00:00:00
All right, everybody. Welcome back to
00:00:01
the number one podcast in the world, the
00:00:04
All-In podcast. Today, the original
00:00:06
squad is back. No guests, just Full
00:00:09
Contact Sachs is back. We're going to
00:00:11
break down a number of topics today,
00:00:12
including chaos in the bond market as
00:00:16
Republicans seemingly have abandoned the
00:00:18
Doge agenda. And we're going to debate
00:00:20
the big beautiful bill, which is going
00:00:23
to add trillions to our national debt.
00:00:25
And it's a new era for Google as they
00:00:28
released an AI first search product.
00:00:30
OpenAI has acquired Johnny IV's company
00:00:33
IO to make some sort of gadget. They've
00:00:36
paid over $6 billion. Is this the real
00:00:39
deal or not? You're going to find out
00:00:40
today. And is energy the solution to
00:00:43
every problem we have? One besties got a
00:00:46
very interesting take. All that and more
00:00:49
on the number one podcast in the world.
00:00:51
Stick with us. For the fourth year in a
00:00:53
row, we're going to do the all-in
00:00:55
summit. If you guys remember four years
00:00:57
ago, I decided, hey, let's do a summit.
00:00:59
And then these three besties were like,
00:01:01
"No, we don't want to do a summit." Now,
00:01:02
it's become a big thing, and everybody
00:01:04
loves it. September 7th to 9th in Los
00:01:06
Angeles. I do give you credit for that,
00:01:07
Jal. Yeah, after you gave me hell for a
00:01:09
year. Okay, that sounds like an apology,
00:01:12
but okay, sure. I'm sorry. You're
00:01:13
wonderful, and I appreciate what you
00:01:15
created. Thank you very much. Wow,
00:01:17
finally some credit. Okay, we all know
00:01:19
the goal of the summit is to have the
00:01:20
world's most important conversations and
00:01:21
this year will be no different. yada
00:01:23
yada yada
00:01:25
allin.com/summit if you would like to
00:01:26
apply to come. Why do you say the most
00:01:28
important conversations and then say
00:01:30
yada yada yada as a way to dismiss it?
00:01:32
Do you think that that increases your
00:01:34
salesmanship and effectiveness? I don't
00:01:36
want to make any grand pronouncements
00:01:37
here, but these have been important
00:01:39
conversations we've had. It's been
00:01:41
pretty great, I have to say. yada yada
00:01:42
yada. No, there's just like they put
00:01:45
like 18 things in the plugs now and I
00:01:47
don't want to do too many. Why don't you
00:01:48
do a modeicum of work before you get on
00:01:50
camera and actually edit it. Here we go.
00:01:52
I get one compliment and then that's
00:01:54
exactly it. He doesn't do the homework.
00:01:57
So, he runs out of words. He's like, "We
00:01:59
get the biggest names to come to our
00:02:01
summit. The vice president came. He said
00:02:04
yada yada yada." Elon Musk showed up in
00:02:06
person. yada yada yada. There was an
00:02:07
astronaut. Freeberg lost his yada yada
00:02:10
yada yada yada yada. Uh, but this year
00:02:12
will be no different. We'll have
00:02:14
incredible parties, blah blah blah, and
00:02:16
let's get to work. Okay. President did a
00:02:19
trip to the Middle East, yada yada yada.
00:02:21
Maybe more peace, a whole new framework.
00:02:23
America's going to yada yada yada yada
00:02:25
yada. World's greatest moderator, yada
00:02:27
yada yada. Executive producer for life.
00:02:29
Deal with it, Fredber. Okay. Uh, what's
00:02:32
in the contract?
00:02:34
Let your winners ride.
00:02:37
Rainman David.
00:02:42
We open sourced it to the fans and
00:02:44
they've just gone crazy with it.
00:02:49
All right, let's get to work here. The
00:02:52
bond market is the captain. Apparently,
00:02:54
Treasury Department sold $16 billion
00:02:56
worth of newly issued 20-year bonds on
00:02:58
Wednesday afternoon, and there was weak
00:03:00
demand. This pushed yields higher across
00:03:02
the board. the 10-year, which Beston has
00:03:04
said to focus on because it's the
00:03:06
benchmark rate for most borrowing costs,
00:03:08
you know, mortgages and stuff like that.
00:03:10
It spiked and here it is. It actually
00:03:12
hit a five handle at one point. A lot of
00:03:14
people were hand ringing about this. And
00:03:16
the S&P dropped 1.5% in about 30
00:03:18
minutes. Here's that chart as
00:03:20
well. And the three major indices were
00:03:23
all down between 1.5 and 2% on the day.
00:03:27
Obviously, in related news, the House
00:03:28
passed the big beautiful bill at the
00:03:31
11th hour last night. Uh this makes the
00:03:34
TCG JA tax cuts permanent. Um and it's
00:03:38
estimated to increase longr run GDP by
00:03:40
60 basis points. Uh that's if all the
00:03:42
cuts were
00:03:43
implemented. It's also going to reduce
00:03:45
tax revenue by 4 trillion over 10 years
00:03:47
is the estimate and it's going to add
00:03:50
between 3 and 5 trillion to the national
00:03:52
debt over 10 years. free break. What's
00:03:54
your take on the tripleB and the bond
00:03:58
weak bond market, all of it? I just want
00:04:00
to take a quick primer on Sure. how the
00:04:03
government gets funded. I know we we
00:04:06
assume everyone understands it, but I
00:04:07
think it's important for folks to really
00:04:08
grock it. For the federal government to
00:04:11
make payments to employees and
00:04:14
contractors and buy stuff, they need to
00:04:17
put money in their bank accounts. And
00:04:18
the way they put money in their bank
00:04:19
accounts is they issue bonds. These are
00:04:21
treasury bonds. So they'll sell treasury
00:04:24
bonds to the public and individuals buy
00:04:28
it, companies buy it, banks buy it, and
00:04:30
foreign governments buy US treasuries.
00:04:32
And they transfer or wire cash into the
00:04:36
federal government's bank accounts,
00:04:37
which they can then use to pay for
00:04:38
stuff. And then the Treasury Department
00:04:40
needs to continuously sell treasuries to
00:04:43
raise cash to fund the government. When
00:04:45
folks don't show up to buy
00:04:48
treasuries, that's a bad thing. And that
00:04:51
means that the government needs to
00:04:54
increase the interest rate that they're
00:04:56
paying on those bonds. So what we saw on
00:04:58
Wednesday was a really weak demand
00:05:01
signal for treasuries on this, you know,
00:05:03
modest treasury auction. Selling 16
00:05:05
billion of bonds is not a lot. There's
00:05:09
hundreds of billions being sold each
00:05:11
quarter. And so this was not a big
00:05:13
number, but there was no buyer. The
00:05:15
market was really dry. And so everyone
00:05:18
that participates in financial markets
00:05:19
saw this and freaked out. And the big
00:05:21
motivation, the big understanding of the
00:05:22
relationship here is to your point that
00:05:25
this big tax bill and spending bill is
00:05:28
passing out of the house. And that bill,
00:05:30
as we talked about last week, has a high
00:05:33
deficit and that'll run up US debt over
00:05:35
time, which makes it more difficult for
00:05:37
the US government to pay its bills
00:05:39
because it has to issue more debt.
00:05:40
Interest payments have to be paid every
00:05:42
year and so on. Now, if you just pull up
00:05:44
this one slide, this is something I
00:05:46
thought was really worth sharing. The
00:05:47
CBO estimates, which is what you
00:05:48
referenced, Jason, are estimates of what
00:05:52
the cost is going to be over time. And
00:05:55
this is from Jessica RLE or Ridle out of
00:05:59
the Manhattan Institute put this chart
00:06:01
together.
00:06:03
Budget Office. Yeah. So the
00:06:05
Congressional Budget Office, the CBO
00:06:08
creates an estimate of the budget, the
00:06:11
spending, and the deficit, and
00:06:13
ultimately the incremental debt that the
00:06:15
US government will need to issue to fund
00:06:17
its obligations over time. This is a
00:06:19
chart that was put together that the
00:06:21
expectation on a baseline basis is that
00:06:24
over the next 30 years or so, US debt to
00:06:26
GDP will climb to 203%. But what a lot
00:06:29
of people don't know and don't talk
00:06:31
about is that in the CBO estimates, they
00:06:34
assume in all of their models that
00:06:36
interest rates are at
00:06:39
3.6%. And with the bond selloff
00:06:42
yesterday, what we are now seeing is
00:06:44
30-year Treasury interest rates at over
00:06:48
5%. It's 5.1% this morning. And for
00:06:51
every incremental 1% above 3.6, six,
00:06:55
you're spending an extra 350 billion a
00:06:58
year in interest. 350 billion a year.
00:07:01
So, as we go up by 1.5% interest over
00:07:04
the next 10 years, we're spending
00:07:06
another 5 trillion just on interest
00:07:09
payments. Half a billion dollars after
00:07:11
trillion dollars a year of incremental
00:07:13
interest on the difference between 3.6
00:07:15
and 5.1. That's an incredible recursive
00:07:19
problem. And I say recursive because the
00:07:22
incremental interest that we now have to
00:07:24
pay because interest rates just went up
00:07:26
because the market is demanding more
00:07:27
payments from the government to fund
00:07:29
this means that the government has to
00:07:30
issue more debt and it quickly gets away
00:07:33
from you. There is a nonlinear
00:07:35
relationship between the deficit and
00:07:38
interest rates which drives up the debt
00:07:40
problem in a nonlinear way and it gets
00:07:42
away from you and you can't fix it. And
00:07:44
that's what the market is telling us is
00:07:46
that the current bill that's being
00:07:47
passed out of the house is showing such
00:07:49
an extraordinarily high deficit that the
00:07:52
market does not want to buy the debt
00:07:53
from the government. Rates are now
00:07:55
climbing and that creates a massive
00:07:56
problem for the government. Okay,
00:07:58
Chimath, you've been tweeting about
00:08:01
this. Uh we thought we were going to get
00:08:05
uh an administration and maybe um some
00:08:09
focus from the government that we would
00:08:10
be have more austerity measures, balance
00:08:12
the budget. Now it looks like we're
00:08:14
going to pour gasoline on the fire. What
00:08:16
are your thoughts? I think we have to be
00:08:18
careful here. So Jason, the political
00:08:20
calculus will be for the president to
00:08:22
decide how much credit he actually wants
00:08:24
to take for this bill because even
00:08:27
though it has his name on it, the
00:08:30
contents of the bill are different in
00:08:33
actual
00:08:35
facts than what I think he intended. And
00:08:39
what I mean by that is when you look
00:08:41
inside of what happened in the 11th hour
00:08:43
last night, it's disappointing. This
00:08:46
thing is like anti-D Doge. If Doge was
00:08:48
meant to be a
00:08:51
reflection of the American voting
00:08:53
population's desire for meaningful
00:08:56
reform in government, cost
00:08:59
controls, some form of austerity, and to
00:09:02
get this debt spiral in check. This is
00:09:05
the opposite of that.
00:09:07
What happened was in the 11th hour, you
00:09:10
had a handful of people abstain. You had
00:09:13
one person that passed away in the last
00:09:15
few days. You had one person that fell
00:09:17
asleep on the floor of the house, so he
00:09:19
wasn't even woken up for the vote. And
00:09:21
in the middle of all of that
00:09:24
chaos, what happened was all kinds of
00:09:26
things were added and attached and
00:09:29
cancelled at the last minute because
00:09:31
what happens is you have to have these
00:09:33
puts and takes as Freeberg described. If
00:09:35
you want to spend over here, you have to
00:09:38
find a cut over there. But I think what
00:09:40
happened was there was really not a lot
00:09:42
of financial literacy used to decide
00:09:45
what to actually put in and what to cut.
00:09:47
And that lack of
00:09:49
discipline is going to create I think a
00:09:54
negative set of consequences. So what
00:09:57
are those
00:09:58
consequences? Today the tenure is around
00:10:01
4.5%. At the rate in which it's
00:10:04
escalating since liberation day, by the
00:10:07
end of this year, we're going to be past
00:10:09
5%. The 30 year is on a rate now to get
00:10:14
past 6 and a quarter, maybe even reach 6
00:10:19
1/2%. Those
00:10:21
are way
00:10:23
beyond what most people thought was a
00:10:28
reasonable place to be for the United
00:10:30
States economy. And so what will the
00:10:32
implications be as rates go to those
00:10:36
levels? You'll delever from the United
00:10:39
States. You'll sell US debt. You'll own
00:10:42
things like gold and Bitcoin. If you're
00:10:44
curious about what's happening to gold
00:10:46
and Bitcoin, they started to spike in
00:10:48
the last few
00:10:49
days. You'll have ratings organizations
00:10:52
that add to this cascade by downgrading
00:10:55
the United States. That happened on
00:10:58
Friday. you'll have very smart people
00:11:01
starting
00:11:02
to signal that this is a much harder
00:11:05
problem than they initially thought.
00:11:08
That's how I personally interpreted
00:11:10
Elon's comments over the last few days.
00:11:12
So, what was the House supposed to do? I
00:11:15
think what they were supposed to do was
00:11:17
implement some form of austerity. They
00:11:20
were supposed to by the will of the
00:11:22
people pass a recision bill. They were
00:11:25
given that recision bill. It was just $9
00:11:28
billion. They couldn't even pass a $9
00:11:32
billion recision. And instead, they
00:11:34
passed a $4 trillion inflation to our
00:11:36
debt. Now, you hand this to the Senate.
00:11:39
The Senate is in a very difficult place
00:11:40
as well. Do they want to quote unquote
00:11:44
claim victory and say, "Here you go,
00:11:46
President Trump. Here's your bill." But
00:11:48
they'll further bastardize this thing
00:11:50
and it will be even further away from
00:11:52
what I think benefits MAGA and benefits
00:11:56
Main Street. So who does it benefit?
00:11:58
Current course and speed right now. This
00:11:59
bill is about traditional Republicans
00:12:02
and traditional Democrats circling the
00:12:05
wagon and putting on a platter a set of
00:12:08
things that I think will be hurtful to
00:12:11
average Americans. You're going to see
00:12:13
energy prices spike. You're gutting the
00:12:16
number of electrons that will be
00:12:17
available for things like AI. You're
00:12:19
going to increase Medicare prices and
00:12:22
the math is wrong. So when you sensitize
00:12:25
this thing to a 4 and a half or five or
00:12:30
five and a quarter rate, so meaning not
00:12:32
what the CBO used but the real
00:12:35
conditions on the
00:12:36
ground, this thing is an albatross. And
00:12:39
I think unfortunately for President
00:12:42
Trump's agenda and for a MAGA movement,
00:12:45
this is the worst of all conditions. The
00:12:47
financial markets will punish this. And
00:12:49
then the last thing I'll say is now to
00:12:52
top it all off, I think that Jerome
00:12:55
Powell will see the writing on the wall.
