Search:

ICE Chaos in Minneapolis, Clawdbot Takeover, Why the Dollar is Dropping

January 31, 202601:30:02
00:00:00
All right, everybody. Welcome back to
00:00:01
the Allin podcast. Your favorite
00:00:04
podcast, podcaster's favorite podcast.
00:00:06
With me again, the original Quartet is
00:00:09
here. Schmoth Poly Hotia in just an
00:00:13
absolutely fabulous winter sweater.
00:00:16
January
00:00:18
looking great. Look at the size of those
00:00:19
buttons.
00:00:20
>> Huge buttons.
00:00:21
>> How many rhinos died to provide those
00:00:23
buttons?
00:00:24
>> Zero rhinos. Zero rhinos.
00:00:26
>> I'm a simple man that lives by simple
00:00:28
means. Okay. Beautiful. Beautiful. And
00:00:31
your sultan of science, David
00:00:33
Freriededberg. What's the background
00:00:35
here? Is that in Is that a melancholy
00:00:37
and infinite sadness background? I'm
00:00:39
trying to figure it out.
00:00:40
>> J, we don't talk about my backgrounds.
00:00:41
Thank you.
00:00:42
>> It looks like melancholy and infinite
00:00:44
sadness by the double album by the
00:00:46
amazing band um Smashing Pumpkins. Am I
00:00:50
close or is it the original artwork of
00:00:52
that?
00:00:53
>> Don't talk about my backgrounds.
00:00:54
>> You don't talk about your backgrounds.
00:00:56
Talk about your background. much giving
00:00:58
me so much to work with here. Luckily, I
00:01:00
have my straight man, my brother in
00:01:03
arms, my Davos party crashing partner,
00:01:07
David Saxs.
00:01:09
>> I got you your first invite to something
00:01:12
elite and exclusive.
00:01:15
>> I mean, I got invited to go 25 years
00:01:18
ago. They just wanted 50 dimes, but we
00:01:21
had a fun time.
00:01:22
>> Yeah, we had a good time.
00:01:23
>> We had a good time.
00:01:25
Uh any uh post Davos WF impressions? We
00:01:29
had a lot of interesting meetings. So
00:01:31
most of which I don't think we can talk
00:01:32
about on air, but um yeah, it was it was
00:01:36
an interesting uh interesting event.
00:01:38
>> We were staying in a log cabin that was
00:01:40
like 300 years old. The ceilings were
00:01:42
like six feet high and the door frames
00:01:44
were like 5t high. So bumped our heads a
00:01:47
couple of times.
00:01:48
>> Yeah,
00:01:48
>> it was pretty crazy.
00:01:49
>> It was brutal. I mean,
00:01:51
>> it looked good on the web. The photos of
00:01:54
the place looked amazing.
00:01:55
>> The Airbnb photos look great.
00:01:57
>> Yeah. But, uh, I think you need to be
00:02:00
inside the circle. You need to be inside
00:02:02
the thick of it. Not driving in every
00:02:04
day, but it was a distinctly different
00:02:06
Davos. Uh, we've mocked Davos here for
00:02:09
many years,
00:02:11
but this one was a business takeover and
00:02:14
a Trump takeover.
00:02:15
>> Correct.
00:02:16
>> Chimath, you would have loved it. It was
00:02:18
1.5 days of everybody hand ringing of
00:02:22
what Donald Trump would say when daddy
00:02:24
got there. Then for the 75 minutes daddy
00:02:27
gave his talk, the entire city shut
00:02:30
down. Everybody ran to a television set.
00:02:33
Then for the next 1.5 days, everybody
00:02:35
talked about what Trump said.
00:02:37
>> Basically, I think there are two really
00:02:38
big differences with the Davos based on
00:02:40
what I heard because this is the first
00:02:41
time I've attended. one is it was much
00:02:43
more business- ccentric and then second
00:02:46
there were a lot more Americans there a
00:02:48
much bigger American presence and I
00:02:51
think that that owes to the fact that
00:02:54
President Trump gave a major speech
00:02:56
there and I think it was Larry Fank
00:02:59
who's chairman of the whole thing now
00:03:02
who sort of orchestrated that
00:03:04
>> and he he wanted to get President Trump
00:03:05
there and I think he's pushed
00:03:07
>> for them to be a little bit more
00:03:09
business centric and it feels like
00:03:12
they're catering a little bit less to
00:03:14
kind of boutique
00:03:16
European political issues, although
00:03:19
that's clearly a big part of it. I would
00:03:22
say in terms of memorable moments, a big
00:03:25
one was I'm only going to talk about
00:03:26
this because I think it was publicly
00:03:28
reported otherwise I wouldn't bring it
00:03:29
up, but there was this opening night
00:03:31
dinner and so Larry Fink calls on like
00:03:35
five people to give little speeches and
00:03:37
the last one he called on was Howard
00:03:39
Lutnik, our Secretary of Commerce. And
00:03:42
Howard just goes up there and he just
00:03:44
starts dropping truth bombs on them and
00:03:49
and uh and again he said all this on the
00:03:51
record in his remarks at Davos I think
00:03:54
either that day or the next day. So
00:03:56
again I'm not talking out of school. It
00:03:58
was all stuff that he said but basically
00:04:00
he said to like all the Europeans he
00:04:01
said look I've been coming here for 30
00:04:03
years and you guys have completely
00:04:04
failed. You've wrecked your economies
00:04:06
with all this net zero stuff and climate
00:04:08
change and energy. and he just started
00:04:11
blasting on that and then all the open
00:04:13
border stuff. He really let them have
00:04:15
it. I mean, he was really the truth. And
00:04:17
then there was this uncomfortable like
00:04:19
rustling in the audience, you know, as
00:04:21
he was like gathering steam and then
00:04:24
it's been reported Al Gore started
00:04:26
booing at the end.
00:04:28
>> Drunk on Chardonnay. Is that true?
00:04:32
I can't verify that it was Al Gore, but
00:04:34
I definitely heard somebody. And I think
00:04:35
it was Al Gore who was
00:04:36
>> just like the two guys from the Muppets
00:04:38
in the Balcony, Statatler and whoever.
00:04:40
And it was just like Waldorf and
00:04:42
Statler.
00:04:43
>> Well, look, I mean,
00:04:44
>> yeah, Walder and Statler.
00:04:45
>> This whole climate change thing was Al
00:04:48
Gore's big hoax going back to the '9s.
00:04:50
And
00:04:51
>> well, did he win an Oscar? He won an
00:04:52
Oscar. I mean, and even absolutely Bill
00:04:55
Gates has acknowledged that this is not
00:04:56
an existential threat. It can be dealt
00:04:58
with,
00:04:59
>> you know. So I think the bloom is off
00:05:02
the rose in terms of that whole agenda
00:05:05
>> I'd say at Davos. Although to be sure
00:05:07
I'd say most people there still probably
00:05:10
agree with Al Gore. I mean they haven't
00:05:12
changed their policies. I'm talking
00:05:13
about the European countries even though
00:05:14
it's wrecking their economies.
00:05:16
>> They are beholden to this net zero idea.
00:05:19
>> It feels like they're in transition and
00:05:21
scrambling to try to figure out where
00:05:24
they what direction they should go in.
00:05:25
they don't feel they can trust America
00:05:27
or that we're not their like reliable
00:05:29
partner. We're not going to bail them
00:05:30
out. We're not going to protect them,
00:05:32
etc. And they're going to have to sort
00:05:34
of get together as the mids mid-tier
00:05:37
economies, economies 5 through 20, and
00:05:40
they're going to just have to build
00:05:42
their own voting block, their own
00:05:43
trading block, and uh go it alone and
00:05:46
spend their own money.
