:59
and I were talking the other day while
01:14:01
you were
01:14:01
>> Yeah.
01:14:02
>> slaughtering me in back gammon when you
01:14:04
won our eight-point match. You're almost
01:14:06
caught up
01:14:08
>> about the taco trade, not Trump always
01:14:11
chickens out. You have a theory about
01:14:13
Trump always cares about optics. Unpack
01:14:15
it for us, Travis.
01:14:17
>> Yeah. So while I was beating you back
01:14:19
Ammon
01:14:21
>> I sort of had a you know there's like
01:14:23
the in because Jay Kal asked me and the
01:14:26
insight is is like and I think this
01:14:28
maybe is just the more simple view of
01:14:30
things which is Trump's weather vein is
01:14:34
the stock market like we see VA is up we
01:14:37
see the VIX is really high but you know
01:14:39
what the S&P is trading in this band
01:14:42
that's actually fairly tight given the
01:14:44
crazy [ __ ] that's going down and
01:14:46
everybody's nervous nervous, but
01:14:48
actually he he moves in the policy
01:14:51
space. He's he's he he does not let the
01:14:54
S&P go down too low. People also get the
01:14:57
sort of the panicking thing now, which
01:14:59
is like he gets people nervous. He makes
01:15:02
moves and then he comes back to sort of
01:15:04
reality and gets things done is
01:15:06
practical and and probably the the
01:15:09
better parts of what he he might be, you
01:15:11
know, what he does well and people are
01:15:13
pricing that in. So there's there's all
01:15:15
the things Sax is saying about the
01:15:17
nuances and the details, but the sort of
01:15:20
super highlevel like I'm in low earth
01:15:23
orbit view is like the stock market is
01:15:27
Trump's weather vein and he will go to
01:15:30
the place that makes the stock market
01:15:32
come up and if he's go and and if it's
01:15:35
up and it's too high, he almost feels
01:15:37
like it's too easy. So he makes it hard
01:15:38
on himself again and then he brings it
01:15:41
back up and the traitors are getting
01:15:42
used to it I think.
01:15:44
>> Yeah. It's like playing Chimoth with
01:15:46
Alan Keading. It's just like how is this
01:15:48
guy solvent? He's literally playing
01:15:50
every hand of poker. He's losing tons of
01:15:52
money and then all of a sudden by the
01:15:54
end of the night he hits two big pots
01:15:56
and he's got the nuts both times and he
01:15:57
pulls out of the stall. What's your take
01:15:59
on the market today?
01:16:03
Yeah, I mean I'll just do it again, but
01:16:05
I think the the Schiller PE Nick I sent
01:16:07
it to you shows near all-time highs.
01:16:10
Then the second is the Buffett index,
01:16:11
which is the sum of all uh US equities
01:16:14
divided by GDP is also at all-time
01:16:16
highs.
01:16:18
So this would generally mean that you
01:16:21
need to be increasingly a little bit
01:16:23
more riskoff. But then the opposite side
01:16:25
of that and I sent you a third one and
01:16:27
this is why it's so confounding is you
01:16:28
have signals showing everything which is
01:16:30
which typically doesn't happen and this
01:16:33
is this dispersion point where when you
01:16:36
see this performance and it's up 5%
01:16:40
in the first you know half of April
01:16:42
typically the market is up almost 32% on
01:16:44
average for the rest of the year and we
01:16:46
were already up as Sax said you know 7
01:16:48
and a half% already
01:16:50
>> trying to read this graph this is crazy
01:16:52
I don't even know what's going on This
01:16:54
is just the this is just a dispersion.