00:12:58
Many aspects of this thing are
00:12:59
inflationary and if they're not handled
00:13:02
well by the
00:13:03
Senate, he has a lot of room to actually
00:13:06
increase interest rates.
00:13:09
So, I would just say that the Senate has
00:13:11
an incredibly difficult job. I think the
00:13:13
president has an even more difficult job
00:13:16
about what to do right now. But the
00:13:17
House did nobody favors. They did not do
00:13:20
anybody a favor yesterday. Here's the uh
00:13:23
prediction from Poly Market on what will
00:13:26
happen to Fed rate cuts in 2025. It's
00:13:30
across the board. It seems like nobody
00:13:32
can tell if it's going to be zero, 1, 2,
00:13:34
or three. 20 22 25 and 15% respectfully
00:13:39
respectively for those four options.
00:13:42
Sax, you don't speak for President
00:13:44
Trump. You're the zar of AI and crypto
00:13:47
specifically. So, let's put that out
00:13:49
there. But I'm curious your thoughts on
00:13:52
what happened with Doge. Why are we
00:13:54
continuing to spend ourselves into
00:13:56
oblivion? And what will the
00:13:59
downstream issues be here? Because this
00:14:01
sounds like it's going to be
00:14:02
inflationary. This sounds like it's
00:14:03
going to be hard for people to buy
00:14:05
homes. It's going to create more
00:14:06
austerity in the future. How bad is
00:14:09
this? Do you feel it's as bad as Dave
00:14:12
and Chimath are framing it? Well, look,
00:14:14
I mean, I wish we had 435 members of the
00:14:17
House who thought like Freeberg about
00:14:19
spending and deficits. Um, we don't. The
00:14:22
Democrats all want a lot more spending
00:14:24
and a lot more taxes. Remember, if it
00:14:26
wasn't for Mansion Cinema, we would have
00:14:27
had that $4.5 trillion build back
00:14:30
better. And among the Republicans in the
00:14:33
House, unfortunately, we have a bunch of
00:14:35
Republicans who are pretty soft on
00:14:37
spending. I mean, you can call them
00:14:38
rhinos if you want. And we only have
00:14:41
what, like a three vote margin in the
00:14:43
House, and the Democrats are not
00:14:44
cooperating in any way. So, you don't
00:14:46
have that many votes to spare. So, what
00:14:49
I'm saying here is you can't make the
00:14:50
perfect the enemy of the good. I think
00:14:51
you have to be realistic about how much
00:14:52
we can get done here. And I do think
00:14:55
that this bill does contain a lot of
00:14:56
good things in it. There are a lot of
00:14:58
priorities for the administration in
00:15:00
here. Do I wish it cut spending more?
00:15:03
Yes. I mean, do I wish that it made all
00:15:06
the Doge cuts permanent through
00:15:08
recisions? Yeah, absolutely. I think
00:15:09
it's outrageous that there were enough
00:15:11
House Republicans who didn't want to
00:15:13
back up Doge that that wasn't enacted.
00:15:16
But there are a lot of good things in
00:15:18
the bill and I should just highlight
00:15:19
them for balance here. So, the number
00:15:22
one thing here is that this bill will
00:15:24
permanently extend the 2017 tax cuts.
00:15:28
And if we don't pass this bill, let's
00:15:30
just assume that we do nothing. Okay, if
00:15:33
we do nothing, you're going to get the
00:15:35
largest tax increase in decades, and it
00:15:38
won't just affect the rich. This affects
00:15:40
the middle class as well. The 2017 tax
00:15:42
cuts increase the standard deduction,
00:15:44
which is important for the middle class.
00:15:46
It raised the child tax credit. So
00:15:49
again, if you do nothing, you get a huge
00:15:51
tax increase as soon as the 2017 tax
00:15:54
cuts, which Trump passed back in his
00:15:56
first term when they
00:15:58
sunset. So that's number one. In
00:16:01
addition, there are promises here that
00:16:03
were made during the campaign to
00:16:05
eliminate taxes on tips and overtime
00:16:07
that have been enacted, and the
00:16:09
president promised to do those things,
00:16:10
and he is accomplishing those things. So
00:16:13
I think that's important. There's
00:16:14
funding for the border wall here, 10,000
00:16:16
new ICE officers, more detention beds,
00:16:19
and so on. On energy, it repeals a
00:16:21
methane tax and unlocks new oil and gas
00:16:24
on federal lands. And there's a bunch of
00:16:26
other things as well. So, there are a
00:16:27
lot of good things in here. Now, to
00:16:29
point, will this bill be bad for the
00:16:32
Republicans prospects in the midterms? I
00:16:35
guess what I would say to that is I
00:16:37
don't understand how you expect to do
00:16:39
well in midterms if Republicans preside
00:16:41
over the largest tax increase in
00:16:43
decades. A tax increase that Joe Biden
00:16:47
tried to get during his administration.
00:16:49
Biden tried to raise taxes and couldn't
00:16:51
get the votes for it. But if we don't
00:16:53
make those tax cuts permanent, if we
00:16:54
allow the sunset to happen, then you
00:16:57
will accomplish what Joe Biden could
00:16:59
not, which is to raise taxes on the
00:17:00
American people. So, look, no bill like
00:17:03
this that's passed by a one vote margin
00:17:05
on straight party lines when you only
00:17:06
have a three vote margin in the entire
00:17:08
house is going to satisfy everyone here.
00:17:11
Let me Well, let me ask you a question
00:17:13
there. You have Trump at his peak power.
00:17:18
You have all these threats of, hey,
00:17:19
we're going to primary anybody if they
00:17:21
don't do like what we're saying here.
00:17:23
Why isn't Trump just coming in and using
00:17:25
his power and just saying, hey, not good
00:17:28
enough. Get back to work. we need to cut
00:17:30
spending. Period. Full stop. And he's
00:17:33
got that power. I don't hear him saying
00:17:34
that. I do think that there are ways to
00:17:37
enact the tax cuts and make them
00:17:41
permanent. And I think that that is a
00:17:43
good part of what this bill
00:17:46
did. The problem was what happened from
00:17:49
11:00 p.m. to 6:00
00:17:51
a.m. And that's all of the nonsense
00:17:54
where I'll just be honest. I think the
00:17:56
people in the House are not nearly as
00:17:59
financially literate as they need to be
00:18:02
and I think what they wanted to do was
00:18:04
pass something and throw it over the
00:18:07
wall and say mission
00:18:09
accomplished. Now it's up to the Senate
00:18:12
and the good news is look, the Senate
00:18:13
has these six-year election cycles for a
00:18:16
reason. It allows them to think
00:18:18
strategically and see past what's right
00:18:21
in front of them. So, I still think that
00:18:24
there's a chance for the president to
00:18:25
have his cake and eat it, too. But I
00:18:27
just want to highlight that as a
00:18:29
supporter of his and a supporter of that
00:18:32
MAGA agenda, which I think a lot of
00:18:34
people voted for, I just want to be
00:18:37
honest and say this is not it. And there
00:18:39
has to be some meaningful reforms in the
00:18:42
Senate version of this thing that gets
00:18:44
sent back to the House. So, I'll give
00:18:46
you an example of one. 81% of the
00:18:49
incremental energy that's generated in
00:18:52
the United States last year came from
00:18:55
private enterprises that were investing
00:18:57
in short-term forms of power that have a
00:19:01
credit and a transfer market tied to it.
00:19:04
This is just the financial machinery.
00:19:06
This is how Blackstone and Goldman Sachs
00:19:09
and Brookfield and all of these big
00:19:11
players move tens and hundreds of
00:19:13
billions of dollars
00:19:15
around. They changed those rules at the
00:19:17
11th
00:19:19
hour. What is the impact of that? Elon
00:19:22
foreshadowed it. Nick, maybe you can
00:19:24
show the headline of the article. He
00:19:26
said, "We could be running into a power
00:19:29
capacity issue by mid next year." Now,
00:19:33
if we knew that, why would we take the
00:19:35
short-term incentive away to generate
00:19:38
more electrons? All the great things
00:19:40
that David has done, that the president
00:19:42
has done on AI, I mean, we'll talk about
00:19:44
what happened in the Middle East.
00:19:46
But when the rubber meets the road and
00:19:49
you actually have a shortage of
00:19:50
electrons because these financial
00:19:52
actors, they're not, you know, and in
00:19:54
fairness to them, they're acting
00:19:56
rationally. You take the financial
00:19:58
incentive away and you can't underwrite
00:20:00
this thing. What are they going to do?
00:20:01
They're just going to stop doing it. So
00:20:02
in the absence of electrons, what
00:20:04
happens? Prices go up. It's
00:20:07
inflationary. And now you'll have to
00:20:09
allocate electrons. Does the electrons
00:20:11
go to an AI data center or does it go to
00:20:14
a home? In a different example, actually
00:20:16
on the residential side, places like
00:20:19
Florida, I think there was like an
00:20:21
incentive for like batteries. So if you
00:20:26
are about to get hit by a hurricane, you
00:20:28
can buy these big battery systems from
00:20:30
Tesla as an example, but there are other
00:20:32
ones, Generrack and whatever
00:20:34
train that was cut out. And so in Texas,
00:20:39
all of these impacts will have negative
00:20:41
consequences. My point is, I think you
00:20:44
have to really look at what happened in
00:20:45
the 11th hour and fix those things
00:20:48
because the problem with these plans is
00:20:51
you can't look at gains that are in
00:20:53
years 6 through 10 to justify pain in
00:20:57
years 1 through 5. You can't do that.
00:21:00
Because I understand how that
00:21:02
mathematically makes sense. It doesn't
00:21:04
make sense practically for the average
00:21:06
American. That's just not how life
00:21:07
works. You don't go into a J curve. You
00:21:10
know, we do that as investors, right?
00:21:12
But you don't do that as a normal
00:21:13
American, which is be on the wrong side
00:21:15
of the economy for four or five years in
00:21:17
the hopes that your six and seven are
00:21:18
good. That's the point that has to get
00:21:21
fixed. I asked Chimat the question,
00:21:24
shouldn't Trump use his capital he's
00:21:28
built up here and say not good enough,
00:21:30
make some cuts here? What would you say
00:21:32
to President Trump if he was asking you
00:21:35
for your advice? Freedber. I think the
00:21:38
writing's on the wall that the system is
00:21:42
winning and the rebels came in and they
00:21:46
tried to fix the inefficiencies in
00:21:49
government with Doge under Elon's
00:21:51
leadership and our friend Antonio's
00:21:53
participation and a great group of
00:21:55
individuals who are true patriots who
00:21:56
are spending time trying to help
00:21:58
reassess how the government operates.
00:22:01
And they, as I've said from day one, for
00:22:03
my first visit to DC during inauguration
00:22:05
weekend and every visit I've done since
00:22:06
then, I always leave after meeting with
00:22:09
members of Congress deeply worried that
00:22:12
they won't actually do what they need to
00:22:13
do to make the changes. And the last
00:22:16
line of defense ultimately is the
00:22:18
president with his veto authority and
00:22:20
his ability to tell people to go back,
00:22:22
rethink this, and cut spending. I feel
00:22:24
like we're in a last ditch effort. And
00:22:25
I'll tell you why I say it's last ditch.
00:22:27
I think that there's a relationship
00:22:29
between what's going on in the United
00:22:30
States and what's going on in Japan.
00:22:32
Nick, if you'll pull up the Japanese
00:22:33
government 40-year bond yield chart. So,
00:22:36
just so you guys can kind of observe, in
00:22:37
the last couple of days, we've seen
00:22:39
Japanese bond yields sore from 2.5% to
00:22:43
3.5%. Remember, Japan owns $1.1 trillion
00:22:46
of US treasuries. I'll just read a quote
00:22:49
from a guy uh who's the chief global
00:22:51
strategist named Albert Edwards at
00:22:52
Society General and he said recent
00:22:55
developments in
00:22:57
Japan may signal the end of the
00:23:00
favorable investment cycle that's been
00:23:02
established since the 2008 global
00:23:04
financial crisis potentially reshaping
00:23:06
global financial landscapes in the
00:23:08
coming months. Can I translate that
00:23:10
because it's it's incredibly important
00:23:11
what Freeberg said but I'll translate
00:23:13
that what he is saying. So this is a guy
00:23:15
from Sockgen since
00:23:17
2007208. Every now and then the markets
00:23:19
give you free
00:23:21
money. One of the most obvious free
00:23:23
money trades was the yen carry
00:23:25
trade. And in order to to execute that
00:23:29
trade, you're effectively leveraging
00:23:32
treasuries against the Japanese bond
00:23:34
market. What he is saying is that there
00:23:37
is increasing risk that that trade
00:23:39
unwinds.
00:23:40
When that trade unwinds, what you're
00:23:42
going to have are net sellers of up to,
00:23:45
as Freeberg said, a trillion plus
00:23:47
dollars of US
00:23:49
treasuries. What will that do? It will
00:23:51
further
00:23:52
exacerbate the market's reluctance to
00:23:55
own US treasuries, which means yields go
00:23:57
even higher, which means interest rates
00:24:00
for businesses and individuals go
00:24:02
higher. Correct. And the only way to
00:24:04
keep those interest rates low will be
00:24:07
for the Federal Reserve to try and step
00:24:09
in and buy those bonds, which will
00:24:12
devalue the dollar. And in his
00:24:15
commentary, he went on to say that uh
00:24:17
the 20-year Japanese bond auction that
00:24:19
just took place, similar to the US
00:24:21
Treasury auction that just took place,
00:24:23
was the worst they've seen since
00:24:26
1987. That's how significant this event
00:24:29
was in Japan over the past week. So the
00:24:31
massive climb in bond yields in Japan
00:24:34
indicates a massive selloff in credit.
00:24:37
And remember if you look at the
00:24:39
historical perspective when the global
00:24:42
reserve currency nation sees their debt
00:24:44
sell off it usually is part of a global
00:24:47
selloff that happens. It's not just the
00:24:49
US that gets affected. And any one of
00:24:52
these markets can cause a cataclysmic
00:24:54
follow on to the rest of the global
00:24:56
financial markets. like if Japan sells
00:24:58
off too much, we are going to start to
00:25:00
see a lot of unraveling happening.