00:05:47
>> They have it. It's called the EU. It's
00:05:49
just yeah
00:05:50
>> has not been very successful because I
00:05:52
think of their own policies, their
00:05:54
energy policies and the open border
00:05:55
policies.
00:05:56
>> I'm not sure what going it alone means
00:05:58
if you don't have
00:05:59
>> best-in-class AI or best-in-class
00:06:01
weapons.
00:06:02
>> Absolutely. I I think it means they're
00:06:03
going to have to start investing in
00:06:04
those things, buying weapons, making
00:06:07
weapons means much of anything.
00:06:10
>> Yeah. I I think it means Canada
00:06:13
and the EU and all these other countries
00:06:16
that feel, hey, we're we're we're gonna
00:06:18
have our rug pulled by the Trump
00:06:21
administration. We're just going to have
00:06:22
to band together and create commerce.
00:06:24
And we saw China and uh Canada do a big
00:06:28
deal. And I think that's them just
00:06:29
trying to say, hey, we have some
00:06:32
sovereignty here. We're going to do a
00:06:33
partnership with Canada and bring BYD
00:06:35
cars in here. And hey, they'll be our
00:06:36
big trading partner. So, you know,
00:06:38
that's the reaction I think on the other
00:06:41
side.
00:06:43
>> But they don't
00:06:45
>> they don't what
00:06:46
>> they don't have as many degrees of
00:06:48
freedom as they think they do because
00:06:50
great powers define the international
00:06:52
system, not mid-tier powers. And a bunch
00:06:54
of second or third tier powers cannot
00:06:56
redefine the international system even
00:06:58
if they band together. And I think at
00:07:00
the end of the day, the Europeans, they
00:07:03
understand the importance of the United
00:07:05
States and specifically they want to
00:07:06
keep the US in Europe. I mean, they're
00:07:09
desperate to keep NATO together and to
00:07:11
keep the US interested and present in
00:07:14
Europe because just remember European
00:07:16
history before the Americans were there.
00:07:19
It's like hundreds of years of wars and
00:07:21
constantly fighting each other.
00:07:22
>> Exactly.
00:07:23
>> Culminating in World War I, World War
00:07:25
II, basically the total self-destruction
00:07:27
of Europe. And the most peaceful period
00:07:30
they've ever experienced has been post
00:07:32
1945 when the Americans are there as the
00:07:35
great pacifier. So, they do not want us
00:07:37
leaving. And I think they're willing to
00:07:39
make large concessions to the US to
00:07:42
ensure that we stay there, even though
00:07:43
they'll probably grumble about it. But I
00:07:46
think that to that end, I think that
00:07:50
what President Trump was saying is,
00:07:51
look, you guys got to share in the
00:07:53
burden here. We've been paying for this
00:07:54
whole thing.
00:07:54
>> Yeah. No, it makes total sense. And they
00:07:56
got their spending up to 3% for NATO and
00:07:59
they're going to go to five. So, mission
00:08:00
accomplished on that.
00:08:01
>> Yes. And on Greenland, he's like, look,
00:08:04
you know, we we we was the best part of
00:08:06
the speech. He's like,
00:08:09
>> "Do you want me to talk about
00:08:10
>> for you?
00:08:11
>> You got to do something for us. This has
00:08:12
got to be a choice."
00:08:12
>> He gets 45 minutes in. He's like, "Do
00:08:14
you want me to talk about Greenland? I
00:08:16
could talk about it. I You ready? I'm
00:08:18
going to talk about it right now." And
00:08:19
then he starts talking about it. But
00:08:21
then he starts calling it Iceland. And
00:08:24
everybody in the room's like, "Wait, he
00:08:25
wants Iceland, too?" And the whole buzz
00:08:28
was Trump's going to take Iceland and
00:08:30
Greenland. And then, you know, he sort
00:08:33
of backed down, but we're not going to
00:08:34
invade. And then there was like a sort
00:08:37
of sigh of relief. But do you think they
00:08:39
actually thought Sachs that he was going
00:08:42
to invade Greenland to take it?
00:08:45
>> Well, I mean, I don't know. I think that
00:08:48
obviously they thought it was a
00:08:49
possibility, but the president took the
00:08:51
use of force off the table. And yeah, I
00:08:53
think you did feel a sigh of relief
00:08:56
there. I think you're right. But I think
00:08:58
that also, I mean, we don't know the
00:09:00
details. They haven't been publicly
00:09:01
reported, but there was some sort of
00:09:03
meeting at Davos that was convened, I
00:09:06
think, by the NATO Secretary General,
00:09:08
Mark Ruda, where they negotiated an
00:09:10
acceptable compromise on the issue. So,
00:09:13
I'm sure we'll find out the details in
00:09:15
due course, but it's safe to say that
00:09:18
President Trump got enough of what he
00:09:19
wanted that he was satisfied with what
00:09:21
they worked out.
00:09:22
>> I don't know how they just like it's
00:09:23
like Lucy with the uh football and
00:09:25
Charlie. It's like he's gonna pull the
00:09:28
football. Like he's just anchoring the
00:09:29
ne negotiations at a military invasion
00:09:32
and takeover. He obviously is just going
00:09:34
to go for a lease. It's like they they
00:09:36
don't know it in year, you know,
00:09:38
whatever we're in now, year five or six,
00:09:41
year six of Trump being president. Like
00:09:43
they should get it by now. He just
00:09:45
anchors things at an impossible insane
00:09:47
level and then he falls back to whatever
00:09:49
he really wanted. It's a classic
00:09:51
negotiating technique. All right, we got
00:09:53
a lot on the docket. Let's get to work.
00:09:55
Everyone wants to hear four venture
00:09:56
capitalists and investors talk about the
00:09:59
horrific situation in Minneapolis. So,
00:10:01
here we go. Uh, for background, I mean,
00:10:04
Jason,
00:10:05
>> for background, last month, the DHS
00:10:07
started an operation called Metro Surge,
00:10:11
sending 3,000 federal agents into
00:10:12
Minnesota to crack down on illegal
00:10:14
aliens. Over the last 3 weeks, two
00:10:17
motans were tragically killed in
00:10:19
altercations with federal agents.
00:10:20
January 7th, 37year-old Renee Good was
00:10:24
shot to death by an ICE agent. This
00:10:26
incident involved Good accelerating her
00:10:28
car, which was surrounded by agents at
00:10:30
the time. We're still waiting for the
00:10:32
final investigation on this one, but
00:10:34
apparently three shots, one through the
00:10:36
front windshield, perhaps two through
00:10:37
the side. Uh, all these details are
00:10:39
still being investigated. Then
00:10:42
tragically on January 24th, Alex Prey,
00:10:45
also 37, was shot and killed by two
00:10:47
border patrol agents, not ICE. Prey was
00:10:49
an ICU nurse at the local VA hospital.
00:10:53
There's a ton of frame by frame
00:10:54
breakdowns available. New York Times and
00:10:57
Wall Street Journal did a good job on
00:10:58
these. So, I think maybe it's best for
00:11:01
us to focus here on maybe the aftermath
00:11:03
of all this and the resolution, but you
00:11:05
guys can feel free to chime in on the
00:11:07
frame by frame breakdowns if you like.
00:11:10
In five parts, Steven Miller tweeted
00:11:12
that Freddy was an assassin trying to
00:11:14
murder federal agents. A source told
00:11:16
Axis that Gnome said, "Everything I've
00:11:19
done, I've done at the direction of the
00:11:20
president and Stephen, Steven being
00:11:22
Steven Miller. Greg Bovino uh has been
00:11:25
removed from duty and had his social
00:11:27
media accounts turned off. President
00:11:29
Trump has pivoted, evolved
00:11:33
and put Tom Hman in charge." Quote, "Tom
00:11:37
is tough but fair and will report
00:11:39
directly to me," Trump wrote. And uh at
00:11:42
the time of this taping, which is on
00:11:44
Thursday's uh there was a press
00:11:46
conference this morning and here is what
00:11:49
Tom Holman said and after 30 seconds
00:11:52
we'll go to you Saxs for your reaction
00:11:54
to all this.