01:16:55
But the the the idea of all of this is I
01:16:59
think we're in a moment where you can
01:17:00
find a piece of data to underwrite your
01:17:04
bias. And I think that's where there's a
01:17:07
lot of danger. I don't know for me
01:17:08
personally I'm generally more risk off
01:17:11
right now and more importantly I'm
01:17:13
waiting for these IPOs so that I can to
01:17:16
be very honest with you delever and get
01:17:18
some chips off the table. I think that
01:17:21
it is crucial that the SpaceX IPO get
01:17:23
done ASAP and then I think it's even
01:17:27
more crucial that one of anthropic and
01:17:30
OpenAI frontr run the other one and get
01:17:33
out first
01:17:34
>> and the first two charts are the reasons
01:17:36
why I it's fairly obvious to me what's
01:17:39
going on here. You you have uh these
01:17:42
multiple trends going on at the same
01:17:43
time. The reason people and traders are
01:17:45
valuing this and I think there is some
01:17:46
smart money in the market right now is
01:17:49
what you're seeing in terms of the
01:17:51
earnings potential of these companies as
01:17:53
they deploy AI as they have unlimited
01:17:56
intelligence and the top employees at
01:17:58
the top companies become 10 or 20 or 30
01:18:01
times more productive that's never been
01:18:03
seen before. Microsoft Office may have
01:18:05
made you 30% more efficient. The
01:18:07
internet may have made you 50% more
01:18:09
efficient but none of these things made
01:18:10
you 10 times. And you had a really
01:18:12
interesting point that you slipped in
01:18:14
earlier, Chimoth, which was nobody knows
01:18:16
how to harness these things yet, right?
01:18:17
And it's producing slop. The truth is
01:18:20
10% of people do know how to harness it.
01:18:22
20% maybe. And I watch this in my own
01:18:24
organizations and I watch it in 600
01:18:26
portfolio companies. The ones that do
01:18:27
deploy it correctly, they are running
01:18:30
the table on the ones that are not. So
01:18:32
the efficiency boom that we're going to
01:18:34
see at these companies, whether it's
01:18:36
Meta or Uber or Airbnb, whoever executes
01:18:39
it properly, is going to be phenomenal.
01:18:42
The earnings will be insane.
01:18:43
>> Maybe I'm an idiot, but it has not
01:18:46
translated into a tsunami of more
01:18:49
revenue and more profit for me yet.
01:18:51
Maybe I'm the only one and maybe it's
01:18:53
everybody else but me, but I haven't
01:18:55
seen it.
01:18:57
>> I see it in a lot of companies. So I
01:18:58
think in private companies we have two
01:19:00
companies as just two examples. Micro
01:19:02
one, which is doing data, you know, dark
01:19:05
pools of data for these language models.
01:19:07
And they have built technology that
01:19:09
allows them to build data and allows
01:19:11
them to collect data to help the large
01:19:13
language model companies grow. And they
01:19:14
are on a tear by using this to identify
01:19:17
and find great people who can then
01:19:19
contribute to these corpuses. And then
01:19:21
we have one tax GPT that is making the
01:19:23
accountants, I think they have six or 7%
01:19:25
of all accountants using their platform.
01:19:27
Jason, you agree? They're just ripping
01:19:30
accountants. I hope you can agree with
01:19:31
me with the following statement.
01:19:34
Small companies nibbling at the edges is
01:19:36
not where these guys will build a multi-
01:19:39
trillion dollar market cap.
01:19:40
>> I think we're we're both in alignment.
01:19:42
Big companies, dumb companies, slow to
01:19:44
implement the technology. Startups. No,
01:19:46
they aren't implementing new technology.
01:19:49
They're always the lagards when it comes
01:19:50
to implementing new technology.
01:19:52
>> The reason that they are slow is not
01:19:53
because they are dumb. The reason that
01:19:55
they're slow is their business is much
01:19:56
more sophisticated and much more
01:19:58
complicated than many other businesses.
01:20:00
And what I'm saying is if you can't
01:20:02
prove that this works in the big time,
01:20:05
prime time, big league use cases, it's a
01:20:09
toy. All right, Saxs, where where do you
01:20:13
stand? Are you in the Jal position? The
01:20:14
startups are showing the way. They're
01:20:15
being massively efficient. They're
01:20:16
growing revenues like we've never seen
01:20:18
in the startup community before with
01:20:20
less people. Or you in the Chimoth camp,
01:20:23
hey, big companies are not having a drop
01:20:24
to the bottom line and they're smart. Or
01:20:26
both things are true. Where do you
01:20:27
stand, Sax? be the uh adjudicator of
01:20:29
this case.