00:25:02
That's the consequence I think of this
00:25:04
particular bill because it's it signals
00:25:06
a lack of a reluctance to actually
00:25:08
address. Let me let me say something
00:25:09
here. I called Freeberg today on the way
00:25:12
here. It's very rare that he and I agree
00:25:17
about this kind of stuff because
00:25:18
normally what I say is you're
00:25:20
catastrophizing. Sure, calm down. But
00:25:23
this is where I will say I'll put my
00:25:25
money where my mouth is and I'll put the
00:25:27
chips on the table to say this is the
00:25:29
moment. This is the moment where we
00:25:32
either reunderwite what we voted
00:25:36
for or we mistakenly pass this Christmas
00:25:40
tree concoction. It does have President
00:25:42
Trump's signature on the top of it. It
00:25:45
does have some features that he asked
00:25:47
for, but what I hope he sees and
00:25:50
realizes is the details here matter.
00:25:53
And without some technical maneuvering
00:25:55
in the Senate, this is a bad bill as
00:25:58
constructed. And if it
00:26:00
passes, forget what any of us think, the
00:26:04
bond markets will act decisively and
00:26:08
they are going to go in one direction
00:26:09
and it's away from us. And I just hope
00:26:13
that we figure this out. Seems like a
00:26:15
classic diffusion of responsibility
00:26:17
here, Sax. You have Congress, Senate,
00:26:19
and POTUS not willing to step up and
00:26:22
say, "Let's cut the spending."
00:26:25
Well, I don't I don't think that's fair.
00:26:27
I mean, look, the the president backed
00:26:29
up Elon and Doge. First of all, he
00:26:30
created Doge. He gave it the US digital
00:26:32
service. He backed it up in every
00:26:34
cabinet meeting, in every public
00:26:35
pronouncement. There's a limit though to
00:26:37
what he can do. Congress has the power
00:26:39
of the purse. And we said on a previous
00:26:41
episode that at the end of the day, for
00:26:43
the Doge cuts to become permanent,
00:26:45
Congress has to incorporate that into
00:26:46
the budget. And the reality is there are
00:26:50
still a number of these like old bull
00:26:52
appropriators on the Republican side who
00:26:55
just are not wired to make spending
00:26:57
cuts. Why are we hearing him vocally
00:26:59
saying stop this madness? He is the most
00:27:02
vocal guy in the world who mixes it up
00:27:05
every topic on the planet. Why is Let's
00:27:07
give him the benefit of the doubt. Let's
00:27:08
give him the benefit of the doubt. He's
00:27:10
probably working it all in the back
00:27:11
channel. There are the things that he
00:27:13
wants and he's in the midst of
00:27:14
negotiating as well. But my my point is
00:27:16
just this. The level of financial
00:27:19
illiteracy in this bill will come back
00:27:22
to bite America in the ass. Look, here
00:27:24
here's the problem is that what what you
00:27:26
guys want to sell right now is
00:27:28
austerity. And the reality is that the
00:27:31
politicians in Washington just are not
00:27:32
buying austerity right now. I mean,
00:27:34
that's what you want is austerity. Yep.
00:27:36
And the reality is the only way you get
00:27:39
true austerity, like a painful set of
00:27:41
changes, is when both parties link arms
00:27:44
and jump off the cliff together and and
00:27:46
make those changes. The Democrats right
00:27:48
now have no willingness to do that.
00:27:50
They're opposed to doing this. This is a
00:27:52
straight line party reconciliation vote.
00:27:55
I think you're trying to overload this
00:27:57
vote with way too much stuff. You're
00:27:58
trying to basically say that this vote
00:28:00
has to fix America's fiscal problems.
00:28:03
Just not going to happen. I mean, I
00:28:04
look, I agree with you that the fiscal
00:28:06
situation needs to get fixed. This is
00:28:07
just not the vote where it's going to
00:28:09
happen. You cannot exacerbate it. Well,
00:28:10
I don't know. I mean, look, I understand
00:28:12
that this doesn't have as much stuff as
00:28:13
as we want. I mean, like I said, I want
00:28:15
to see the Doge cuts made permanent, but
00:28:18
the alternative to this bill, let's say
00:28:20
the alternative is just do nothing. That
00:28:22
is actually meaningfully worse because
00:28:24
you will get a massive tax increase.
00:28:27
Massive. I think President Trump
00:28:28
inherited a Biden economy that was
00:28:30
already feeling pretty shaky. I don't
00:28:32
know what the argument is for a massive
00:28:33
tax increase in an economy that's not
00:28:36
doing great. I'd rather see the tax cuts
00:28:38
made permanent. I'd rather see the no
00:28:40
tax on tips and overtime implemented.
00:28:43
Important campaign promises. I'd rather
00:28:44
see funding for the border wall. I mean,
00:28:46
these are all important things that we
00:28:48
need to have happen. And I just don't
00:28:50
think that this bill is how you're going
00:28:51
to start. Good. But what I'm saying is
00:28:53
why do you pay for that with electron
00:28:55
supply for homes and AI? I don't know. I
00:28:57
mean, you're talking about a pretty
00:28:58
esoteric issue, Chimoth. This is not the
00:29:00
same as solving the US budget situation.
00:29:03
Why do it by doing Medicaid cuts? Why?
00:29:05
Well, let's talk about the Medicaid cuts
00:29:07
first of all. I mean, you're being a
00:29:08
little bit schizophrenic about this. I
00:29:10
mean, you guys are saying that you want
00:29:11
to see more cuts, but then you also
00:29:14
believe that the Medicaid cuts are a bad
00:29:15
idea. So, you got to kind of choose. I'm
00:29:17
not saying that. I think all cuts are
00:29:18
great. Cuts across the board. Yeah, cuts
00:29:21
across the board. Spread it out. You got
00:29:23
to decide what you're opposed to. Are
00:29:24
you opposed to cuts or you want more of
00:29:26
them? I want more cuts. And given the
00:29:28
fiscal situation, I'm actually if if
00:29:30
someone showed up and said, "Hey, we're
00:29:31
going to increase taxes." I'd say,
00:29:33
"That's what we got to do right now.
00:29:34
Everyone's got to take a hit on this."
00:29:35
But hey, I know I'm going to be alone in
00:29:37
that. So, I'm not going to try and make
00:29:38
that argument. I guess the question for
00:29:39
you, you're not alone in that. I agree
00:29:41
with you. Well, let me ask a question,
00:29:42
Sax. What do you think is the moment
00:29:44
when the two parties lock arms and do
00:29:47
jump together? What is the conditions
00:29:49
that need to be kind of prevalent for us
00:29:53
to take, you know, draconian austerity
00:29:55
measures? probably the bond market
00:29:56
forces it on the politicians force it on
00:29:58
Washington. Yeah, I think you're right.
00:30:00
You're an economist by background. Do do
00:30:01
you see that happening right now? Like
00:30:02
do you see this the sell-off that's
00:30:04
happening not just in the US, but
00:30:06
everyone that's linked to the US with
00:30:07
Japan and so on. Don't you think that
00:30:09
the markets telling us that this might
00:30:11
be the moment? It could be the moment,
00:30:13
but you know, I remember the first week
00:30:14
of April when Larry Summers came on the
00:30:16
pod and said that the markets were
00:30:18
melting down and that was a different
00:30:20
moment that he was trying to sell in
00:30:22
favor of his policy. So my point is just
00:30:24
there's market volatility and we won't
00:30:26
know for a while whether this means
00:30:28
anything. It could just be a blip,
00:30:29
right? So look, you ask me what are the
00:30:32
conditions for getting austerity. I
00:30:34
think that austerity will have to be
00:30:35
imposed on Washington from the outside.
00:30:38
Washington's not going to find the will
00:30:40
internally. We had a wonderful moment
00:30:43
here where you have Elon come in with
00:30:45
Doge, which the president made possible.
00:30:48
Okay. No other president would have even
00:30:49
have done this. Agreed. Agreed. Big
00:30:51
credit for that.
00:30:53
where you get a total outsider to come
00:30:54
in and bring a team of like young big
00:30:58
bald geniuses and you let them go
00:31:01
through these departments line by line
00:31:02
and start deleting things. Okay. And
00:31:05
they made I think great progress. I
00:31:06
think they found a 160 billion a year of
00:31:09
cuts that can be made administratively.
00:31:12
Beyond that they need Congress to
00:31:13
approve it. And I will acknowledge that
00:31:16
there are still a number of soft
00:31:19
onspending Rhino Republicans and there's
00:31:22
just not a majority for the types of
00:31:25
austerity and cuts that you guys would
00:31:27
like to see. I mean, that's just the
00:31:28
reality. Yeah. But I still think that
00:31:30
the question with this bill is just how
00:31:32
pure you want to be. My point is just
00:31:35
are you going to make the perfect the
00:31:36
enemy of the good? I don't think you're
00:31:37
going to get austerity out of Washington
00:31:38
right now. It's just not in the cards.
00:31:40
The question is whether this bill is
00:31:42
basically better than the status quo.
00:31:44
And I and I think it is because of the
00:31:46
tax cuts. To your point, without opining
00:31:49
on it, I think what's going to happen is
00:31:50
the bond market will sensitize this
00:31:53
budget at real rates. They're not going
00:31:56
to use the 3.6 that Freeb pointed to.
00:31:58
They're just going to rewrite this at 5,
00:32:00
five and a quart, or 5 1/2. That has a
00:32:03
very
00:32:04
different risk exposure, I think. And so
00:32:07
to your point, they'll react. If it
00:32:10
looks like it passes as is, then that's
00:32:12
what they'll do. And to your point,
00:32:13
David, we'll know in the next 60 to 90
00:32:15
days. Let's say that this is just a blip
00:32:17
in the market and it's not some larger
00:32:19
once in a century type event. The
00:32:21
reality is we don't know exactly what's
00:32:23
going to happen. Look, I don't like
00:32:24
America's fiscal picture at all. But
00:32:26
what if the AI and robotics revolution
00:32:28
plays out in the most optimistic way
00:32:30
over the next decade and is massively
00:32:32
deflationary and we get basically AI
00:32:36
powered robots expanding the economy
00:32:38
massively. You need power for that. I
00:32:40
get that and we need to do power
00:32:41
generation for sure. But my point is
00:32:43
just what if the new technology provides
00:32:46
an answer to the fiscal situation that's
00:32:50
not currently on the table. I think
00:32:51
again. So, this is my point is is
00:32:53
Washington's not going to embrace I
00:32:54
understand, but you're going to have
00:32:55
brownouts and blackouts of average human
00:32:57
homes to do that. No, you shouldn't have
00:32:59
to make these decisions. Those are dumb
00:33:00
trade-offs. You're not going to be able
00:33:01
to solve the power generation problem in
00:33:03
a reconciliation bill. And I want to
00:33:04
just add my two cents at the end here.
00:33:07
Leadership starts at the top. This
00:33:09
president acts unilaterally all the
00:33:11
time. Whether it's immigration, DEI, he
00:33:15
has no problem. And this is what he is
00:33:18
doing. He's endorsing this bill. He's
00:33:22
saying it's big. It's beautiful. He is
00:33:24
saying this is historic. I think this is
00:33:27
bad leadership. I'll just say it
00:33:28
straight out. He was elected to balance
00:33:30
this budget and to have austerity. He's
00:33:32
going to put more onto the debt. And we
00:33:35
sat here on this very podcast, the four
00:33:36
of us, and said this was the most
00:33:38
important issue on the world. And Trump
00:33:40
is putting gasoline on the fire. I think
00:33:44
it's bad leadership. Let's continue on
00:33:47
the docket. I suggest you look at the
00:33:48
Constitution because the power of the
00:33:50
purse rests with Congress. He's
00:33:52
endorsing this. He's endorsing it. He's
00:33:54
not even putting up a fight. When it
00:33:56
comes to immigration, these other areas,
00:33:58
the president has far more unilateral
00:34:00
power and he's used that every chance he
00:34:01
gets. But when it comes to spending, you
00:34:04
have to get Congress on board. This is a
00:34:06
bill that passed with a one vote margin.
00:34:08
I don't know where you think the
00:34:10
opportunity is to squeeze out more
00:34:12
concessions from Congress. I don't think
00:34:14
he could speak up. He could speak up and
00:34:16
say do better. He's doing the opposite.
00:34:18
He's doing the opposite. This bill
00:34:20
passed by a one vote margin. It was
00:34:21
hanging on by a thread. And if it
00:34:23
failed, you get basically hundreds of
00:34:26
billions of tax increases, which we
00:34:28
don't need right now. And you know, in
00:34:30
terms of in terms of I think he's trying
00:34:32
in terms of Republican prospects. Look,
00:34:35
people vote for Republicans to do two
00:34:37
things. Cut taxes and cut spending. I
00:34:40
wish that this bill did two out of two.
00:34:42
It does one out of two with some of the
00:34:45
second one like with the Medicaid stuff.
00:34:47
If we didn't at least cut taxes, there'd
00:34:50
be no reason to vote for Republicans.
00:34:52
None. You all are much closer to him
00:34:53
than I am. You all should be challenging
00:34:55
him to not endorse this and to do the
00:34:57
opposite. That would be the profile and
00:34:59
courage. Then we get the biggest
00:35:00
increase in decades. Great. No, he
00:35:03
should just go to everybody and say do
00:35:04
better. I don't agree with that tactic.
00:35:05
I think the better tactic is to do what
00:35:07
he did, but then work behind the scenes
00:35:09
and help the Senate implement the
00:35:11
technical guard rails that make the bill
00:35:14
better than it as is. Look, nobody
00:35:16
thinks that this bill is going to pass
00:35:17
as is. It's going to go through its own
00:35:20
set of changes in the Senate. Now, I
00:35:22
think the opportunity is to understand
00:35:24
these puts and takes with a little bit
00:35:25
more detail and a little bit more
00:35:27
mathematical precision and fix it. So, I
00:35:30
still hold out hope that we're going to
00:35:31
get to a better place. But he's going to
00:35:33
have much more influence, Jason, by
00:35:35
actually taking the positive path than
00:35:36
throwing everybody under the bus. The
00:35:38
real question will be in 30 days or 45
00:35:41
days, what is the final version of this
00:35:43
thing? And whatever gets passed and
00:35:46
signed by him, there will have been
00:35:48
enough time to understand it. And you'll
00:35:50
see the real-time reactions. All I'm
00:35:52
trying to indicate is I think that this
00:35:56
was a moment where people were very
00:35:58
pensively optimistic that the approach
00:36:01
of Doge would really be embraced and
00:36:03
that we'd see some broadscale attempts
00:36:05
to at least and you know Freebrook said
00:36:07
this just to go back to 2019 even if we
00:36:10
don't do anything else just go back to
00:36:11
2019 pretend COVID didn't happen. Yeah.