00:11:55
>> No agency organization is perfect.
00:11:57
President Trump and I along with others
00:11:58
in administration have recognized that
00:12:01
certain improvements could and should be
00:12:03
made. That's exactly what I'm doing
00:12:04
here. So if we get these agreements in
00:12:07
place, that means less agents on the
00:12:09
street, more agents in the jail. Matter
00:12:12
of fact, I have staff from CBP and from
00:12:15
ICE working on a draw down plan. What
00:12:18
does that look like based on the
00:12:19
cooperation?
00:12:20
>> Sax, your thoughts?
00:12:22
>> Well, first let me say the deaths of
00:12:24
Renee Good and Alex Petty are
00:12:27
regrettable and and tragic. So are the
00:12:30
deaths of Lincoln Riley, Joselyn
00:12:32
Nungare, Rachel Morren, Victoria
00:12:34
Harwell, Ivory Smith, and too many
00:12:37
others to mention who are murdered by
00:12:38
criminal illegal aliens. And the media
00:12:40
won't ever tell you their names. But
00:12:42
President Trump was hired by the
00:12:43
American people to do a job, which is to
00:12:46
seal the border and deport criminal
00:12:47
aliens so that more of these tragedies
00:12:49
do not occur in the future. And this is
00:12:51
a popular policy. Over 55% of the
00:12:53
American people say they want all
00:12:56
illegal aliens deported and over 90%
00:12:59
want criminal aliens removed. And by
00:13:00
criminal aliens, I'm referring to the
00:13:02
ones who commit additional crimes after
00:13:04
they enter the country illegally. Now,
00:13:07
this policy is working. Uh murders were
00:13:09
down 21% last year. It's one of the best
00:13:12
years in record. And in most states, the
00:13:15
process is smooth and doesn't make
00:13:17
national news. And the reason for that
00:13:19
is because local authorities are
00:13:20
cooperating with ICE. But Minneapolis
00:13:24
has taken a different approach. They've
00:13:25
engaged in a campaign of quote massive
00:13:28
resistance to federal authorities. So
00:13:30
let's talk about what's actually
00:13:32
happening there on the ground. I think
00:13:34
the first thing to understand is that
00:13:37
what's happening is much more than just
00:13:38
protests. And obviously I have no
00:13:40
problem with people peacefully
00:13:41
protesting and making their opinions
00:13:43
known, but that's not what's going on
00:13:45
here. These are Antifa style operations
00:13:47
designed to thwart
00:13:50
the enforcement of federal immigration
00:13:52
law. They're highly organized. They're
00:13:54
communicating in encrypted chat groups.
00:13:57
They stalk and dox ICE agents. They
00:14:00
follow them around town. They surround
00:14:02
them at their hotels. They use their
00:14:03
cars to block roads. And they use
00:14:06
bullhorns and whistles to alert
00:14:07
criminals who are about to be arrested.
00:14:09
And remember, ICE is a law enforcement
00:14:11
agency. They have warrants to arrest
00:14:14
known criminal aliens. Despite this
00:14:17
rhetoric of them being like the Gustapo,
00:14:19
they are going after specific named
00:14:20
individuals for whom they have warrants
00:14:23
to arrest. These are dangerous missions
00:14:26
and these agitators are interfering and
00:14:28
making these missions even more
00:14:29
dangerous. Now, the media has tried to
00:14:32
portray good and pretty as simply
00:14:34
innocent bystanders or people who were
00:14:37
peacefully protesting ICE policies. They
00:14:40
weren't. They were foot soldiers in
00:14:42
these Antifa style operations and most
00:14:45
importantly they brought deadly weapons
00:14:47
to the fight. So Renee Good hit an
00:14:49
officer with her SUV which under a
00:14:52
Minnesota law signed by Tim Waltz
00:14:54
himself in 2020 justifies the use of
00:14:56
deadly force by an officer to defend
00:14:58
himself. And Alex Prey was even more
00:15:01
reckless. I think we've probably all
00:15:02
seen the video by now where he sought
00:15:05
confrontation with ICE officials. He was
00:15:07
kicking the car. He was in a rage. This
00:15:09
wasn't his first time doing this. And
00:15:11
any experienced gun owner will tell you
00:15:13
that if you're armed and you're dealing
00:15:16
with law enforcement, you have to be the
00:15:17
world's biggest pacifist because you're
00:15:19
putting your life in danger by making
00:15:22
them fear that their lives are in
00:15:23
danger. And I think the mainstream media
00:15:25
didn't tell people these facts. They
00:15:27
just presented highly selective camera
00:15:29
angles. They even airbrushed and
00:15:31
facetuned Prey to make him appear to be
00:15:33
a more, I guess, handsome victim, which
00:15:36
is truly sick. Now, in a way, I feel
00:15:39
sorry for Good and Pretty because they
00:15:42
were the victims of a tinder box that
00:15:44
was created in Minnesota by the extreme
00:15:48
rhetoric and decisions of Tim Waltz,
00:15:52
Jacob Fry, and the political
00:15:54
establishment. The local police in
00:15:56
Minneapolis should have been allowed to
00:15:58
keep conditions safe on the street by
00:16:01
creating a perimeter and keeping
00:16:02
protesters away from ICE officers who
00:16:05
were executing lawful warrants of
00:16:08
arrest. But the police were told not to.
00:16:11
And then the agitators stepped in and
00:16:13
they took advantage of this sort of
00:16:14
vacuum of authority to physically
00:16:17
intervene. So I think it was almost
00:16:19
inevitable that some sort of tragedy was
00:16:21
going to result from this abdication of
00:16:23
public safety. Now why would Walson Frey
00:16:27
want to risk such tragedies with their
00:16:30
massive resistance? I think there's two
00:16:32
reasons for this. This is the last point
00:16:33
I want to make. First, they are
00:16:35
desperate to change the subject from the
00:16:36
billions of dollars of fraud that they
00:16:39
allow to occur on their watch. Remember
00:16:41
we had Nick Shirley on the show just a
00:16:43
few weeks ago talking about the 8
00:16:44
billion so that was stolen by Somali
00:16:47
fraudsters and this campaign of
00:16:50
resistance in Minneapolis has done a
00:16:52
really good job making everyone forget
00:16:54
about that. But I think there's a second
00:16:56
and bigger reason that applies and I
00:16:58
think it applies to to national
00:17:00
Democrats which is they want to thwart
00:17:03
mass deportations because illegal
00:17:05
immigrants are a vital part of their
00:17:07
power base. And you can see this in the
00:17:10
2030 aortionment forecast which just
00:17:12
came out. Illegal aliens count towards
00:17:15
the census which occurs every decade.
00:17:18
And the census determines the
00:17:19
aortionment of congressional seats and
00:17:21
electoral votes. And what you see in
00:17:23
these maps is that
00:17:26
citizens of blue states have been
00:17:29
migrating to red states because those
00:17:31
blue states are failing. And as a result
00:17:34
of that, blue states are expected to
00:17:37
lose nine house seats and electoral
00:17:40
votes because of the changing population
00:17:42
numbers. Illegal aliens in blue states
00:17:44
have been propping up those numbers. And
00:17:47
so, for example, in the last election,
00:17:49
President Trump would have won an
00:17:50
additional nine electoral votes
00:17:53
if we had an accurate counting. So,
00:17:56
look, this is not about principle. This
00:17:59
is bare knuckle politics. The Democrats
00:18:01
are playing for keeps. They don't really
00:18:03
care how many innocent Americans get
00:18:05
hurt or killed in the process. This is
00:18:08
about thwarting a popular policy of
00:18:10
deportations
00:18:12
and sealing the border which the
00:18:14
American people voted for. So don't let
00:18:16
the media fool you.