01:20:30
>> I mean, honestly, I'm probably closer to
01:20:31
you, Jal. I hate to say that. Um,
01:20:35
>> listen,
01:20:36
>> because look, I think people are still
01:20:38
figuring out how to drive business value
01:20:40
out of AI and change management is hard
01:20:43
and the bigger the company, the harder
01:20:45
it is. So, what's happening right now is
01:20:48
there are a lot of transformation
01:20:49
projects at large enterprises that are
01:20:51
failing. There's like a big Mackenzie
01:20:53
study on that. But if you look at
01:20:55
activity from the bottom up, I think
01:20:57
it's very interesting and it's becoming
01:20:59
more interesting. And over the last
01:21:01
several months, obviously with coding
01:21:04
reaching a new level, we're starting to
01:21:06
see very interesting things happening
01:21:07
there. And obviously the revenue that's
01:21:10
now being generated from these coding
01:21:11
models is, you know, again, it's
01:21:13
exponential like we've never seen
01:21:14
before. So the ROI
01:21:17
is finally there at the model layer.
01:21:20
Meaning before, you know, people were
01:21:22
were saying that it's a bubble because
01:21:25
>> you had all this massive capex at the
01:21:26
data center level and there was no ROI
01:21:29
coming at the model layer. Now we have
01:21:31
the ROI at the model layer and I guess
01:21:34
we're sort of questioning whether the
01:21:37
ROI will be there at the application
01:21:38
level. But any event, I think things are
01:21:40
progressing. I mean, look, I
01:21:41
fundamentally I'm bullish on this whole
01:21:43
thing.
01:21:43
>> Okay. No, let me be clear. Hold on. Hold
01:21:45
on.
01:21:45
>> Okay. Obviously, I'm bullish. I'm in
01:21:47
this I'm in the space. I'm doing it.
01:21:49
you're in the arena.
01:21:51
>> We have seen in every single wave in the
01:21:54
mobile wave, we needed consumers to show
01:21:56
up at scale. So, we needed consumer
01:21:58
experiences. And it was very obvious
01:22:01
that there were these consumer
01:22:02
businesses that were going to be, if not
01:22:04
already, incredibly, incredibly
01:22:06
profitable. Google and Facebook were
01:22:08
profitable within the first few years.
01:22:10
They never looked back. And all I'm
01:22:13
trying to point out to you guys is there
01:22:15
is not one great example yet. If we
01:22:18
believe enterprise is where all the
01:22:20
money is and if we believe that's what's
01:22:22
going to underpin these trillion dollar
01:22:23
valuations, just please somebody show me
01:22:26
a couple of good examples of scaled
01:22:28
profits.
01:22:29
>> Okay, fair enough. Travis, uh, final
01:22:32
word here on
01:22:33
>> By the way, can I say one other thing? I
01:22:34
think I think Chimath has a good point
01:22:36
about being riskoff right now because if
01:22:38
you just look at valuation metrics, they
01:22:40
do seem to be quite high and I've seen
01:22:42
other uh versions of those metrics as
01:22:44
well. So, look, it's very very hard to
01:22:46
time the market. I don't try, but I
01:22:49
think you could make an argument just
01:22:51
based purely on valuation levels, not
01:22:54
events, that you want to adopt a more
01:22:56
conservative posture right now. I I
01:22:58
think the bet you're making then,
01:23:00
Travis, is are these valuations
01:23:04
so high right now on the standards that
01:23:07
the
01:23:09
efficiency that AI could bring to these
01:23:12
companies is uh not real enough to
01:23:16
continue the growth from here. It's not
01:23:18
a catalytic enough technology. And you
01:23:21
know, when you look at
01:23:21
>> No, it's the opposite. No, all all these
01:23:23
traditional companies have horrible
01:23:25
valuations. They've been crushed. So my
01:23:26
point is if AI is real, the upside is
01:23:29
also real. All I'm saying is
01:23:32
>> the details guys matter. It's very hard
01:23:35
to take a very complicated business and
01:23:37
all of a sudden quote unquote transform
01:23:39
it.
01:23:40
>> It's not as easy as it sounds. That word
01:23:42
easy to say
01:23:44
>> for sure. Travis, go ahead. Final word.
01:23:46
Final word from Travis. Couple things.