00:36:14
And we would be, by the way, if we did
00:36:15
that, we would have a budget surplus and
00:36:18
the bonds would be trading down to 2%
00:36:21
yields again. I mean, we would be in
00:36:22
such an incredible condition. We've got
00:36:23
2% inflation right now. Anyway, we don't
00:36:25
need to beat the dead horse. Let's move
00:36:27
on. I I wish I wish there were 218
00:36:30
members of the House who thought like I
00:36:31
hear you. Oh gosh. Listen, this guy is
00:36:34
saying he's going to primary anybody
00:36:35
who, you know,
00:36:37
doesn't he should be speaking up right
00:36:40
now. I'll tell you what, give him the
00:36:41
line item veto and then let's talk,
00:36:43
right? How about you saying guys spend
00:36:45
less money? He doesn't have to throw him
00:36:47
under the bus, Jimoth. It could be
00:36:49
something very subtle. But you make it
00:36:50
that way. What you the way that you
00:36:52
position it is like you make it such a
00:36:54
black and white thing and it's I speak
00:36:55
up is is black and white. He's there's a
00:36:57
lot going on bill. No, but what I'm
00:36:59
saying is he
00:37:01
is I'm guessing I don't know, but I'm
00:37:04
going to assume working behind the
00:37:06
scenes. And what you're saying is he
00:37:08
should virtue signal and prognosticate
00:37:09
in public. And I'm saying that I I never
00:37:12
said virtue signal. That's what you're
00:37:14
asking him to do. No, I'm not. I'm
00:37:16
asking him to say reduce spending in the
00:37:18
bill because that's good for America.
00:37:21
This bill cuts 880 billion for Medicaid
00:37:23
over a decade, which is something that
00:37:26
already is politically tough and
00:37:28
controversial. It imposes work
00:37:29
requirements for able-bodied adults.
00:37:32
This is similar to what Bill Clinton did
00:37:34
back in 1996 with welfare reform.
00:37:36
Basically saying that you can't be a
00:37:37
layabout and get welfare. So these are
00:37:40
relatively tough things to do
00:37:42
politically. It's like do I want to see
00:37:44
even more cuts? Yeah, absolutely. But
00:37:46
again, when you have a one vote margin,
00:37:48
whose vote are you going to pick up by
00:37:49
doing more cuts? I guess you get Thomas
00:37:51
Massie maybe. But you probably have to
00:37:52
go all the way to balance the budget to
00:37:54
get his vote. I don't know what it takes
00:37:55
to get a Thomas Massie vote. So, I don't
00:37:58
know what votes you think you're going
00:37:59
to pick up by cutting even more. And we
00:38:02
can't afford to lose even one vote. It's
00:38:04
literally the art of the game. We got to
00:38:05
move on. I will move on. But if you're
00:38:07
going to tell me it's I'm asking virtual
00:38:08
signal, you're going to lie about my
00:38:09
position, I will correct it. It's been
00:38:11
40 minutes. So, my position is not to
00:38:12
virtual signal. It's to go flip a couple
00:38:14
people and say, "Do better." Okay. It
00:38:16
was a huge week for AI again. Google and
00:38:19
the ghost of Steve Jobs took center
00:38:21
stage. Let's start with Google here.
00:38:22
They had their IO conference on Tuesday.
00:38:25
This is where they show a bunch of new
00:38:27
stuff. They get developers to come
00:38:29
together to embrace their products and
00:38:32
the stock ripped 5% on the day which
00:38:35
might be a turning point for Google and
00:38:38
up again today. Search was the big
00:38:40
announcement and we had a discussion
00:38:42
with Sergey about that in Miami and we
00:38:44
had multiple discussions about it here
00:38:46
over the last couple of months. They
00:38:49
demoed something called AI mode. No, not
00:38:51
founder mode, AI mode and they compared
00:38:53
it to regular search. Here it is on the
00:38:55
screen if you're watching us on YouTube
00:38:56
or Spotify. A very elegant app. And so
00:39:00
here you have somebody searching and
00:39:02
what you can see is in AI mode. It looks
00:39:04
like a nice comprehensive search.
00:39:06
Instead of 10 blue links and tons of
00:39:09
advertising. So you can flip over and
00:39:12
see all or you can do a perplexity like
00:39:14
search. Here we're showing you what this
00:39:16
search would have looked like. The AI
00:39:17
summary at the top and then your 10 blue
00:39:19
links. It's distinctly different. It is
00:39:22
exactly Chimath what you talked about.
00:39:23
Somebody has to have the courage to flip
00:39:25
the switch here. And so that's was the
00:39:29
big drop. And I think that in your
00:39:31
interview Dave, didn't Sundar say we're
00:39:33
going to integrate ads into that kind of
00:39:35
result? Yes. The comprehensive search
00:39:37
result. I think there were a couple of
00:39:39
things that Sundar said in the interview
00:39:41
that showed up in a really important way
00:39:44
at IO. The first was exactly this which
00:39:46
is to make AI mode more ubiquitous in
00:39:48
search and effectively over time replace
00:39:51
search with this AI mode experience
00:39:53
which they've been testing in a small
00:39:54
group and now they've expanded very
00:39:57
specifically to kind of Chimath's point
00:39:59
a couple weeks ago they should try and
00:40:00
flip the switch. Well, it appears
00:40:02
they've done that. The key question and
00:40:03
the challenge has always been what's the
00:40:05
revenue per query? How are you going to
00:40:06
make money? And I think one of the other
00:40:09
kind of interesting announcements that
00:40:11
we saw come out of IO which indicates
00:40:13
the business model opportunity here is
00:40:16
not just AI mode and search but some of
00:40:18
the other tools that they launched
00:40:20
bundling them together and they have
00:40:22
this product offering called AI ultra
00:40:24
for $250 a month that includes you know
00:40:26
YouTube premium. It includes 30
00:40:28
terabytes of storage on your Google
00:40:30
account. It includes access to flow
00:40:32
which is their movie creating AI model
00:40:35
where you can use they have V3, Gemini
00:40:37
and imagen below it. Those are the three
00:40:39
models that contribute to this movie
00:40:41
creation studio tool and they've
00:40:44
launched a number of other kind of highv
00:40:46
value Gemini models as standalone
00:40:48
research apps and you can get access to
00:40:50
all of these kind of high-owered tools
00:40:52
for $250. So I do think this sets a new
00:40:55
direction that Google is likely testing
00:40:57
as a business model which is
00:40:59
subscriptions you're saying. Can we make
00:41:00
yeah a very high dollar volume
00:41:02
subscription model for consumers that
00:41:04
over time could create a very meaningful
00:41:07
shift in the revenue mix for Google away
00:41:10
from ads and more towards subscription
00:41:12
revenue that we already see in some of
00:41:14
the consumer services like YouTube and
00:41:16
YouTube TV. This is I think a really
00:41:19
important turning point. I would say if
00:41:20
anyone were to identify the week that
00:41:22
Google really pivoted into the AI
00:41:24
business model it might be this week.
00:41:26
They launched over 15 products at this
00:41:28
thing. And another important one I'll
00:41:29
highlight, Sundar talked about this in
00:41:31
the interview uh I did with him. In the
00:41:33
early days of Google from 2002 till
00:41:35
about 2011, Google had this kind of
00:41:38
playground of products called Google
00:41:40
Labs where they would introduce new
00:41:41
stuff. I don't know if you guys remember
00:41:43
checking out new stuff. So they
00:41:44
relaunched Google Labs like a year ago
00:41:46
or so and as Sundar said they're putting
00:41:50
a lot more stuff in labs. So Labs is now
00:41:52
becoming the new test bed and they
00:41:54
launched a bunch of the announcements in
00:41:56
labs. So that becomes the place where
00:41:58
they'll say what's the business model,
00:42:00
what's the use case, do people like it,
00:42:01
do they love it? If they do, it
00:42:02
graduates out of labs into full
00:42:04
production. And I think that opens up
00:42:06
the opportunity for some of the things
00:42:07
like Chimath was talking about where
00:42:09
they could experiment new ideas, new
00:42:10
modalities, and then productize them if
00:42:12
they work. And they discontinued labs in
00:42:14
2011. They just brought it back last
00:42:15
year. They just brought it back. Yeah.
00:42:17
Saxs, you I think mentioned, hey, they
00:42:19
have the I'm feeling lucky button here.
00:42:20
They put the AI search up in the top
00:42:23
bar. We have images, video, shopping,
00:42:26
etc. news. We're all familiar with that.
00:42:28
What do you think? Just swap out I'm
00:42:29
feeling lucky for Delight Me with AI and
00:42:32
and M or maybe just if you were a
00:42:34
product manager there because you've
00:42:35
done a lot of great product work in your
00:42:37
career before you were in politics,
00:42:39
would you just AB test this and say
00:42:41
anybody who let's say doesn't click on
00:42:44
ads, let's send them to the AI search
00:42:46
and just see how they do because they
00:42:47
don't click on ads anyway and then just
00:42:48
let the ad clickers keep going to the ad
00:42:51
product.
00:42:52
Well, I think you know we were all
00:42:54
calling on previous podcasts for Google
00:42:56
to risk disrupting their dominance in
00:43:00
search by moving to more of an AI based
00:43:03
model. And I think they've taken an
00:43:05
important step in that direction.
00:43:06
They've sort of threaded the needle here
00:43:09
between keeping their old UI and product
00:43:12
and then also incorporating the the new
00:43:14
sort of AI UI. So, this is a compromise.
00:43:18
I mean, they're trying to have their
00:43:19
cake and eat it too. It ultimately feels
00:43:21
a little bit like a transitional move. I
00:43:22
doubt that this will be the end point,
00:43:25
but given their need to protect their
00:43:28
search business while also developing
00:43:30
their AI business, this feels like a
00:43:33
pretty good compromise, I think. I mean,
00:43:35
it certainly shows that they're in the
00:43:36
game and they're not going to get caught
00:43:39
totally flatfooted and let the whole
00:43:40
world disrupt them while they're trying
00:43:42
to figure out AI. I mean, they are
00:43:43
responding now and the market seemed to
00:43:47
like it. I think they were up like 5% on
00:43:49
this news. Yeah. And then up again
00:43:50
today. Chimatha, you talked about having
00:43:53
some uh design sense and maybe some
00:43:56
bravery here. Looks like uh they're
00:43:59
pretty much on the cusp of doing that.
00:44:01
The search alerts results in AI look
00:44:04
elegant. They look uh dare I say
00:44:05
perplexity like when you look at them. I
00:44:08
would
00:44:09
say I would like to have that as my
00:44:12
default. I think when I'm doing search,
00:44:15
what do you think of the product? We're
00:44:16
showing it here again. And what do you
00:44:18
think the road map will be? I gave one
00:44:21
possibility. Sax is totally right. I
00:44:24
think that this is not the destination,
00:44:26
but they've taken a really important
00:44:28
step and now they have to follow
00:44:30
through. We all know what has to happen.
00:44:33
So, I think we're all just going to
00:44:35
debate when it happens and I think what
00:44:37
the market is betting is that it's going
00:44:39
to happen in the next 18 months. So what
00:44:41
is it exactly? It's when AI mode becomes
00:44:44
the default for a large swath of
00:44:46
existing Google users. How will we know
00:44:49
that happens? They're probably running
00:44:50
AB tests right now. They're probably
00:44:52
gauging behavioral patterns of which
00:44:54
kinds of users, what the actual impacts
00:44:57
to search are, and what impacts exist to
00:45:00
CPCs to cost per click, which is one of
00:45:02
the inventory types that they sell. But
00:45:04
I think this is the beginning of a
00:45:06
process. I suspect it'll be done in less
00:45:08
than a year. And for a large swath of
00:45:11
users, they are going to put AI in the
00:45:14
front door. I think that's a feta
00:45:15
compli. I think we're just now debating
00:45:17
the mechanics of getting there per how
00:45:19
Sax alluded to it. That's a really
00:45:21
powerful step. But here's Jason. I think
00:45:24
I go back to what they also have to keep
00:45:27
in mind. What you don't want to have
00:45:31
when you
00:45:32
are going through an innovator's dilemma
00:45:36
is to have the competitor be on to the
00:45:39
next lily pad while you are figuring out
00:45:42
the current lily pad. And the reason I
00:45:44
bring this up, and I'm sure we'll talk
00:45:46
about this, is it's impressive what Sam
00:45:50
Alman is doing right now at OpenAI. I
00:45:53
don't know if it was a troll or not, but
00:45:55
to buy IO, Johnny Ives firm at the end
00:45:58
of Google IO. And I read an article that
00:46:01
said OpenAI has been trying to do this
00:46:03
for a couple years in a row. But what
00:46:05
Sam alluded to in that video was some
00:46:08
nextgen whisbang device that they've
00:46:10
been working on. That was the industrial
00:46:13
logic for the M&A.
00:46:16
I think what you don't want to have
00:46:18
happen if you're Google is to finally
00:46:20
get the search thing right and then now
00:46:24
have to play catchup on some device. So
00:46:27
I would encourage them to just run the
00:46:28
AB tests and shorten the window of
00:46:32
measurement. I think we all know what
00:46:34
has to happen. So, it's probably better
00:46:37
to just do it sooner than later and then
00:46:39
start to allocate the incremental
00:46:41
resources to these new form factors and
00:46:44
other things so that they shorten the
00:46:47
distance between them and open AI. Let
00:46:49
me tee up exactly what happened. OpenAI,
00:46:52
they're the makers of Chat GPT, bought
00:46:55
Johnny IV's startup. The startup is
00:46:57
called IO. They bought it for 6.5
00:46:59
billion in an all stock deal. If you
00:47:01
remember, Johnny IV was Steve Jobs
00:47:03
collaborator and designer. iPhone, iPod,
00:47:06
iPad, all that beautiful stuff. So,
00:47:09
OpenAI is buying IO. IO is not loved
00:47:12
from, if you've heard about that. That's
00:47:13
his design agency. And so, they did a a
00:47:16
really interesting romcom video. I don't
00:47:17
know if you guys saw this. It's
00:47:19
incredibly ridiculous. Let's play it.
00:47:22
Here are two people having a meat cute.
00:47:25
Dude, you're right. It's like a love
00:47:27
story. It literally is like a meat cute.
00:47:29
Here comes Sam. He's walking around San
00:47:32
Francisco and he's a genius according to
00:47:34
this video. Here's what I have said
00:47:37
about Johnny. I was like, I've never met
00:47:39
a man who's so brilliant and so humble.
00:47:41
I mean, it's embarrassing the language.
00:47:43
The responsibility that Sam bears is
00:47:45
actually honestly beyond my
00:47:47
comprehension. He's a co-founder of IO.
00:47:50
Yeah. Open AAI already owns 20% of it.
00:47:51
Yes. No, no, 23%.
00:47:55
As all Sam Alman deals, if you learned
00:47:57
anything, they're conflicted.