00:18:17
>> Freedber Chimath, you want to give your
00:18:19
opinion on these two tragic deaths
00:18:23
>> or not? I mean, I don't feel you're
00:18:24
obligated to comment on this if you
00:18:27
don't want to.
00:18:27
>> I'm happy to.
00:18:29
Nick, I sent a clip into the chat.
00:18:31
>> Deported all immigrants who are here
00:18:32
illegally, 55% of the New York Times,
00:18:34
Marquette, 64%. CBS News, 57%. ABC News
00:18:38
with a slightly different question, 56%.
00:18:40
So, what you're seeing essentially here
00:18:42
is a very clear indication that a
00:18:45
majority of Americans, in fact, when
00:18:47
they're asked this blunt question, which
00:18:48
I believe gets at the underlying
00:18:50
feelings, do in fact want to deport all
00:18:52
immigrants who are here illegally.
00:18:54
There's no arguing with these different
00:18:56
numbers cuz they're all essentially the
00:18:57
same across four different pollsters.
00:18:59
>> I think Sax is right that there's a very
00:19:01
very vocal minority. But if we just put
00:19:03
that aside, it's important right now to
00:19:05
just stick to the facts. Democracy is
00:19:08
supposed to be the will of the majority,
00:19:13
but also defense and protection for the
00:19:15
minority. In this example, the will of
00:19:17
the majority is pretty clear. As the CNN
00:19:19
clip just showed, everybody wanted the
00:19:22
southern border shut and the northern
00:19:24
border shut
00:19:26
and a structured path to deal with
00:19:29
illegal immigration. David's right that
00:19:31
that creates a cascade of second and
00:19:32
third order effects
00:19:35
that have huge implications with respect
00:19:38
to the Democrats and their ability to
00:19:40
have and curate power.
00:19:43
I don't know whether this is what's
00:19:45
motivating them or not. I don't want to
00:19:46
speculate on that. But the conceptual
00:19:49
problem and the conceptual desire of
00:19:52
Americans is undisputable. I think
00:19:54
that's why Donald Trump won. Now I think
00:19:57
though we have to explore the tactics.
00:19:59
I think the reality is that both of
00:20:01
these two deaths were complete and total
00:20:04
tragedies
00:20:06
and it has created such an upswell
00:20:09
that it has the potential to spin out of
00:20:12
control. And if it does that, it risks
00:20:16
his ability to continue doing his job
00:20:18
and delivering on the conceptual promise
00:20:20
that everybody wants. The other thing is
00:20:22
that I think that he has otherwise, the
00:20:24
president, done an incredible job up
00:20:26
until now. The fact that Tom Hman is
00:20:28
going there is a really good thing. He
00:20:30
was the same person that was awarded a
00:20:32
medal by Obama for how he managed
00:20:35
Obama's deportation process. It's time
00:20:37
to just get control of the process and
00:20:39
dial down the temperature because the
00:20:42
structural things that they are doing
00:20:44
are correct. There are people here that
00:20:46
broke the law. There are criminals that
00:20:48
are here illegally. We need to remove
00:20:51
them because that is the will of the
00:20:53
majority. Now we just need to find a way
00:20:55
of doing it that creates some freedom to
00:20:58
operate for all of law enforcement so
00:21:00
that these tragedies stop. That's my two
00:21:02
cents.
00:21:03
>> Okay. Freedberg, would you like to
00:21:04
comment on this or pass? I
00:21:07
>> I'll comment. Can I just ask you to
00:21:08
react to what Sach said? Do you agree or
00:21:11
disagree with his point about the
00:21:14
Democrats needing to remove people
00:21:15
because they do count in the census and
00:21:17
they increase the seats in the House.
00:21:20
>> Elon's talked about this a whole bunch
00:21:22
as well that the immigration is being
00:21:24
done to boost the voter roles. I don't
00:21:27
know enough about the census
00:21:29
specifically because that occurs x
00:21:30
number of years and uh if it's accurate
00:21:33
at all. So I'll leave that aside and do
00:21:36
some research on it. In terms of
00:21:38
importing people for votes, this uh
00:21:42
strategy does not make a lot of logical
00:21:44
sense. And so many workingclass people
00:21:47
voted Trump into office and so many
00:21:49
specifically uh Hispanic people, Mexican
00:21:52
people are all voting for Trump now
00:21:54
because he's a populist and he appeals
00:21:56
to to that group. So whatever if Biden
00:21:59
and the Democrats were doing that for
00:22:01
that reason, that makes no sense. And
00:22:04
also those people would have to become
00:22:05
citizens in order to vote and that's a
00:22:09
20 30year process. So, you know, they'd
00:22:12
be playing an incredibly long game uh on
00:22:15
that front
00:22:16
>> unless there is cheating in the voting.
00:22:17
>> Yeah.
00:22:18
>> And we've talked about that as well
00:22:19
here.
00:22:20
>> For example, if there's no voter ID, you
00:22:23
know,
00:22:23
>> you know, the Heritage Foundation, I
00:22:24
think David, you worked there at some
00:22:26
point, right? Did an internship.
00:22:28
>> That uh organization, that think tank
00:22:30
has done tremendous research into this.
00:22:32
They have a database. I think they've
00:22:33
collected now 3,000 cases of voter fraud
00:22:36
over a 40-year period. So really the
00:22:39
whole concept of voter forward being
00:22:40
able to tilt a presidential election is
00:22:42
just ridiculous and has not been proven.
00:22:44
I think Trump filed about 50 or 60
00:22:46
lawsuits and lost all 50 of them. So
00:22:48
there's there's no credence to that. Uh
00:22:50
but you know I have a couple of thoughts
00:22:52
broader on what we've seen and I do
00:22:55
actually agree with you David. There
00:22:56
should be voter ID everywhere. I I don't
00:22:58
think anybody should be able to vote
00:22:59
without a driver's license. If you can't
00:23:00
or ID if you can't take the time to get
00:23:02
ID why should you vote? It doesn't make
00:23:04
a lot of sense right.
00:23:04
>> Why do you think that's such a push on
00:23:06
the other side though Jal? Like what's
00:23:07
the motivation for not having IDs?
00:23:11
>> The stated
00:23:12
>> if it's not about getting people to vote
00:23:13
that aren't allowed to vote.
00:23:15
>> The stated reason, which I don't
00:23:18
believe, is that it's more democratic
00:23:20
and you want to get as many people to
00:23:22
vote as possible. So, but I don't agree
00:23:23
with that. I think everybody should have
00:23:24
ID.
00:23:25
>> Yeah. Like you got to have ID to buy a
00:23:26
beer, right? So,
00:23:27
>> yeah. I mean, to get on a plane to I
00:23:29
mean, even to ride a train, like you
00:23:31
need an ID. I don't I don't understand.
00:23:32
>> I think the answer speaks for itself. I
00:23:34
mean, uh, to me it's obvious the reason
00:23:36
why you prohibit, by the way, it's not
00:23:38
just saying you can vote without an ID.
00:23:40
They actually prohibit the people
00:23:41
administering the polls from checking.