01:23:47
First is when it comes to big companies,
01:23:49
I think the big thing about let's call
01:23:51
it the the autonomous enterprise is
01:23:54
change management. is the big boy. And
01:23:58
and change management actually is about
01:23:59
all the people that already work there,
01:24:01
the middle managers, the technocrats,
01:24:02
the bureaucrats, the whatever, the
01:24:07
>> getting the change management going. The
01:24:10
change management going there is it's a
01:24:12
human thing and it's very tricky with
01:24:15
very complex processes, many of which
01:24:17
are not even documented. And in theory,
01:24:20
it's just all going to happen real fast.
01:24:21
But in practice, like that's hard. So
01:24:23
that that's part one. Part two, what I'm
01:24:26
seeing with true tech companies,
01:24:28
>> real companies, public, I would say
01:24:31
public like hardcore public companies. I
01:24:35
mean like you know founder, you know,
01:24:37
folks that are really cranking public
01:24:40
companies and tech companies that are
01:24:42
sort of upandcomers.
01:24:45
I talked to CEOs across the board and
01:24:47
they're like they are fired up about the
01:24:50
development the sort of the the
01:24:53
productivity and deployment schedule and
01:24:56
like the new features they were able to
01:24:59
roll out much much much faster because
01:25:01
they've pivoted their culture sort of
01:25:03
very pro- AI development. Um, and I'm
01:25:07
getting like almost at this point a
01:25:09
consistent feedback from real founder
01:25:12
CEOs that like this stuff's real and
01:25:16
it's it's it's not just hype. Now, there
01:25:18
are folks pushing, you know, selling
01:25:20
their book that are like, "Oh, we're at
01:25:22
AGI and all this." Anybody who's worked
01:25:25
with these agents and done the AI dev
01:25:29
stuff,
01:25:31
there's a lot of good stuff, but they're
01:25:33
not they're not that smart yet. They're
01:25:36
just not that smart. Anybody who's done
01:25:38
like I got a side quest where I'm just
01:25:39
investing. I have agents investing and
01:25:42
betting on Kali and Paulie and you know
01:25:45
these other places and it's silly how
01:25:49
dumb the agents are. Even their best
01:25:50
agents to be honest.
01:25:53
>> Yeah.
01:25:54
>> The agents are what you you have to be
01:25:56
human in the loop with the agent. The
01:25:58
agent has no taste. The agent can do
01:26:00
repetitive tasks. the agent is not going
01:26:02
to do something novel and they quickly
01:26:04
can get lost in the forest.
01:26:06
>> We had to spend a lot of time we had to
01:26:07
spend a lot of time with our investing
01:26:10
agents getting them on board with the
01:26:12
idea that if you want to make money
01:26:15
investing, you can't be on both sides of
01:26:17
the same bet.
01:26:20
>> Yeah.
01:26:21
>> You know what I mean?
01:26:22
>> There it is.
01:26:22
>> It's just like that. That's where we're
01:26:25
at. It's the AGI is not here and it's
01:26:27
kind of silly for folks, I think, to
01:26:28
sort of suggest it is.
01:26:30
>> Absolutely. Hey guys, we got a wrap.
01:26:33
Travis, you didn't get to play uh the
01:26:35
price is wrong. I have two more. You
01:26:37
guys want to do a bonus round of the
01:26:39
Price is Wrong?
01:26:39
>> Yeah, do it. Do it. See if See if Travis
01:26:41
gets it.
01:26:42
>> Okay, two more. But you guys can steal.
01:26:44
You guys can steal here. Let me get my
01:26:46
back. I got to give me the music.
01:26:47
Travis, you're going to get to do the
01:26:49
bonus round here.
01:26:50
>> Let's go.
01:26:50
>> Hello everybody. We got Travis Kalanick.
01:26:53
He's here. He is a professional water
01:26:55
sports player from Tallahassee, Florida.
01:26:59
Uh, how is it down there in Tallahass,
01:27:00
Florida? I understand you do a little
01:27:02
handy work uh to pay the bills, but you
01:27:04
spend your most of your time out there
01:27:05
on the lakes and the oceans doing water
01:27:07
sports. Yeah, you like the water sports.
01:27:09
>> I love the water sports. I I doing water
01:27:12
ski lessons on the weekends. Anybody's
01:27:14
interested?
01:27:15
>> All right, I'll be out there uh at
01:27:17
Tallahassee, Florida. Okay, now here we
01:27:19
go. It's your job to figure out the
01:27:23
price is wrong. The price is wrong.