00:48:00
But the the copy here I just want to get
00:48:01
your general the gentleman's comment on
00:48:04
what I see you worrying about are other
00:48:06
people or about customers about society
00:48:09
about culture and to me that tells me
00:48:11
everything I want to know about
00:48:12
surprised you guys are such haters on
00:48:14
this I'm excited to see what this thing
00:48:15
is but I just thought this was like
00:48:18
you're falling for
00:48:20
it the
00:48:23
deepest I don't know why we're even
00:48:25
watch but they're going to watch the
00:48:27
9minute video I wanted my 9 minutes back
00:48:29
Nothing happened. I want the device.
00:48:31
They say they have a device. Okay, I'm
00:48:33
giving them the benefit of the doubt.
00:48:35
They've invented something that they say
00:48:37
is world changing. Okay, but they this
00:48:40
is again I'm just using their words and
00:48:42
he's sipping a macchiato there. So
00:48:44
that's important. Oh my god, you are so
00:48:46
falling for the espresso video. This is
00:48:48
your problem. Jam, look at the glasses.
00:48:50
I think the clue is in the video. Look
00:48:52
at Johnny's glasses. The fact that we're
00:48:54
even talking about this such a problem.
00:48:56
Let me let me tell you my 2007 my 2007
00:49:00
Apple iPhone moment. Steve announced the
00:49:03
iPhone. I got one of the I don't know
00:49:06
first like few hundred few thousand. I
00:49:09
was going to Vegas. So I got it on a
00:49:11
Friday and I had it in my hand. The way
00:49:15
that people looked at me for those three
00:49:16
days. I've never been looked like at
00:49:18
that ever or since. They liked you.
00:49:21
People like what you were feeling was
00:49:23
love and admiration. All I had All I had
00:49:25
was music. There was no multiple
00:49:26
screens. You couldn't buy app stores.
00:49:28
Yeah. But it was so mesmerizing. And my
00:49:32
point is, you're right. I may have
00:49:34
fallen for it, which may seem crazy, but
00:49:36
when I saw that video, I thought, hey,
00:49:38
hold on. All I'm saying is if there is a
00:49:40
next generation device that they or
00:49:42
somebody else, my my only comment is I
00:49:45
think Google is so wonderful. I love the
00:49:47
people there. I just want them to be on
00:49:49
the forward foot and I just don't want
00:49:51
them to be reacting for the next five
00:49:52
years. That's all I'm Here's what Sam
00:49:54
said about Johnny. Johnny is the deepest
00:49:56
thinker of anyone I've ever met. I mean,
00:49:58
this was a love story. I'm not
00:50:01
justifying what they said about each
00:50:02
other. Okay, let's handicap it here.
00:50:04
I'll tell you exactly what's happening.
00:50:05
It's the glasses. He's wearing the
00:50:07
glasses in there. They're making smart
00:50:08
glasses. Sachs, what are they making?
00:50:10
What are these two guys up to? You know
00:50:12
these guys. You know both of them for 20
00:50:14
years. Come on now. You know what's
00:50:17
going on. Just spill the beans. Let me
00:50:19
spill it. Let me express a hot take
00:50:21
here. But let me express in the form of
00:50:23
a question cuz I don't want to I don't
00:50:25
want to be too definitive with the
00:50:27
rhetoric. Okay, go which is is Johnny IV
00:50:31
just a media creation?
00:50:34
I mean no seriously just did Steve Jobs
00:50:37
manifest him? Well, I know he did good
00:50:40
work at Apple, okay, don't get me wrong,
00:50:41
but the software was done by someone
00:50:45
else, right? So, he did like the
00:50:46
beautiful Eddie. So, you know, the whole
00:50:49
and Scott industrial design. Yes, the
00:50:52
hardware looks beautiful with a beveled
00:50:53
edge to paint a glass with a beveled
00:50:56
edge. Okay, great. So, he has this
00:51:00
beautiful English accent which everyone
00:51:02
associates with intelligence, right? So
00:51:03
he could describe anything and it's
00:51:05
going to sound amazing and taste and
00:51:07
taste 15% smarter, right? So I don't
00:51:12
know. I mean, maybe he's like parlayed
00:51:14
this incredible reputation that he's
00:51:16
built in the media into 2% of open AI.
00:51:21
Open AI. It's incredible. I mean, what a
00:51:22
trade. What a trade. Do you think it was
00:51:25
done at that $300 billion price? You
00:51:27
think it was done at like a cheaper
00:51:28
price? Yes. Yes. Yeah. I mean, there is
00:51:31
an argument. I'm going to go in Sax's
00:51:33
corner here. I'm going to go into
00:51:34
Conspiracy Corner here. I'm going to say
00:51:37
Dita Rams designed the iPhone, not
00:51:39
Johnny IV. Pull it up, Nick. Well, I'm
00:51:41
going to say Tony Fidel designed the
00:51:42
iPhone. Okay. But if you don't know who
00:51:44
Dita Rams is, I mean, if you look at the
00:51:47
transistor radio, made the iPod. If you
00:51:49
pull up the U Braraw radio, here is the
00:51:52
power Mac. If you look at the iMac,
00:51:55
basically what did Steve Jobs Are you
00:51:57
accusing Johnny IV of ripping off a
00:51:59
German industrial designer? What I'm
00:52:01
saying is immature artists copy,
00:52:03
immature artist steal. I am saying 90%
00:52:07
of what Apple did was steal Dita Rams
00:52:09
design. I will die on that hill. 90%
00:52:12
stolen. Johnny IV has been kned. He's
00:52:14
Sir Johnny. He took Der Rams' nighthood.
00:52:18
That's what he did. Sir Johnny. I mean,
00:52:21
Sir John, he did put the handle on an
00:52:24
iMac, right? I mean, so you got that.
00:52:26
Stop. He did. You can move those iMacs
00:52:28
around. So slanderous. Stop you guys.
00:52:31
This is so brutal. So great, dude. Look
00:52:34
at this one. Pull this one up, Nick.
00:52:36
Everybody knows the iconic iMac. I hate
00:52:38
to break your hearts, folks. Totally
00:52:39
ripped off from deer. It was all
00:52:41
inspired by the Germans. Sorry. So
00:52:44
you're saying deer brawn is like the
00:52:46
Kaiser So of our entire physical world.
00:52:48
Look, here's the speaker on the left.
00:52:50
Look at the iMac on the right. These
00:52:52
guys took every design that deer did in
00:52:55
the 60s and 70s, even the 50s, and they
00:52:58
photocopied it and they put an Apple
00:53:00
logo on it. I couldn't Stop. Stop. Stop.
00:53:03
Stop. Sax, am I wrong? Stop. I didn't
00:53:06
know about the Der Braun angle.
00:53:09
I mean, listen, it is what it is. I
00:53:12
mean, look, everyone at Apple did
00:53:13
amazing work. There's no question about
00:53:15
it. But but I but I'm just saying like I
00:53:18
don't quite get this like design god
00:53:22
reputation that's been created a little
00:53:23
bit. But but I would say that it's
00:53:25
certainly a brilliant idea to parlay
00:53:28
that into 2%. By the way, is that what
00:53:30
San Francisco looks like? Like I never
00:53:32
thought every time I go to San Francisco
00:53:34
it's so disgusting and it's cloudy and
00:53:36
cold. But yes, did they find the one day
00:53:38
where like the sun was shining in that
00:53:40
hellscape? Yeah, there were no homeless
00:53:43
people and there was no fentinel. I
00:53:44
don't know what happened. Yuck. On a
00:53:46
serious note, what I would say is maybe
00:53:48
what this deal indicates is, I mean,
00:53:50
obviously, is that OpenAI is going to do
00:53:52
something with hardware. Sure. It could
00:53:54
be a new kind of phone that is AI first
00:53:56
as opposed to the wall of applications.
00:53:59
Or maybe it's like some sort of like
00:54:01
pendant thing that like tracks what's
00:54:03
going on around you and feeds it into AI
00:54:05
so it can give you more context. You
00:54:07
know, maybe like in the movie Her. Who
00:54:09
knows? But it seems like they'll do
00:54:10
something with devices in the physical
00:54:14
world.
00:54:15
So that would be the play, I guess. I I
00:54:18
guess the question you have to ask is,
00:54:19
will it
00:54:21
make Open AI 1% more valuable? If it's a
00:54:24
$300 billion company, 2% more valuable.
00:54:26
Yes, thank you. Will it make it more
00:54:29
than 2% more valuable? If it makes it 1%
00:54:31
more valuable, it was worth the the
00:54:33
swing, I guess. Right. I heard an
00:54:35
interesting factoid yesterday which is
00:54:38
by somebody who was looking at OpenAI
00:54:42
equity. All of the rounds before the
00:54:44
$300 billion round from SoftBank had
00:54:47
essentially a feature where there was
00:54:49
like a hard rate of return and the
00:54:51
ability tax or something. No, but the
00:54:53
ability to put the stock back if they
00:54:55
weren't able to do the conversion to a
00:54:57
for-profit company, etc., etc. And in
00:55:00
the $300 billion round, apparently it's
00:55:04
much cleaner and it's essentially like
00:55:06
sub new co that is owned. And I thought
00:55:10
to myself like, wow, at at those
00:55:13
valuations, I would have expected the
00:55:15
terms to ratchet up. And instead, it
00:55:17
seems like the terms have ratcheted
00:55:18
down, which is
00:55:21
I don't It's either it's either
00:55:22
commentary on the negotiating prowess of
00:55:25
OpenAI and their team and or the
00:55:27
riskmanagement philosophy of SoftBank
00:55:30
and the lack
00:55:31
thereof. I do think you know he is a
00:55:34
great designer and he will make it one
00:55:36
or two% better. Design matters. So if he
00:55:38
makes it I mean there's a chance right
00:55:40
Sachs that he could make the product
00:55:41
five or 10% better through design. I
00:55:43
don't think it's going to be one or two%
00:55:44
better. I think it's going to be either
00:55:46
zero or a doubling of the company's
00:55:48
value. Something like that. Right. like
00:55:50
either he develops an iPhone like
00:55:52
amazing hit product no one's seen before
00:55:54
some new consumer device that you know
00:55:57
really could take the company to another
00:55:58
level or probably they do something that
00:56:01
kind of doesn't work. So it's more of
00:56:03
like a 2% option against a strike of
00:56:06
doubling the value of the company. I'm
00:56:07
surprised it hasn't leaked because in
00:56:09
the in Sam's update or whatever he said
00:56:12
that the that the executive team was
00:56:14
shown the device and that he has it.
00:56:17
Who? So what is it? It's it's definitely
00:56:20
going to be a pendant or glasses. It's
00:56:22
got to be those two form factors. He was
00:56:25
the investor in Humane. Remember, wasn't
00:56:27
that the pendant from the Apple people?
00:56:28
Yeah. I will take the field against
00:56:29
that. What are the odds and how much
00:56:31
money do you want to bet? Okay, here we
00:56:33
go. So I think it's a a pendant like a
00:56:36
wearable. We'd have to define that. So
00:56:38
like humane. I'll say it's not a
00:56:40
pendant. Okay. Glasses. I get What are
00:56:41
you going to give me? You give me odds.
00:56:44
I'll give you Give me odds. Well, we'll
00:56:46
just do even money. Let's do even money.
00:56:48
Even money.
00:56:50
I don't know. Whatever you want. How
00:56:51
However many thousands of dollars you
00:56:53
want to bet. Even money. How about you
00:56:54
give me two to one 10
00:56:57
dimes? Cuz you have every option. I have
00:57:00
to pick them. No. No. No. That's a
00:57:03
pretty good bet. This is my You want to
00:57:05
get in on the bet here? All right, let's
00:57:08
move on. All right. We covered Trump's
00:57:10
trip to the Middle East last week, but
00:57:12
our bestie Sachs was there as well. Um
00:57:15
and uh you did uh you got a couple
00:57:17
photos to show and uh you met a lot of
00:57:20
friends there. Tell us about the
00:57:22
experience of did you go first of all
00:57:25
did you go on Air Force One? That's what
00:57:26
everybody wants to know. Or did you go
00:57:28
on Airsacks? I took my own plane. So I
00:57:31
know that sounds a little bit like a
00:57:32
flex but
00:57:34
were you were you in formation with Air
00:57:36
Force One? Were you like you know just
00:57:39
right behind it or on your own travel?
00:57:41
You were on your own travel schedule. I
00:57:43
was I was doing my own my own thing. But
00:57:45
did you have the option to go on Air
00:57:46
Force One? I probably could have if I
00:57:48
wanted to, but I I actually wanted to
00:57:49
get to the region a little bit early cuz
00:57:50
I hadn't been before and I kind of
00:57:52
checked it out and so I was waiting
00:57:54
there when the president arrived. Oh
00:57:55
wow. So it was your first time in the
00:57:57
region. Yeah. Well, tell us about your
00:57:59
impressions generally and uh what were
00:58:01
you trying to accomplish there as the
00:58:03
ZAR? Well, I call it AI diplomacy. First
00:58:05
of all, these countries are obviously
00:58:08
they're very resourcerich. They've got a
00:58:10
lot of capital to deploy and they're
00:58:14
also very interested in diversifying
00:58:15
their economies. They're very interested
00:58:16
in high-tech and they want to do things
00:58:18
in AI. They've got very significant
00:58:21
aspirations there and so I was over
00:58:23
there just to kind of listen and learn
00:58:25
and see what they're interested in doing
00:58:27
and I also got to see some of their tech
00:58:29
scenes and just Oh, you went to Digital
00:58:32
Garage in Riad, right? I've been there.
00:58:34
Oh, you've been there? Okay. Yeah. Yeah.
00:58:35
No, no, it's um they are really pushing
00:58:37
entrepreneurship, trying to generate
00:58:39
more ideas and teach people how to start
00:58:41
companies and co-working spaces, all
00:58:44
kinds of incentives in terms of visas to
00:58:47
come to Riad, to come to UAE, you know,
00:58:49
Dubai, Abu Dhabi specifically. So, yeah.