00:23:43
There's only one reason to do that. You
00:23:45
want to allow cheating, obviously. And
00:23:48
one of the things that Doge found was
00:23:50
that there were lots of illegal aliens
00:23:51
being added to the social security
00:23:53
roles. Now, they weren't necessarily
00:23:54
collecting social security, but in a lot
00:23:57
of
00:23:57
>> they were actually paying into it. So,
00:23:58
it was quite the opposite. My point is
00:24:00
that when you get an SSN number, in a
00:24:02
lot of states, all you have to do is
00:24:04
check a box when you get a driver's
00:24:06
license, which they also give to illegal
00:24:07
aliens in order to be added to the voter
00:24:09
roles.
00:24:10
>> So, they are finding illegal aliens on
00:24:12
voter roles. But look, regardless of
00:24:14
where you are on cheating in elections,
00:24:17
the census just counts total population
00:24:20
and then they aortion house seats and
00:24:23
electoral votes based on total
00:24:24
population, including illegal aliens.
00:24:26
And there is data on this. Trump would
00:24:28
have won the last election by an
00:24:29
additional nine electoral votes if these
00:24:32
changes had been made before the last
00:24:34
election in 72.
00:24:35
>> Do you think uh
00:24:36
>> so just to put a pin on that point? It's
00:24:39
really important because what it means
00:24:40
is that a future Republican would not
00:24:44
need to somehow crack the Democrats blue
00:24:46
wall of Pennsylvania, Michigan, and
00:24:48
Wisconsin. Right? So nine electoral
00:24:51
votes is a lot and it does put the
00:24:54
presidency that much further out of
00:24:56
reach for future Democratic candidates.
00:24:59
It provides a powerful incentive and
00:25:01
it's again it's not just the electoral
00:25:02
votes. It's also these House seats.
00:25:04
California would have lost House seats.
00:25:07
I think it was like four if this had
00:25:10
been recognized because again there's
00:25:11
been a huge migration. Jake, you know
00:25:13
this better than anybody since you're
00:25:14
one of the people who've left. The state
00:25:16
is so badly mismanaged that people have
00:25:18
been fleeing blue states in droves. And
00:25:21
the fact that you've got illegal aliens
00:25:23
then replacing that population mass that
00:25:25
effect and allows these blue states to
00:25:28
maintain their level of representation
00:25:30
in the house. That is a huge incentive.
00:25:33
>> Freeberg, do you support
00:25:36
the use of ICE agents DHS to go into
00:25:40
these cities and do you think they were
00:25:42
too violent? And
00:25:45
do you agree with that strategy that
00:25:47
Trump's now seems like pulling back?
00:25:51
>> I mean, I know the answer. We talked
00:25:52
about it on private chat. Yeah.
00:25:54
>> Okay. I'll be very matter of fact.
00:25:56
>> Yeah.
00:25:56
>> Neither of these people should be dead.
00:25:58
It's sad that it happened.
00:25:59
>> Who's to blame?
00:26:00
>> Neither of these people should have been
00:26:02
doing what they were doing.
00:26:05
Federal law enforcement agents
00:26:08
should not wear masks.
00:26:10
They should identify themselves when
00:26:12
asked. Ideally, they wear body cams like
00:26:14
local law enforcement does. I like body
00:26:16
cams. I actually like watching videos of
00:26:18
police wearing body cams on YouTube and
00:26:21
Tik Tok. I think it's a very important
00:26:24
way to hold law enforcement accountable
00:26:26
to their work. There may be reasons why
00:26:28
federal law enforcement can't do that.
00:26:29
If they're undercover and whatnot, they
00:26:30
shouldn't have to identify themselves,
00:26:32
but I think those are important ground
00:26:34
rules. I think that federal law
00:26:36
enforcement should have warrant or
00:26:38
probable cause. They should not be
00:26:40
allowed to randomly ask people for
00:26:43
papers.
00:26:45
I don't agree with that. If that's what
00:26:47
they're doing, and I'm not saying that
00:26:48
that's what I agents are doing.
00:26:49
>> No, no, that's what they were doing.
00:26:50
That's all been proven.
00:26:51
>> I don't know that. I don't know if
00:26:52
that's been proven.
00:26:53
>> Wearing masks, not identifying
00:26:54
themselves.
00:26:54
>> That's not what I'm talking about. Hold
00:26:55
on a second. Hold on a second. There
00:26:57
have been a lot of claims made that
00:26:58
they're going around quote rounding
00:27:00
people up asking for ID and papers. I
00:27:03
don't know if that's true. I have
00:27:04
personally not been convinced of that.
00:27:06
You may have seen something else, but
00:27:07
I've not been convinced of that. But if
00:27:09
that is what they're doing, there is no
00:27:11
probable cause and there's no warrant
00:27:12
attached to that. That should not be
00:27:14
allowed. I think the mechanism for
00:27:16
fixing that, the mechanism for
00:27:18
addressing that is number one to go to
00:27:20
federal courts and file injunctions,
00:27:22
which is actively happening through the
00:27:24
legal process. And there's a due process
00:27:25
that then would take place that would
00:27:27
determine whether these agents are
00:27:29
following the law and doing things by
00:27:30
the rule of law or are not. And if
00:27:33
they're not, injunctions will be imposed
00:27:34
and they can't do what they're doing.
00:27:37
Short of that, no one should take it
00:27:39
into their own hands to obstruct law
00:27:40
enforcement, federal or local law
00:27:42
enforcement. I think that that is wrong.
00:27:44
I don't think that individuals that
00:27:46
disagree with a law or disagree with the
00:27:49
actions of law enforcement officers
00:27:51
should obstruct their actions. The
00:27:53
correct path is number one to protest
00:27:55
peacefully, number two to go to the
00:27:57
courts and file injunctions, and number
00:27:59
three to change the law. To go to your
00:28:02
local,
00:28:03
>> vote for someone and change the law. I
00:28:06
think it's totally okay for protesters
00:28:08
to not like the law. There's a million
00:28:10
laws and regulations and other [ __ ]
00:28:12
that I fundamentally disagree with, but
00:28:14
those don't give me the right to impose
00:28:17
myself physically in the driveway with a
00:28:20
car blocking law enforcement from doing
00:28:22
their job. If law enforcement was
00:28:24
arresting a non-ilgal immigrant murderer
00:28:28
and someone did that, would people be up
00:28:30
in arms? It's because people disagree
00:28:32
with the law. And if you disagree with
00:28:34
the law, you've got to change the law.
00:28:37
And that's okay. Now, right now, it is
00:28:40
the case that it is the law that if
00:28:42
someone came here without going through
00:28:44
the immigration process as defined by
00:28:46
the law, then they are technically here
00:28:49
illegally. That is the law. That's what
00:28:51
it is. And so, there is a course or a
00:28:53
path or a point of view on how do you
00:28:55
enforce that law and that is what people
00:28:57
are having disagreements over. And
00:28:59
again, file injunctions if you don't
00:29:01
like the methods, change the law if you
00:29:03
don't like the law. My personal view on
00:29:06
immigration, just so I can wrap this all
00:29:08
up, I don't see how we're going to do
00:29:10
this in a humane and just way of
00:29:12
removing people from this country who
00:29:13
have been here for a period of time and
00:29:15
have paid taxes and have been good
00:29:16
contributors to this country. I don't
00:29:19
know how you're going to do it. I don't
00:29:20
know how you're going to do it without
00:29:21
inciting a civil war. So I think the
00:29:23
compromise has to be that there has to
00:29:25
be a path to permanent residency and
00:29:27
eventually citizenship for individuals
00:29:29
who have been in the United States for
00:29:31
some period of time who have followed
00:29:33
other laws who have paid taxes and who
00:29:36
have not taken advantage of public
00:29:38
services that they are not supposed to
00:29:39
use. With those conditions met there
00:29:42
should be some path otherwise we're
00:29:44
going to have civil war. I think that
00:29:47
separate to this there's a bigger point
00:29:48
to be made and Ray Dalio has made this
00:29:50
and I will restate it and you guys can
00:29:51
roll your eyes at this but I think that
00:29:53
there's a deep emotional driver to all
00:29:55
of this. People look at this and they
00:29:57
get incited. They get activated
00:29:59
emotionally. Everyone I speak to is
00:30:01
activated emotionally over this issue
00:30:02
and we have to ask ourselves the
00:30:04
important question of why are people
00:30:05
activated so emotionally right now? What
00:30:07
is going on? What is the fundamental
00:30:10
root cause of this? What is the
00:30:11
fundamental root cause of this
00:30:13
prediction model that Ray Dalio has
00:30:14
talked about saying that there's a 35 to
00:30:17
40% chance of a civil war in the United
00:30:18
States based on predictions that he made
00:30:20
in 2020. And it is rooted fundamentally
00:30:24
in the fact that everyone to some degree
00:30:26
feels some amount of oppression right
00:30:28
now. And this is a manifestation of that
00:30:30
oppressed feeling. That oppression comes
00:30:33
from the fact that the world is racing
00:30:34
ahead and people are not a part of it.