01:27:26
Here we go.
01:27:27
>> It's so funny. The
01:27:28
>> price is wrong here. Here we go. This is
01:27:30
your bonus round.
01:27:32
>> Yeah.
01:27:33
>> All right. This startup raised $900
01:27:36
million
01:27:37
>> at a $9 billion valuation. This is a big
01:27:40
one.
01:27:41
>> This is a big one. Before dissolving in
01:27:43
2018
01:27:45
>> and their famously deep voiced founder
01:27:48
CEO
01:27:49
>> is currently serving an 11-year sentence
01:27:51
in a federal prison.
01:27:53
>> Yes.
01:27:54
>> Can you name that mispriced startup? You
01:27:56
name the mispriced story.
01:27:59
>> It rhymes withinos.
01:28:02
>> Okay, you're you're closing in here. Be
01:28:05
careful. You have to formulate a
01:28:06
question. You could lose twist the
01:28:08
>> final answer. Theronos.
01:28:09
>> Ferrronos. Okay, very good. 100 points.
01:28:11
You're on the board. And here is your
01:28:13
next one. This is your next one.
01:28:16
>> Okay, this is the overtime. All three
01:28:19
players get to play this one. Whoever
01:28:21
zooms in first and says it, this is your
01:28:23
final.
01:28:24
That's the mispriced startup here. The
01:28:27
price is wrong on this startup. Here we
01:28:29
go.
01:28:29
>> Come on, you can do this. Come on,
01:28:30
Jimoth.
01:28:30
>> You got this now. Here we go.
01:28:32
>> You can do this.
01:28:33
>> This short form mobile first streaming
01:28:36
platform raised 1.7 billion
01:28:39
>> from top investors. Jeffrey has got the
01:28:42
the the the uh what is it?
01:28:43
>> You got to YELL IT OUT. TOP INVESTORS
01:28:45
ACROSS Silicon Valley and shut down in
01:28:48
just 6 months.
01:28:48
>> Quicky quickie
01:28:49
>> quickie. Wrong. David Sax wrong
01:28:55
company. JEFFREY JACKSONBY
01:28:59
WINS.
01:29:01
>> YES.
01:29:02
>> There's your champion everybody. Let's
01:29:04
tell him what he won. He uh apparently
01:29:06
he's won a Pier
01:29:09
in downtown uh East Austin. A Pier in
01:29:12
East Austin by the airport. God knows
01:29:15
what shenanigans are going down there. I
01:29:16
think if you need methamphetamine or a
01:29:18
date, you're going to have an easy time
01:29:20
there. Chimoth. Okay. Oh, hey, Chimath.
01:29:22
Um, I'm getting hammered with people who
01:29:23
want to come to the soldout liquidity.
01:29:25
500 was the cap you put on it. You've
01:29:27
been [ __ ] around as dictator. Can we
01:29:30
add 50 seats?
01:29:31
>> I think we can add 50 or 100. I don't
01:29:33
know. Lisa, can we add 100?
01:29:34
>> Get a Simocast room going. All right,
01:29:37
we'll see you all. Oh, and the All-In
01:29:39
Summit tickets are on sale. Go to
01:29:40
allin.com. Uh, and yeah, don't get shut
01:29:43
out of the summit, folks, because this
01:29:44
happens every time. You guys email me
01:29:47
two months out, the tickets have been
01:29:48
sold out for four months. So, get your
01:29:49
tickets now uh at allin.com. Another
01:29:53
amazing episode and we'll see you next
01:29:55
time. Bye-bye.
01:29:56
>> Bye-bye.
01:29:58
>> We'll let your winners ride.
01:30:01
>> Rain
01:30:05
and we open sourced it to the fans and
01:30:08
they've just gone crazy with it. Love
01:30:10
you. Queen of
01:30:13
>> your
01:30:18
besties are gone.
01:30:21
>> That is my dog taking your driveways.
01:30:26
>> Oh man, my appetasher will meet me at
01:30:29
>> We should all just get a room and just
01:30:30
have one big huge orgy cuz they're all
01:30:32
just useless. It's like this like sexual
01:30:34
tension that they just need to release
01:30:35
somehow.
01:30:39
Wet your feet.
01:30:42
>> We need to get merch.
01:30:52
I'm going all in.