00:58:51
Right. I'm curious your impression of
00:58:53
the individuals who are leading a lot of
00:58:56
these different groups. I was really
00:58:59
taken back by how much time they had
00:59:02
spent in the West. It seemed like every
00:59:04
single person I had met had taken one of
00:59:06
these scholarships to go to Oxford to
00:59:09
Harvard to you know Caltech whatever it
00:59:11
was they all had spent massive amounts
00:59:13
of time in the west and then come back
00:59:15
to either Saudi or UAE or Oman etc. So
00:59:19
maybe you could talk a little bit about
00:59:20
the people and their knowledge base
00:59:23
motivation etc. I had the same
00:59:25
observation. The elites of all these
00:59:27
countries have all been educated in the
00:59:29
west, usually in the United States, some
00:59:30
in the UK. The leadership tends to be
00:59:34
young, visionary, future oriented,
00:59:36
intensely interested in AI, like I was
00:59:39
saying, and they want to do big things
00:59:41
in AI. The thing that that's been in
00:59:44
their way is that in October 2023, the
00:59:48
Biden administration basically put a a
00:59:50
blanket on the whole region acquiring
00:59:53
semiconductors, GPUs, and it it required
00:59:57
that every export of a GPU or a server
01:00:02
that contains a GPU had to get a
01:00:04
specific license from the commerce
01:00:06
department. And it put a major damper on
01:00:08
their efforts to do things like build
01:00:10
data centers in the region.
01:00:12
or to have their own local AI efforts.
01:00:14
So, they've been kind of in this holding
01:00:16
pattern where the Biden administration
01:00:18
decided to kind of alienate them. And,
01:00:20
you know, it wasn't just AI, there was
01:00:22
other things, too. Remember the fist
01:00:23
bump? And the Biden administration was
01:00:25
very, I'd say, hostile, standoffish,
01:00:29
cool to, and even hostile towards these
01:00:32
states in the region. And I think from a
01:00:34
geopolitical standpoint, it was just
01:00:35
stupid because we're in an intense
01:00:38
competition, a high-tech competition, a
01:00:40
security competition, economic
01:00:41
competition with China. And these states
01:00:44
want to be aligned with the US and we
01:00:46
are pushing them into China's arms. And
01:00:48
if we don't give them the ability to buy
01:00:50
the American tech stack, it's not like
01:00:52
they're going to sit on their hands and
01:00:54
do nothing. They're going to be forced
01:00:56
to buy the Chinese tech stack. They're
01:00:58
going to participate one way or the
01:00:59
other, right? And China still has some
01:01:01
limitations on how many chips it can
01:01:03
produce, but they clearly are building
01:01:05
their own Chinese tech stack. It's a
01:01:07
Huawei plus Deepseek tech stack. There's
01:01:10
actually a story in Malaysia just in the
01:01:12
last couple of days where they were
01:01:14
talking about building a local data
01:01:16
center using the Huawei Ascend GPU plus
01:01:20
Deep Seek. And then they walk back that
01:01:22
story because I think they're afraid of
01:01:24
getting in trouble with with the US
01:01:25
government. But the point is that if we
01:01:28
drive these countries away, if we
01:01:30
alienate them, they will end up
01:01:32
partnering with China instead of the
01:01:33
United States. I don't understand how
01:01:34
that's in the United States's interest.
01:01:37
Now, as part of this trip, we rolled out
01:01:40
a new framework for an AI acceleration
01:01:43
partnership with these countries. And
01:01:45
there's a few very important terms that
01:01:47
I think it's worth going into. So, so
01:01:49
first of all, the the purpose of the
01:01:50
framework is to replace the Biden export
01:01:54
control to the region with a framework
01:01:56
that gives them the ability to buy GPUs
01:01:58
and build data centers. Uh, but there's
01:02:01
a few very important provisions. Number
01:02:02
one is there's a matching investment
01:02:04
provision. So, for every dollar of
01:02:07
investment they make in terms of
01:02:08
building out data centers in the region,
01:02:10
they have to invest a dollar in the
01:02:12
United States and building out our AI
01:02:14
infrastructure. So it's a huge win for
01:02:16
the United States in terms of
01:02:17
accelerating our own infrastructure
01:02:19
buildout. So that's point number one.
01:02:21
Number two is even with respect to the
01:02:23
data centers they're going to build in
01:02:25
the region, at least 80% of the chips
01:02:28
have to be owned and operated by
01:02:30
American cloud service providers or
01:02:32
American hyperscalers. So even with
01:02:34
respect to the data centers they're
01:02:36
building, the vast majority of the
01:02:38
compute is going to be run by American
01:02:41
companies. And so I mean this to me is
01:02:44
just a huge was there critique were
01:02:46
people critiquing it like uh you know
01:02:48
this uh expansion I guess Ro Khan has
01:02:50
said like hey we should build more data
01:02:51
centers here that was sort of a light
01:02:53
critique maybe or a
01:02:56
misunderst that that Ro was making and
01:02:59
Ro is a friend but just you know I
01:03:00
respond to him on on X which is we are
01:03:03
building out data centers in the US. We
01:03:05
want to do as much as we can. It's the
01:03:07
Trump administration's policy towards
01:03:09
energy that's making that possible.
01:03:11
We're making it easier to spin up new
01:03:13
power generation and we're making it
01:03:15
easier to permit. Both both those things
01:03:17
were virtually impossible during the
01:03:18
Biden administration. So we are building
01:03:21
more here in the US. In addition to
01:03:23
that, you've got these countries who are
01:03:25
resourcerich and they have their own AI
01:03:27
dreams and aspirations and we can either
01:03:30
partner with them or drive them into the
01:03:31
arms of China. And the deal we've made
01:03:33
is to enable their aspirations while
01:03:35
getting them to fund more AI
01:03:37
infrastructure in the United States and
01:03:39
even the infrastructure they're building
01:03:40
over there. First of all, it's going to
01:03:41
be on an American tech stack, right? We
01:03:43
want it to be Nvidia and AMD and those
01:03:45
types of companies. That improves our
01:03:47
balance of trade. It also ensures that
01:03:50
these data centers across the world are
01:03:52
going to be built on American
01:03:53
technology. We want American technology
01:03:55
to become the standard of course and if
01:03:58
you allow those data centers to be built
01:04:00
on Huawei plus deepseek that will become
01:04:01
the standard. So I think this is
01:04:03
something that people in Silicon Valley
01:04:05
intuitively grasp but people in
01:04:07
Washington don't which is the way that
01:04:08
you win these technology battles is you
01:04:11
create the largest ecosystem right you
01:04:13
want to have the most partners you want
01:04:14
to have the biggest app store you want
01:04:16
to have the API everyone use the most
01:04:18
data power exactly so the point is we
01:04:21
want to involve the whole world on our
01:04:23
tech stack I think export controls on
01:04:25
China that makes sense but we don't want
01:04:27
to reduce our market share in the rest
01:04:29
of the world we want them getting hooked
01:04:31
on our platform platforms, making those
01:04:33
platforms the standard and getting lock
01:04:35
in before China can catch up to us. Sax,
01:04:38
did Jensen just say that he thought
01:04:40
export controls were kind of pointless?
01:04:43
Did he just say that recently? Yeah, he
01:04:45
he's against all export controls. And
01:04:47
you know where I would split the baby
01:04:49
here is that I think that most people in
01:04:52
Washington believe that the export
01:04:54
controls on China make sense because we
01:04:56
simply can't allow our most powerful
01:04:59
semiconductors to be to be used by China
01:05:02
because of the dual use.
01:05:04
Now, where I agree with Jensen is that I
01:05:08
think we should not place undue
01:05:09
restrictions on the rest of the world
01:05:12
using the American tax stack because we
01:05:14
will simply concede the market share to
01:05:16
China. And so I think he's right with
01:05:18
respect to the Middle East. Now the the
01:05:20
other concern that you heard a lot is
01:05:23
around this quote
01:05:25
diversion the idea that somehow the GPUs
01:05:28
or somehow the IP if we allow it to be
01:05:31
used in data centers in the Middle East
01:05:33
will somehow find its way to China and
01:05:36
yeah I mean the expression on your face
01:05:37
Jal shows that makes no sense like these
01:05:40
guys have unlimited resources they have
01:05:42
unlimited ambition that's just a naive
01:05:43
point of view they want these for
01:05:45
themselves they understand the power of
01:05:47
AI they don't want to give it to China
01:05:48
they want it for themselves pounds. They
01:05:50
make trillions of dollars selling oil
01:05:52
and gas. They don't need to smuggle our
01:05:53
GPUs. It doesn't make any sense. The the
01:05:55
other thing is that that people don't
01:05:56
understand, you know, they're called
01:05:57
chips, but really these things are now
01:05:59
the size of mainframes. The NVIDIA data
01:06:02
center product, NVL72, is this giant
01:06:04
server cabinet. It's 8 ft tall. It
01:06:06
weighs 3600 lb. It's not like diamonds
01:06:09
you smuggle in a briefcase, right? All
01:06:11
you have to do is send You're not
01:06:13
putting them up and getting them across
01:06:15
the border. That's definitely not
01:06:16
happening. No, it's not happening. All
01:06:19
you have to do is send an inspector to
01:06:21
the data center to count the servers and
01:06:23
you can see that they're there.
01:06:24
Ridiculous. So this idea that somehow
01:06:25
they're going to get smuggled. It's like
01:06:27
this totally fake concern that's been
01:06:29
created. It's almost like a totally fake
01:06:30
narrative, you know. Then people, well,
01:06:33
they won't physically smuggle the
01:06:34
servers, but the IP will somehow be
01:06:36
transferred to China. Look, there's
01:06:38
nothing in a data center that China
01:06:40
hasn't already seen. They have one of at
01:06:42
least one of everything. The issue is
01:06:44
that they can't reverse the advanced
01:06:46
semiconductors just from the chips,
01:06:48
right? Because it's about the process
01:06:50
technology that it takes to create an
01:06:52
advanced semiconductor. There's hundreds
01:06:54
or even thousands of steps in that
01:06:55
process. And you can't figure out the
01:06:57
recipe just by seeing the end product. I
01:06:59
would like to say that I'm very proud of
01:07:01
my bestie. I got three calls.
01:07:04
I do want to tell you, Sax, you crushed
01:07:06
it. I think you Yeah, it was pretty
01:07:08
impressive. put your best foot forward
01:07:10
and people were really impressed. But I
01:07:13
got calls I got three calls from the
01:07:15
powers at be. They were very psyched to
01:07:18
meet you and spend time with you and
01:07:19
they thought you were super impressive.
01:07:21
So it was awesome. Thank you for doing
01:07:23
that. Well, I appreciate that. Yeah,
01:07:25
it's an incredible region. I think
01:07:26
you're doing like a really patriotic
01:07:28
thing. If you look at that region, like
01:07:30
you said, it could tip one way or the
01:07:32
other. They could be involved with the
01:07:34
West. They we could be building
01:07:35
businesses together. And if we build
01:07:37
businesses together, that's better than
01:07:40
them building
01:07:41
businesses with China. Pretty obvious, I
01:07:44
think. I mean, totally. And this is the
01:07:46
thing is I just don't get the opposition
01:07:47
in Washington to this. And I especially
01:07:49
don't get how the opponents of this can
01:07:51
call themselves China hawks because at
01:07:53
the end of the day, this is a tech
01:07:55
battle between the US and China. And we
01:07:57
can have these countries be aligned with
01:07:59
us and in our orbit, or we could push
01:08:01
them into the arms of the Chinese. The
01:08:03
choice is do we want these countries to
01:08:04
be the piggy bank for American AI or for
01:08:07
Chinese AI? It's like say they have an
01:08:10
investment level that they can do that
01:08:13
dwarfs anybody else in the world with
01:08:15
the exception of Norway and the United
01:08:16
States. I mean they can put a lot of
01:08:18
money to work and here's the thing they
01:08:19
want to. It just hit the wire that
01:08:21
OpenAI is partnering on a 5 gawatt data
01:08:24
center cluster in Abu Dhabi with G42.
01:08:26
Well, let's hold on. Let's just be clear
01:08:28
about so the the 5 gawatt cluster or
01:08:30
it's more of a a campus that's going to
01:08:33
support many American hyperscalers and
01:08:36
cloud service providers and they're
01:08:38
going to be one of the tenants. So, I
01:08:39
think they're saying that they're going
01:08:40
to have 1 gawatt. There's still 4
01:08:42
gawatts to go that they plan on
01:08:44
supporting. Well, well, the the the most
01:08:46
interesting part of the deal it said was
01:08:48
that as part of the deal, G42 plans to
01:08:50
make a reciprocal dollar for-dollar
01:08:52
investment in AI infrastructure in the
01:08:54
US. So to your point, Saxs, it's like
01:08:56
perfect. Cooperating like this, I think
01:08:58
just makes a ton of sense. Absolutely.
01:09:01
And this is not dumb money, by the way.
01:09:03
These are very smart people who are
01:09:04
investing very strategically with a very
01:09:07
crisp, you know, group of individuals,
01:09:09
like you said, who have been educated in
01:09:11
the west. You know, when people explain
01:09:13
to me, oh yeah, just go to Abu Dhabi and
01:09:14
come back with a a suitcase full of
01:09:16
money. No, these are folks who want to
01:09:18
be on the board of companies. They want
01:09:19
to company build. They see themselves as
01:09:22
owners in business, not just investors.
01:09:24
And then they want to do it domestically
01:09:26
and abroad. The amount of investment
01:09:28
that they're going to do in the next 30
01:09:30
years is going to dwarf what we see out
01:09:33
of universities and retirement accounts.
01:09:35
I don't people understand the scale of
01:09:37
what's happening. This is why I just
01:09:38
don't understand the point of the Biden
01:09:40
policy of alienating these countries and
01:09:43
making them feel uneasy about their
01:09:46
place because you you don't know who it
01:09:48
was that made the decision. Yeah, I
01:09:51
mean, let's face it, that guy was
01:09:52
confused. We can have Bernies. I've
01:09:55
actually have a pretty good idea at this
01:09:56
point of who was controlling the auto
01:09:58
pen on AI. There was a group of of very
01:10:01
powerful staffers in the Biden
01:10:02
administration, about half a dozen
01:10:04
people who kind of controlled AI policy.
01:10:06
And the interesting thing is none of
01:10:08
them have ever worked in Silicon Valley.