00:30:36
that people feel like they're being left
00:30:38
behind and they see victims of that
00:30:40
world and they want to act. And I think
00:30:42
we need to pay attention to that and be
00:30:43
very cognizant of it because I think
00:30:45
fundamentally it is an inevitability
00:30:47
that there will be a civil war if we
00:30:49
don't recognize it and address it and
00:30:51
find some paths of compromise to solve
00:30:53
these problems.
00:30:54
>> Okay, I'm done.
00:30:55
>> Uh, superb.
00:30:57
>> So, it ship it. Episode done. I mean,
00:31:02
>> let's go home.
00:31:03
>> So, yeah, I'm strongly in agreement with
00:31:05
you. the police should um not be
00:31:07
interfered with. I said that here and I
00:31:09
said it on Twitter. You should stay home
00:31:10
if you can't peacefully protest and also
00:31:14
agree with you strongly. The police uh
00:31:16
should be trained to deescalate, wear
00:31:17
body cameras, not wear masks, be trained
00:31:19
properly, all that stuff.
00:31:22
But I I want to make three points here.
00:31:25
And the first is, you know, President
00:31:26
Trump has surrounded himself with a lot
00:31:28
of really competent people, yourself
00:31:30
included, David Rubio, Bessant, Lutnik,
00:31:33
Kushner. We had a lot of them here. my
00:31:35
pal Chris Wright, but he's also
00:31:37
surrounded himself with very
00:31:39
inexperienced sick offense that he
00:31:41
picked based on their loyalty to him and
00:31:43
that group needs to go because that
00:31:45
group is sinking his second term. I'll
00:31:48
put on that list Steven Miller, Christy
00:31:49
Gnome, Cash Patel, and Pam Bondi. Uh
00:31:52
these folks have been a disaster for
00:31:53
Trump. They're not qualified to be in
00:31:55
the positions they're in, and they have
00:31:57
caused a lot of chaos. And back in
00:32:01
August, I explained exactly how
00:32:03
unpopular these,
00:32:06
you know, ICE actions were. And I
00:32:08
implored people, hey, stay home if you
00:32:10
cannot be peaceful. If you want to
00:32:12
record people, that's fine. But I
00:32:14
predicted as well that somebody's going
00:32:15
to get killed. And that these agents
00:32:18
were acting just without training. They
00:32:20
were not deescalating. And they were in
00:32:22
fact provoking a lot of this which I
00:32:24
think was part of the concept that
00:32:26
Steven Miller had was to provoke these
00:32:28
kind of reactions. Nick, you can pull up
00:32:30
my first chart. You remember back in
00:32:32
October was talking about Trump's
00:32:35
sinking approval rating back when these
00:32:39
ICE issues were happening. The Epstein
00:32:42
files weren't being released. Then in
00:32:43
November, I brought this up again.
00:32:45
Trump's uh net approval rating hit 13%.
00:32:50
And then same chart coming up now. He's
00:32:54
uh 18%.
00:32:58
What's And you know, a lot of people
00:32:59
like to use this term taco. Trump always
00:33:02
chicken out chickens out. I don't like
00:33:04
when people use that because I think
00:33:06
they're trying to goat him on. I think
00:33:08
he needs to react to these plummeting
00:33:11
ratings here. If you could play the
00:33:13
Tillis quote, Nick,
00:33:14
>> if I were in her position, I can't think
00:33:16
of any point in pride over the last
00:33:18
year. She's got to make her own decision
00:33:20
or the president does. But she has taken
00:33:22
this administration into the ground on
00:33:24
an issue that we should own. We should
00:33:25
own the issue of border security and
00:33:27
immigration. But they have destroyed
00:33:29
that for Republicans. Something that got
00:33:31
the president elected. They have
00:33:32
destroyed it through their incompetence.
00:33:34
David Miller is in the same boat. this
00:33:36
guy after doing the stupid comments he
00:33:38
made about Greenland getting the
00:33:39
president in a difficult circumstance is
00:33:42
one of the people that came out publicly
00:33:44
and said that this guy was a terrorist
00:33:46
before he had even talked with anybody
00:33:47
on the ground and that's clearly not the
00:33:49
case now. So I mean it's just I mean
00:33:52
Steven Miller never fails to live up to
00:33:54
my expectations of incompetence.
00:33:56
>> And here's Marowski uh with a similar
00:33:59
opinion another Republican.
00:34:00
>> Senator, do you have faith in Christine
00:34:01
Noam as DHS secretary? I I've already
00:34:04
made a statement on that.
00:34:05
>> Oh, I I wasn't there for it.
00:34:07
>> Yeah, I I said that I've lost confidence
00:34:09
in her.
00:34:09
>> Do you think that President Trump should
00:34:11
remove her from the position? Do you
00:34:12
think she should resign?
00:34:13
>> Obviously, up to the president. Um I
00:34:17
think we would be we would be better
00:34:20
served
00:34:22
with leadership.
00:34:24
>> At the end of the day, Americans don't
00:34:27
like to see this violence. They don't
00:34:28
like cruelty. They don't like chaos. The
00:34:30
reason Trump lost to Biden, who's not a
00:34:33
very strong candidate, is because of the
00:34:34
general chaos people felt
00:34:37
with the immigrant ban the last time and
00:34:40
people do want to see the border closed
00:34:42
and that has had a great effect I think
00:34:44
on the country and I think that's
00:34:47
fantastic and he should take a victory
00:34:49
lap for that. But leadership starts at
00:34:52
the top. President Trump put Steven
00:34:54
Miller in charge of all this. Uh I think
00:34:56
he's a bad actor. I think these people
00:34:59
who he hired were picked strictly to do
00:35:01
lawfare and to do this kind of sadistic
00:35:04
violent behavior to feed a MAGA base
00:35:06
that's not going to keep Trump in office
00:35:09
and in fact going to cause him to lose
00:35:11
the second uh to lose the midterms. So
00:35:15
the easy solution which I brought up
00:35:17
here over and over again is if you want
00:35:20
to stop having people come into this
00:35:22
country, you have to look at why they're
00:35:24
coming to this country. They're coming
00:35:25
to this country because they want to
00:35:26
have a better life. That's why they're
00:35:28
coming here. And that's why they'll pay
00:35:30
a coyote 10 or 20 or $30,000 and risk
00:35:32
their own safety coming across that
00:35:34
border where many people die and abused
00:35:36
and are abused. So if you want to stop
00:35:39
this, all you have to do is go to the
00:35:40
business owners and find them. And if
00:35:42
they keep hiring illegal aliens, you
00:35:44
keep finding them and you put them in
00:35:46
jail. But you don't see Steven Miller
00:35:48
doing that. Why? Well, because a lot of
00:35:50
those businesses are the voters who put
00:35:52
Trump in office. They're Republicans.