01:10:10
They're all basically lawyers. And they
01:10:12
don't really have a good understanding
01:10:14
of the idea of a platform or a
01:10:15
technology ecosystem or a stack. And I
01:10:18
don't think they realized that by
01:10:22
alienating these very resourcer-rich
01:10:26
countries that they were creating a huge
01:10:28
opportunity for China to come in and
01:10:30
serve the region. And this is why I
01:10:32
think this, you know, again, what I call
01:10:33
AI diplomacy is so important is it boxes
01:10:35
out China. It's good for America. It's
01:10:38
good for our companies. It's good for
01:10:39
our trade balance. It makes us the
01:10:41
standard and it boxes out China. What's
01:10:44
not to like about this? I really don't
01:10:45
get it. I you know I think they were
01:10:47
living in a world where they put people
01:10:49
into very rigid buckets. If you wanted
01:10:52
to look at human rights as an issue,
01:10:54
which I think is what some people were
01:10:56
doing in that administration, I look at
01:10:58
the progress that's being made and if
01:11:00
you go to these societies and you look
01:11:02
at what's happening in that region, the
01:11:04
pace at which human rights and
01:11:07
individual freedoms is growing and the
01:11:09
rate at which those societies are moving
01:11:11
forward, it's going in the right
01:11:12
direction. So what do we want to do? Do
01:11:14
we want to steer people towards
01:11:15
collaboration and working together to
01:11:17
build the future or do you want to have
01:11:19
it devolve? I mean, it's such a stupid
01:11:21
decision by the Biden administration. I
01:11:22
It was perplexing. Yeah, I I really
01:11:24
don't get it. I I don't understand what
01:11:26
the point of alienating these Gulf
01:11:28
States was. Well, I like Trump's
01:11:32
approach there when he said, "Listen,
01:11:33
you're going to make decisions in your
01:11:35
own time." That's what they want to
01:11:37
hear. They want to be respected as they
01:11:39
respect us. Hey, we're going to make
01:11:40
decisions in our own time what's right
01:11:41
for our people. That's a process that
01:11:43
we'll decide. Now, if you want to
01:11:45
collaborate on building a company, a
01:11:47
data center, a technology platform, hey,
01:11:49
we're here for that, but you can't tell
01:11:51
us how to run our society and just like
01:11:53
we wouldn't want them to come here and
01:11:54
tell us how to run our society, right?
01:11:56
So I agree with what you just said Jal
01:11:58
and I thought the best part of the
01:12:00
president's speech which the whole thing
01:12:01
was excellent but the best part was when
01:12:03
he talked about how the neocons and the
01:12:05
interventionists had come into the
01:12:07
region and lectured people and come in
01:12:09
with this very moralistic point of view
01:12:11
and made all these interventions that
01:12:12
basically wrecked countries and it's it
01:12:16
wasn't the the neoconsi interventionists
01:12:18
who have led to so much progress in the
01:12:20
Middle East and in the Arab world it's
01:12:22
been the countries themselves whose own
01:12:25
leaders ers have set the the path and
01:12:27
have had the vision. That was an
01:12:29
incredible speech. It was incredible.
01:12:32
Trump for me. Yeah, he did a great job.
01:12:34
Credit do and that's the message they've
01:12:36
wanted to hear for a long time and
01:12:38
that's the message that moves society
01:12:40
forward. All these other messages just
01:12:42
sort of set it back and um I think I'm
01:12:44
I'm super hopeful about the region.
01:12:46
Okay, let's do Science Corner. you know.
01:12:47
Oh, by the way, I just I was just
01:12:49
checking um X here and it looks like
01:12:52
there is a um announcement on what the
01:12:56
Open AI device is going to be. You guys
01:12:58
see this? Here we go. Oh, the puck. The
01:13:01
little puck. I saw that puck, but it
01:13:03
hasn't been confirmed yet. Oh, it's a
01:13:04
puck you put on your desk and it Yeah,
01:13:06
it's like a puck with a camera and a
01:13:07
hole for a mic. Oh, that's a pin. Yeah,
01:13:09
that's a pendant. It's a pendant. I
01:13:11
could have made 10 times straight up
01:13:12
versus a pendant. That's a puck.
01:13:15
It fits in the palm of your hand. It's
01:13:17
too small to be a puck. A puck's like
01:13:19
the size of your hand. It looks like
01:13:20
your thumb. Okay, this is AI generated.
01:13:23
It's not actually the device. They
01:13:24
haven't shown the device from Johnny IV
01:13:26
yet. Somebody just took what's in the
01:13:28
Wall Street Journal report. They mocked
01:13:30
it up. This is their best guess as to
01:13:31
what it might look like. This is already
01:13:34
out. And by the way, this is going to
01:13:36
get you punched in the face. You show up
01:13:38
with a listening device to a party or a
01:13:40
meeting, I'm gonna knock you out. This
01:13:42
is ridiculous. And it's got a camera in
01:13:44
it, too. I don't want this surveillance
01:13:46
device anywhere near me. How are you
01:13:47
going to do it? Like this?
01:13:52
Yeah. Get that buck away from me.
01:13:56
Yeah. Okay. Anytime you're ready, how
01:13:58
dare you record me and violate my
01:13:59
privacy. I was literally at dinner the
01:14:02
other day. I kid you not. And somebody
01:14:04
came up with one of these things and it
01:14:06
was clipped on his thing. I said, "Hey,
01:14:07
uh, what's the thing on your lapel?" He
01:14:09
said, "It's an AI recording device. It
01:14:11
does. Uh, it summarizes what we're
01:14:12
doing. I said, he goes, "Oh, do you do
01:14:15
you care?" I said, "Do you care if I
01:14:16
punch you in the face for covertly
01:14:18
recording me?" And he said, "Yes." I
01:14:19
said, "Well, then turn it turn it off."
01:14:22
I mean, literally, let's talk about
01:14:24
science. Let's talk about science. I
01:14:25
mean, well, you rais a good point, Jay
01:14:26
Cal, that these um these new devices
01:14:30
No, these new devices
01:14:33
requires a level of trust that is kind
01:14:35
of unprecedented. Yeah, I mean you
01:14:37
already had to trust, you know, Google
01:14:39
and iPhone and so on because they do
01:14:41
have a huge level of access to your
01:14:43
data, but the new let's call them AI
01:14:46
connected devices. I mean they are
01:14:48
surveillance devices. I mean their whole
01:14:50
point is to listen and see everything
01:14:52
that you can hear and see and you know
01:14:56
and then to give that data to AI to do
01:14:58
something with and um man you better
01:15:00
trust that company completely. Oh yeah.
01:15:02
Which company's doing it? Oh right, that
01:15:04
company. Yeah. Hm. The company that had
01:15:07
every leader leave because they didn't
01:15:09
trust the founder. Okay. All right.
01:15:12
Yeah. Go ahead and wear the puck. And by
01:15:13
the way, with these pucks, I'm going to
01:15:14
make one more point, Freeberg. I'm
01:15:15
sorry. These devices need to have a red
01:15:18
flashing diode on them when they're
01:15:19
recording. We need to pass legislation
01:15:21
that it does that. If it's recording, it
01:15:23
should have a flashing red light on it.
01:15:25
Let's go to Science Corner because
01:15:26
everybody Chim loves Science Corner.
01:15:29
Freeberg. Some breaking news, some uh
01:15:32
innovations in Crisper. Yes. It's got a
01:15:34
lot of press coverage, but there was a
01:15:36
baby born named KJ in the research paper
01:15:40
that was published was born
01:15:43
with a mutation that was inherited from
01:15:46
the mother on a specific gene called
01:15:48
CPS-1. And the same mutation inherited
01:15:51
from the father, which resulted in the
01:15:53
fact that this gene has a deficiency in
01:15:55
it.
01:15:57
CPS1 produces a protein that's part of
01:16:00
the ura cycle. So basically breaking
01:16:02
down nitrogen or protein compounds in
01:16:05
the blood so that they can be excreted.
01:16:08
And in the absence of this protein, if
01:16:10
this protein is not being made in the
01:16:11
body, ammonia accumulates in the blood
01:16:14
and that has extraordinarily harmful
01:16:16
consequences including brain damage and
01:16:19
ultimately death. This gene had a G that
01:16:23
was mutated into an A on both of the
01:16:26
copies of the gene meant that the
01:16:29
protein wasn't being correctly produced.
01:16:32
And as a result, the breakdown of
01:16:34
nitrogen and ammonia was not possible.
01:16:38
And so this leads to a very short life
01:16:41
for the very very very very few people
01:16:43
that have ever had this double mutation
01:16:45
that this particular child was
01:16:47
unfortunately born with. So we have
01:16:49
technology today, crisper gene editing
01:16:51
technology. In fact, crisper gene
01:16:53
editing technology is what we use to do
01:16:55
some of the work we do at Oh, the
01:16:57
company I run. And um some advanced
01:17:00
forms of gene editing technology include
01:17:02
what are called base editors where you
01:17:04
can change a single letter on a strand
01:17:06
of DNA into another letter. So in this
01:17:08
particular case, changing an A to a G
01:17:11
was the goal. So the physician Rebecca
01:17:13
Arens Nicholas, giving her a big shout
01:17:15
out for the incredible work she did in
01:17:17
partnership with scientists at the
01:17:18
University of
01:17:19
Pennsylvania, they developed very
01:17:22
quickly a very specific gene editing
01:17:26
target to go in and find the letter A
01:17:29
and turn it to G in this particular
01:17:32
patient's body. Now remember, we have
01:17:35
the same DNA in all of our cells in our
01:17:37
body, but specific cells have specific
01:17:39
functions and genes are expressed in
01:17:42
only those cells. So in this case, it's
01:17:44
the liver that makes this this CPS1
01:17:48
protein. And so they had to get the gene
01:17:50
editing into the liver cells to edit the
01:17:52
liver cells. And if they could edit the
01:17:54
liver cells, then those liver cells
01:17:56
would start to become more functional
01:17:57
and you would have a functional protein.
01:18:00
So they quickly tested dozens of base
01:18:02
editors. These are different proteins
01:18:04
that can go in and edit. They tested
01:18:06
different guide RNAs. These are the
01:18:08
guides that tell that editing protein
01:18:10
where to go, what part of the DNA to go
01:18:12
to, which is trying to get it to this
01:18:13
exact same site. They tested it in petri
01:18:16
dishes. Then they tested it in mice,
01:18:17
then in monkeys to make sure that the
01:18:19
selection that they came up with worked.
01:18:22
They then put it in the child. And so
01:18:24
basically by putting it in the blood
01:18:27
found its way into the liver and that's
01:18:29
the the the the fortunate benefit here
01:18:31
was at the end of the day you're trying
01:18:33
to edit specific cells and we can get to
01:18:35
the liver by putting the crisper and the
01:18:38
guide into the blood. It made its way to
01:18:39
the liver and they coded the gene
01:18:42
editing system with liquid nanop
01:18:44
particles which helped it get into the
01:18:46
cells. It got absorbed by the liver. It
01:18:48
got went into the cells. It found its
01:18:49
way to the site on the DNA. That's
01:18:51
amazing. converted the A to G and as a
01:18:54
result those liver cells became
01:18:56
functional to make the protein that
01:18:58
breaks down the ammonia in the child's
01:19:00
blood. And so they did the first one
01:19:02
just to make sure the child would handle
01:19:03
it. It worked. The child didn't have any
01:19:05
adverse effects. They then did the
01:19:07
second dose, able to take the child off
01:19:09
some of the medicines and compounds they
01:19:11
were giving the child to break down
01:19:12
ammonia in the blood and they gave it
01:19:13
the third dose and they're monitoring.
01:19:16
So this is the first time that we've
01:19:18
seen this kind of custom crisper gene
01:19:20
editing for a genetic mutation being
01:19:22
applied to a patient in vivo in the
01:19:25
blood for a specific target for a
01:19:27
specific particular uh treatment. I just
01:19:30
asked rock how many
01:19:33
diseases are because of an adg point
01:19:36
mutation and it says it's hard to
01:19:38
estimate through the human genome
01:19:41
database or clinvar but it estimated
01:19:44
between 20 and 30,000 conditions are
01:19:47
because of an ad point mutation and
01:19:49
isn't that incredible? Yeah. And I'll
01:19:51
tell you one other story about this. I
01:19:54
saw a geneticist here at UCSF a couple
01:19:57
years ago and he was telling me about
01:19:59
all of these kids that come in and they
01:20:01
don't know what's wrong with them and he
01:20:03
can see on the kid's face. He can see a
01:20:05
disease condition. They have all these
01:20:07
weird lab tests and the kids are dying
01:20:08
or they're really deficient in some way
01:20:11
and they can't afford to do the DNA
01:20:13
sequencing because the child is on
01:20:14
medical or some other sort of like
01:20:16
healthcare plan and it costs like $5,000
01:20:18
to do the DNA sequencing needed to
01:20:20
figure out what genetic mutation this
01:20:23
particular child has. So there aren't
01:20:25
easy offthe-shelf targeted tests or lab
01:20:27
tests you can run to figure out when
01:20:29
people have genetic disease. And so this
01:20:32
is a an incredible like undiscovered set
01:20:35
of problems. Chim off to your point that
01:20:36
there are likely millions of people out
01:20:38
there that have some health condition
01:20:40
that they've not yet been able to
01:20:42
identify because it's a genetic mutation
01:20:44
and not everyone can afford to go out
01:20:45
and just get their genome sequence to
01:20:47
figure out where the mutation might lie
01:20:49
in their body. And so as we discover
01:20:51
more and more genetic mutations, we
01:20:53
build targeted tests for identifying
01:20:54
that people have them and hopefully over
01:20:56
time can start to build more targeted
01:20:58
therapies like this one to go after and
01:21:01
make the changes at specific sites to
01:21:03
fix the the genetic mutations. Bad ass.
01:21:06
It's incredible. Okay, somebody wake
01:21:08
Sachs up.
01:21:10
Sax, just so you know, one thing about
01:21:12
America is we've got a pretty good
01:21:14
advantage in gene editing technology
01:21:16
today. And China is on its way coming
01:21:18
after the US on this technology front as
01:21:20
well as AI. You woke him up. You should
01:21:22
peak his interest. China's coming for
01:21:24
Uranus. And you will see some very good
01:21:27
posts recently on people talking about
01:21:29
how a lot of technology that's been
01:21:31
developed in the United States is being
01:21:33
transferred to China. and they are
01:21:34
building biotech companies and biotech
01:21:36
solutions that take American IP and put
01:21:38
them into market and this has also had a
01:21:41
very big dampening effect on the biotech
01:21:42
market. So what do you think we should
01:21:44
do about that?
01:21:46
It should be in the trade negotiations
01:21:48
some tighter IP controls. I can tell you
01:21:50
Saxs that a very large company I had a
01:21:53
conversation with their CEO and they
01:21:56
said this exact thing which is that the
01:21:58
amount of IP that's available for
01:21:59
purchase or for
01:22:02
licensing questionable provenence but
01:22:05
how effective and good it is is causing
01:22:07
them to figure out how to establish more
01:22:09
of a presence to just buy stuff there.
01:22:11
It's kind of like so what does that
01:22:13
mean? Meaning that let's just say you
01:22:15
have an R&D effort in the United States.