00:35:54
Yeah. Sure, there's Democrat owned
00:35:55
business owners as well, but you can
00:35:57
solve this whole problem without sending
00:35:59
mass agents in to beat the [ __ ] out of
00:36:01
people, to be violent, to provoke these
00:36:03
kind of reactions. And that doesn't
00:36:06
absolve people driving their cars into
00:36:07
the police. That's horrible and it's a
00:36:10
terrible tragedy. And these people, as
00:36:12
I've said three or four times, uh should
00:36:14
not go out and protest. But you could
00:36:17
solve this. If there are no job
00:36:19
opportunities for illegal aliens, they
00:36:21
will stop coming here. finally agree
00:36:22
with your point strongly, Freedberg,
00:36:24
that we should be compassionate to the
00:36:26
people who have been here 10, 20 years
00:36:27
and paid their taxes. I've said this
00:36:28
here many times. They should be given a
00:36:30
path. America is a country built by
00:36:32
immigrants for immigrants, including the
00:36:34
three of you, my bestie immigrants.
00:36:38
>> And we all came here illegally or our
00:36:41
parents did
00:36:42
>> and we didn't seek into the country.
00:36:44
>> Yeah, absolutely.
00:36:45
>> Now, you kind of gave this diet trial. I
00:36:47
don't know why you've picked Steven
00:36:49
Miller as the vessel of your hatred.
00:36:51
It's like you've
00:36:52
>> No, it's not hatred. I based on
00:36:53
>> There's been a transference. There's
00:36:55
been a transference of your
00:36:56
>> trans just based on his TDS to Steven
00:36:58
Miller derangement syndrome or something
00:36:59
like that.
00:37:00
>> If he if his actions were if his actions
00:37:02
were kind and compassionate
00:37:05
>> if if his if he was compassionate
00:37:08
towards immigrants
00:37:10
and uh recognize the importance of
00:37:12
immigrants to this American story, I
00:37:14
would be fine with it. It's just based
00:37:15
on his behavior. David, I don't have
00:37:17
TDS. You can name call all you want. I'm
00:37:19
just basing it on your behavior.
00:37:21
>> You've decided to do this ad homonym
00:37:23
against him. And look, the big picture
00:37:24
is that threat is basing it on their I'm
00:37:27
basing it on their qualifications.
00:37:29
>> You're not really presenting You're not
00:37:30
really presenting evidence. You're
00:37:31
showing what other people think. You're
00:37:33
showing polling and then what a few
00:37:35
senators think. So basically, you're
00:37:37
trying to build a Hold on. You're Hold
00:37:39
on. Let me finish my point. You're
00:37:40
basically trying to build a case against
00:37:42
him by using the opinions and
00:37:44
declarations of other people rather than
00:37:46
actually presenting evidence and
00:37:48
building that case other than just a
00:37:49
name call and say that he's not
00:37:51
compassionate. The big picture is that I
00:37:53
would say that Steven Miller has been
00:37:54
more correct about immigration than you
00:37:57
have been over the last several years.
00:37:59
For years, you were denying that we had
00:38:01
an open border problem, did you not?
00:38:03
>> No. No. for one year when the spike went
00:38:05
up I said let's get the data on this
00:38:07
because it doesn't make sense that it
00:38:09
would triple year-over-year and then we
00:38:11
found out that it was in fact and with
00:38:13
the new information I changed my opinion
00:38:14
and said yeah this is obviously
00:38:16
happening
00:38:17
that was literally what happened
00:38:20
>> well I don't know I mean during the
00:38:21
buying years I remember we had a bunch
00:38:22
of debates about this and you were
00:38:24
saying
00:38:24
>> my position was let's see the data cuz
00:38:26
the data if you look I I I played this
00:38:28
chart on the podcast many times we
00:38:30
showed like the last four years and then
00:38:31
there was a spike and I said this
00:38:32
doesn't make any sense How did this
00:38:33
spike happen? It's just my opinion. You
00:38:35
don't have to have a you can have a
00:38:36
difference of opinion.
00:38:37
>> Let me go back to the origin of the
00:38:40
chaos in Minnesota. You don't see this
00:38:42
happening in other states. And I just
00:38:44
want to point out one of the major
00:38:45
reasons why. So in other states, you
00:38:48
know, take your pick. When you have an
00:38:50
illegal alien and they get arrested,
00:38:53
they get handed over to ICE or Border
00:38:56
Patrol for deportation. That's what's
00:38:58
happening in Minnesota. the politicians
00:39:00
have given an instruction not to do
00:39:02
that. So, literally, you have criminals,
00:39:05
people who've been arrested. There's a
00:39:06
case that was just um posted today of an
00:39:09
illegal alien in Minneapolis who killed
00:39:13
an innocent mother and severely injured
00:39:15
two others backto back in uh drunk
00:39:18
driving crashes. He did not have a valid
00:39:21
driver's license, did not have
00:39:22
insurance. He was arrested for vehicular
00:39:24
homicide. ICE asked him to be turned
00:39:27
over to them and the authorities in
00:39:30
Minneapolis refused to do that and they
00:39:32
eventually released him. So that puts
00:39:34
ICE in the position of having to go out
00:39:37
and find these guys and arrest them.
00:39:39
That's how all these operations started.
00:39:41
And you make it sound like these guys
00:39:43
are the Gestapo and they just randomly
00:39:45
round people up. That's not what's
00:39:46
happening. They have arrest warrants
00:39:48
with the names of specific people that
00:39:51
they're going after. There are arrest
00:39:52
warrants, but there were also Steven
00:39:54
Miller told them to just go round people
00:39:56
up at specific locations. So,
00:39:58
>> there is that as well. And they've been
00:40:00
trying to hit numbers that they just
00:40:01
can't possibly hit. You have to ask
00:40:03
yourself
00:40:05
>> why didn't we have you have to ask
00:40:06
yourself, I think, Sax, why didn't we
00:40:08
have this kind of violence and unmarked
00:40:11
people, breaking people's rights under
00:40:13
Obama who did far more?
00:40:14
>> I'll tell you why. You want to know why?
00:40:16
Because Obama said that we should
00:40:18
deport. Ultimately, our nation, like all
00:40:21
nations, has the right and obligation to
00:40:25
control its borders and set laws for
00:40:28
residency and citizenship. And no matter
00:40:30
how decent they are, no matter their
00:40:34
reasons. The 11 million who broke these
00:40:37
laws should be held accountable.
00:40:39
>> I'm saying he was able to do it without
00:40:40
sending masked agents in to beat the
00:40:42
>> No, the Democratic party changed. That's
00:40:44
what happened 15 years ago. the position
00:40:46
that Obama Can I just talking about the
00:40:48
policing? You ask me a question. Why
00:40:49
wasn't the policeing? Why didn't this
00:40:50
happen under Obama? Let me tell you the
00:40:51
reason why the policing happen years
00:40:53
ago. The Democrat party was still
00:40:55
somewhat rational on this issue and it
00:40:58
was basically a bipartisan position to
00:41:00
support deportations of illegal aliens
00:41:02
and not to have an open border. The
00:41:04
Democrat party changed their point of
00:41:05
view on this issue. Now,
00:41:07
>> that's not what I'm talking about
00:41:09
policing.
00:41:09
>> Let me finish my point. Tom Hman has
00:41:11
said over and over again, we don't want
00:41:12
to go out on the street. It's dangerous.