01:22:17
you can almost hedge your bets by having
01:22:20
an outpost in or near China, let's say
01:22:22
Hong Kong, where there's an equivalent
01:22:25
project for almost anything that we're
01:22:27
working on happening there at the same
01:22:30
time in real time. And so if you can't
01:22:32
figure out how to get it here, you can
01:22:33
get it there and it's much murkier and
01:22:36
the the laws are sort of unclear and
01:22:39
it's just different enough where you can
01:22:41
kind of get to the same place. So it's
01:22:44
just to say what Freeberg said, which is
01:22:46
that the amount of R&D that the Chinese
01:22:49
are doing is roughly equivalent to the
01:22:50
amount of R&D that the United States is
01:22:52
doing. Yeah. And so as a result,
01:22:55
everything that we discover, they
01:22:56
discover. And so it really is a foot
01:22:58
race. And we're kind of neck and neck.
01:23:02
For sure. This is a perfect time to make
01:23:04
sure that we have an electricity
01:23:06
deficit. Perfect time. Well, I agree
01:23:08
with you. We need to build out power,
01:23:10
but I think the administration's
01:23:11
committed to doing that. Drill, baby,
01:23:12
drill. I'm not questioning Trump's
01:23:14
desire. I'm talking about the actual
01:23:16
congressional bill details. I saw
01:23:18
yesterday the Tennessee Valley Authority
01:23:20
filed the first ever application for a
01:23:23
small modular reactor. Let's go.
01:23:25
Hopefully, it gets approved very
01:23:27
quickly. This is, by the way, I I I will
01:23:29
watch this. I think others should watch
01:23:30
this to see how quickly they're able to
01:23:32
get through the regulatory process to
01:23:34
get that thing approved. And that will
01:23:36
be a very good indicator if it gets
01:23:37
approved and they can get fast
01:23:38
permitting, fast deployment. That would
01:23:40
be an incredible insight on how quickly
01:23:42
the US will be able to scale power
01:23:44
production for AI over under 2032. What
01:23:46
exactly? Let's do a poly market. Great.
01:23:49
Let's do a poly market and let's you
01:23:50
should create a special Trump
01:23:53
designation like a speedrun designation.
01:23:55
When something's super important, he
01:23:57
puts the speedrun designation on it and
01:23:59
they they actually attempt to do it as
01:24:00
fast as possible. I don't know if you
01:24:02
guys watched my interview with Bergam,
01:24:03
but he said that there's a permit for a
01:24:05
copper mine that's been outstanding for
01:24:08
30 years. They've been trying to get
01:24:09
through the permitting process. 30 years
01:24:11
to get this copper mine started to run.
01:24:14
And now he and the team are trying to
01:24:15
fasttrack this thing. But 30 years it's
01:24:17
been sitting through a Here's the good
01:24:18
news. The Manila folder has made its way
01:24:20
down the elevator. So, it's down there
01:24:22
in the subb
01:24:25
uh shipped. But let's go ahead and get
01:24:26
rid of Doge. Let's get rid of Doge and
01:24:28
keep things Let me ask Let me ask
01:24:29
Freeberg. Well, I don't want to get rid
01:24:31
of Doge, but let me ask you a very
01:24:32
question. No, I'm joking. I'm like,
01:24:34
yeah, I know. Given the amazing advances
01:24:36
that you see coming with crisper,
01:24:39
biotech, but also AI, robotics, what are
01:24:42
the odds that we can grow our way out of
01:24:45
this fiscal problem? Yeah.
01:24:48
Yeah. I do I I I'll be honest. I think
01:24:51
the limiting fac so the crisper side is
01:24:54
going to have profound effects in
01:24:55
longevity, human health, and food
01:24:57
abundance. So those are I think about
01:24:59
abundance in four ways. One is abundance
01:25:02
of food or calories. One is in abundance
01:25:05
of labor through automation. One is in
01:25:08
abundance of lifespan or longevity. And
01:25:10
one is an abundance of what I think is
01:25:12
the most important and gating factor
01:25:14
which is energy which leads to
01:25:15
everything else. So you can't have the
01:25:17
labor savings without the energy and the
01:25:18
productivity improvements without the
01:25:20
energy. So the thing I do watch the
01:25:21
most, Sax, and that I do think you and I
01:25:23
have talked about this that I do think
01:25:24
is the critical lynch pin for all of
01:25:26
those other points of abundance being
01:25:28
unleashed is the energy equation. And
01:25:30
this is where I look very clearly and
01:25:32
plainly at what China's doing versus
01:25:34
what where we're at today. We've got
01:25:36
stated intentions, but the actions are
01:25:38
where we're still, you know, we got to
01:25:39
start showing up. That's it. But let's
01:25:41
assume that we have enough power. I
01:25:43
mean, my point's really about Yeah. If
01:25:46
we have enough power, the 38 trillion of
01:25:47
debt doesn't matter. Really? You you're
01:25:49
saying that in a No, I'm very very
01:25:51
serious way. That's a pretty that's a
01:25:54
pretty important conclusion because if
01:25:57
that's true and we can grow our way out
01:26:00
of this problem, then that's an
01:26:02
alternative to having this austerity
01:26:04
approach where there's for which there's
01:26:06
no political, you know, will sax listen.
01:26:09
If I saw us adding a terowatt of
01:26:11
electricity production capacity per year
01:26:12
in the US, I would shut the up about the
01:26:14
debt. I wouldn't care. terowatt. We're
01:26:18
talking about gigawatts. Yeah, I'm
01:26:20
talking about one terowatt. No, I know.
01:26:21
But I'm saying I'm like chi China, you
01:26:23
guys know China right now is dealing in
01:26:25
in gigawatts. Yeah, I know. Have a
01:26:28
terowatt and China is dealing in
01:26:30
terowatts. China is moving. I mean,
01:26:32
we've been through the numbers, but
01:26:33
they're going to they're
01:26:35
adding an entire United States every 18
01:26:38
months. Every 18 months. Yeah. So, every
01:26:41
18 months, China adds all of the power
01:26:43
production capacity of the entire United
01:26:45
States. additional to their grid mostly
01:26:48
in nuclear and solar. Solar is a big
01:26:49
component.
01:26:51
Hydro hydro is look at the look at the
01:26:53
number of times they're adding like
01:26:56
hundreds of gigawatts just on the on the
01:26:59
Yang. Yeah. Look at the number of times
01:27:00
that Elon has said this. Like it's like
01:27:02
the amount of energy that's available if
01:27:05
we just harness it from the sun and
01:27:07
redirect it with some storage would
01:27:10
solve all of our problems. But the
01:27:12
problem is that we're at like just a
01:27:14
little over a terowatt. We're inching
01:27:16
along in gigawatts whereas China is in
01:27:20
the terowatts adding terowatts. Yeah,
01:27:22
they're
01:27:23
in that chart. There are three ter
01:27:26
corner of Arizona or Utah going to eight
01:27:30
and we're at one going to two. Yeah,
01:27:31
I've shown this many times before on
01:27:33
this show. Crazy. This is in this is
01:27:35
total production over a year. So this is
01:27:37
terowatt hours and this shows 4,000
01:27:40
terowatt hours is where the US is at. If
01:27:43
you break that down on basically a
01:27:45
continuous production, you measure it in
01:27:47
watts. So the number of watts that the
01:27:49
US can make in any given point in time
01:27:50
or is making any given point time is 1
01:27:52
terowatt, 1 trillion watts. Meanwhile,
01:27:54
right now, China is making three
01:27:57
terowatts and they're scaling up to make
01:28:00
eight. We are at one going to two over
01:28:03
the next 15 years.
01:28:05
So you consider this graph to be more
01:28:07
important than the the budget deficit
01:28:10
graph? I do. And I'll tell you why. Um,
01:28:13
I've got another chart. I can pull it up
01:28:14
for next time. So, now do you think it's
01:28:16
esoteric sachs to like add these line
01:28:18
items that this market? Well, I don't I
01:28:21
just don't I mean, you're talking about
01:28:23
I I just I'm not sure. You're going to
01:28:25
have to provide a lot more information
01:28:27
about what exactly is No, but I mean,
01:28:29
look, it's obvious. It's obvious and
01:28:31
Blackstone. And I just want to translate
01:28:34
this. I It's so obvious that any
01:28:36
reporting on that, Jamath, I just
01:28:37
haven't seen anything. It just happened
01:28:39
last night. This is the point. I'll just
01:28:41
say one thing because I think it's I
01:28:42
just want to add it at the end. Sure.
01:28:44
Automation which is unlocked by AI. We
01:28:47
are going to see an explosion of robots.
01:28:49
They're not going to all be humanoid
01:28:50
robots, but like
01:28:51
robotics allows humans to get an
01:28:54
incredible amount of work done. Imagine
01:28:56
having the cost to build a giant
01:28:58
building go down by 50x. That's what
01:29:01
this unlock unleashes is all of this.
01:29:03
Optimist doing it. Yes, we'll have
01:29:05
optimist but we'll have these robotic
01:29:06
devices doing it. You can see it in
01:29:08
China. Go look at how they build
01:29:09
bridges. It takes like 3 days to build a
01:29:11
multi-m bridge in China because the
01:29:13
whole thing is automated. They do
01:29:15
automated drilling. They do automated
01:29:16
mining. They do automated building. All
01:29:18
of that If we unleash that
01:29:20
capacity in the United States, we can
01:29:22
build things that are economically
01:29:23
productive for America. All of this gets
01:29:25
unleashed. The technology is here today.
01:29:28
The only thing that's missing is the
01:29:29
power. We want this. David, I would just
01:29:32
then ask folks, go talk to like Duke
01:29:34
Energy, go talk to Constellation Energy,
01:29:37
go talk to the the drillers, but they'll
01:29:39
all tell you the same thing. If you want
01:29:41
a NAT gas turbine, we have to wait until
01:29:44
2030. Why? Because China owns them all.
01:29:47
They make them all. So, if you say,
01:29:49
"Hey, Chimat, drill, baby, drill, and
01:29:51
use that gas." And I say, "Okay, I'm
01:29:53
waiting in line for 5 years to get a
01:29:55
turbine." So, then it's like, "Okay,
01:29:57
what can I do? Nuclear? I can't because
01:30:00
that'll take till
01:30:01
2035. So then I'm like, okay, I'll go
01:30:03
into the tax equity markets and I'll
01:30:05
finance myself residential or industrial
01:30:07
solar. And when that goes away, there is
01:30:09
no energy. Somebody should just go and
01:30:12
figure this out. By the way, this is our
01:30:14
Manhattan project. This there was the
01:30:15
Manhattan project, then there was the
01:30:17
Apollo project. This generation's
01:30:19
Manhattan and Apollo project is energy
01:30:21
production scaling in the United States.
01:30:23
That's it. All the other flows. We don't
01:30:25
need the government to do AI. We don't
01:30:26
need the government to do automation.
01:30:28
private market. Let us let us do the
01:30:30
energy. We need to figure out how to get
01:30:32
energy and you can't decapitate the
01:30:34
markets that finances it. We're all
01:30:35
willing to put up the risk capital.
01:30:36
That's the crazy part. We don't even
01:30:37
need the government. I don't know any of
01:30:39
the details of what Jama's talking
01:30:40
about, but I mean generally like you
01:30:41
think and your chairman dictator
01:30:45
dictator. Yeah. Chairman dictator. Who
01:30:48
you dictate for your chairman dictator?
01:30:51
I'm trying to close the show for the
01:30:52
seventh time. your chairman dictator
01:30:54
Chimath Payatia Yar David Sachs Sultan
01:30:57
of science I am the world's greatest
01:30:59
moderator I'll be in Singapore next week
01:31:01
if anything's popping off
01:31:04
Jasonallin.com hit me up why do you
01:31:06
broadcast your location I don't what
01:31:08
what are you I'm going to Singapore I'm
01:31:09
doing a bunch of public events people
01:31:11
know you're the same one that wants to
01:31:12
like for the pin and then you're just
01:31:14
telling people where you're at put on
01:31:16
see what happens tough guy put the pin
01:31:18
on the pin I'm sorry you took allin.com
01:31:22
as your email address. Jason.com. We
01:31:24
each have that. Wait, but don't
01:31:27
broadcast that out, bro. Come on. Just
01:31:29
email me.
01:31:30
Jasonin.com. You know how long
01:31:33
that domain? That's your domain. This is
01:31:37
not your domain. This is
01:31:40
all you ungrateful pricks.
01:31:44
Why do you use that for negotiation? Oh
01:31:46
my god. I can't even goddamn email. It's
01:31:49
not in the contract that you get the
01:31:50
email.
01:31:51
Jasonallin.com for life. I think we
01:31:53
should charge you 50k a month for that.
01:31:55
Email for life. Executive producer for
01:31:58
life and email for life. We'll see you
01:32:00
all next time. What are you going to
01:32:01
Singapore for? He's trying to pay you 25
01:32:04
grand. What are you doing? This is a
01:32:05
plug. This is a
01:32:06
plug. Singapore. There's no plug. I'm in
01:32:08
Singapore next week. That's it. You were
01:32:10
in the Middle East last week.
01:32:12
Yeah, but I didn't pre-announce it. You
01:32:15
didn't pre-announce it. You were on CNN.
01:32:17
You pay for your business class ticket,
01:32:19
Jake Hell, to come to a talk. Is that
01:32:21
what's going on here? You got a free
01:32:22
ticket. Let's go. First of all, first of
01:32:23
all, let let me just tell you I speaking
01:32:25
gigs are six figures now, I flown
01:32:29
private to Singapore. Never again. That
01:32:30
is the dumbest thing. Fly Singapore
01:32:32
Airlines. The best. The best. Are you
01:32:35
doing Singapore first, Jal? The Yeah, I
01:32:38
don't know. I think I'm United,
01:32:39
actually. But do not fly United. Do not
01:32:40
fly United to Singapore. Oh, really?
01:32:42
Okay, I'll switch it out. Fly Singapore.
01:32:44
What do you think? Raffles? You like
01:32:45
raffles?
01:32:47
Raffles is the best. That's what I keep
01:32:49
hearing. Let me have the best. The best.
01:32:51
The best. Beautiful. All right,
01:32:52
everybody. We'll see you next time.
01:32:53
Byebye. Love you guys. Byebye.
01:32:57
We'll let your winners ride.
01:33:00
Rainman David S.
01:33:04
And it said we open sourced it to the
01:33:06
fans and they've just gone crazy with
01:33:08
it. Love you. Queen
01:33:10
[Music]
01:33:13
of your life.
01:33:17
Besties are gone.
01:33:20
That is my dog taking notice of your
01:33:21
driveways.
01:33:25
Oh man, my appetiter will be. We should
01:33:28
all just get a room and just have one
01:33:29
big huge orgy cuz they're all just
01:33:31
useless. It's like this like sexual
01:33:32
tension and we just need to release
01:33:34
something else.
01:33:38
Your
01:33:40
feet. We need to get mercy.
01:33:45
[Music]
01:33:50
I'm going all in.