00:41:14
They don't want to be put in this
00:41:15
position. ICSE has an incredibly
00:41:17
dangerous job to do. Border Patrol has
00:41:19
an incredibly dangerous job to do. By
00:41:21
the way, the reason why they wear masks
00:41:22
is because they're getting doxed by
00:41:24
these groups who are organizing with
00:41:26
each other on signal and they're being
00:41:28
followed around. They're being stalked.
00:41:30
They're being chased to their hotels and
00:41:32
these guys follow them around. They
00:41:33
block their cars by um interposing
00:41:35
vehicles. They're interfering in in
00:41:38
official operations. They're interfering
00:41:40
with arrest and they're getting in
00:41:42
physical altercations with them. This is
00:41:44
very dangerous for eyes. They don't want
00:41:45
to be in this position. But the reason
00:41:47
why all of this is happening is because
00:41:49
Tim Waltz and Jacob Fry and the rest of
00:41:52
these guys, Lieutenant Governor, they
00:41:53
are saying, "Do not cooperate with ICE.
00:41:56
Do not cooperate with the border
00:41:57
patrol." Telling local authorities in
00:41:59
Minnesota, when you arrest an illegal
00:42:01
alien, do not turn them over to ICE.
00:42:03
It's better to release them. These
00:42:05
people are such anti-ICE zealots that
00:42:07
they would rather release a killer onto
00:42:11
the street than keep them in jail where
00:42:14
they can be turned over to ICE. And
00:42:16
that's the whole reason why they don't
00:42:17
want to keep them in jail is because
00:42:18
that would create a central repository,
00:42:20
if you will, for ICE to go collect all
00:42:22
these illegal aliens. Last year there
00:42:24
was something like 470 convicts, okay?
00:42:28
Illegal aliens, criminals. And I'm not
00:42:30
talking about that their crime wasn't
00:42:32
breaking into the country. I'm talking
00:42:33
about subsequent crimes that were
00:42:35
released
00:42:37
in Minnesota so they they cannot be
00:42:39
rounded up by ICE and deported. These
00:42:41
people are zealots and that is the
00:42:43
source of the problem. And then on top
00:42:45
of that, you've got rhetoric by people
00:42:47
like Tim Walls and the lieutenant
00:42:49
governor calling ICE and Border Patrol
00:42:52
agents who are just trying to do their
00:42:53
jobs the Gestapo and Nazis and riing
00:42:57
people up. And I think that people like
00:43:00
Alex Pretty or Renee Good, they're
00:43:01
victims of this type of rhetoric. These
00:43:04
are left-wing activists who are in this
00:43:07
leftwing echo chamber and they're
00:43:09
baronating in all this rhetoric that's
00:43:13
portraying law enforcement officials as
00:43:15
Nazis and they're embibing this
00:43:17
constantly to the point where they're
00:43:19
full of rage and anger. You see that
00:43:20
video of Alex Pretty attacking Border
00:43:23
Patrol. This is two weeks before he gets
00:43:25
killed. He's kicking the car. He's like
00:43:27
foaming at the mouth. He's in some sort
00:43:29
of rage. This is someone who's
00:43:30
emotionally disturbed and he's got a a
00:43:33
gun on his waistband, which look, every
00:43:36
gun owner knows that you have to be
00:43:38
exceedingly careful if you are carrying
00:43:41
a weapon and then you are dealing with
00:43:43
law enforcement. The last thing you want
00:43:45
to do is go out looking for an
00:43:46
altercation with them, which is what he
00:43:48
did. And by the way, I think it's
00:43:49
completely tragic that the two of them
00:43:52
were killed. I don't want to see that
00:43:53
happen. But again, they are victims of
00:43:56
this political environment, this chaos
00:43:58
that's being whipped up by the elected
00:44:00
officials in Minnesota. This would not
00:44:02
be happening if they simply turned over
00:44:04
the illegal aliens in their custody to
00:44:06
ICE and Border Patrol, which is all that
00:44:08
they're asking for.
00:44:09
>> I've said on this podcast like many
00:44:11
times, but none of you will respond to
00:44:12
it. Why are they doing this instead of
00:44:14
just finding the business owners?
00:44:16
Wouldn't that be a better policy? And
00:44:17
you're in the White House, you're around
00:44:18
these people. Why don't you tell them
00:44:19
why don't we do that? Why don't we just
00:44:21
give fines to business owners who hire
00:44:24
illegals and stop the incentive to come
00:44:26
here?
00:44:27
>> How do you know that they hired
00:44:28
illegals? How would you know that? And
00:44:30
if you know that, then you know the
00:44:32
individual.
00:44:32
>> It's super simple. You would just go do
00:44:35
uh a little investigation. You go to the
00:44:37
car wash, you videotape it like any
00:44:39
other detective or any other federal
00:44:41
agent or a local police officer would,
00:44:43
and then you track those people down.
00:44:45
You very quietly go pick one or two of
00:44:47
them up on the way home. You ask them
00:44:48
for their papers or you go to the owner
00:44:50
and say, "Hey, we have these six people.
00:44:51
Here's their pictures. Here's the
00:44:53
pictures of the six people. Show us
00:44:54
their papers. Show us their social
00:44:56
security number." That's all you have to
00:44:57
do. Freeberg,
00:44:59
>> here's my here's my here's my proposal.
00:45:01
Okay, Jacob, you may have a point there,
00:45:02
Jacob. You may have a point, but I'm not
00:45:04
sure
00:45:05
>> that it's good enough in the case of an
00:45:07
illegal alien who's already been in this
00:45:09
country committing crimes. Okay. Yeah.
00:45:11
So, if you have if you hold on, if you
00:45:12
have a drunk driver
00:45:14
>> who has killed people through drunk
00:45:16
driving and they're they're captured, I
00:45:18
think they should be turned over and
00:45:19
deported immediately.
00:45:20
>> But if you want to do big numbers, like
00:45:21
big numbers, you could just go to a
00:45:23
farm, you could go to a car wash, you
00:45:25
could go to a restaurant, you take
00:45:26
pictures of everybody coming in and out
00:45:28
for their shift, you go to the business
00:45:29
owner and say, "We have these people.
00:45:31
They can you show us their uh paychecks
00:45:34
and their pay stubs?"
00:45:36
>> All right. So, we'll agree to disagree
00:45:38
on some of this and uh we'll agree that
00:45:41
this is a tragedy and that we need
00:45:43
leadership to calm these things down. Uh
00:45:46
and we need to agree on a very
00:45:48
reasonable immigration policy which is
00:45:51
not what we have right now in my
00:45:53
opinion. All right. Claudebot has gone
00:45:55
viral overnight. I have been claed. I
00:45:58
have been one claude. I am all in on
00:46:00
this. Um this is an open-source project
00:46:03
created by a gentleman named Peter
00:46:05
Steinberger. He's an Austrian developer
00:46:07
and entrepreneur.
00:46:09
So, what is it? It's uh basically an
00:46:11
open-source personal assistant. Uh think
00:46:14
like Siri or maybe Jarvis. Have any of
00:46:17
you guys used it yet? Has anybody
00:46:19
installed it yet?
00:46:20
>> I did. I spent 15 minutes and I saved
00:46:22
15% on my car insurance.
00:46:25
>> Are you joking? Sure. You installed it
00:46:26
and then asked it to go.
00:46:27
>> You know what's so funny, Nick? I posted
00:46:29
this. The number of people that didn't
00:46:31
understand that that was a joke.
00:46:33
>> No, I understand it's a joke. Yeah,
00:46:34
>> I know you did. Yeah. Yeah.
00:46:35
>> But there was a bunch of people that
00:46:37
were like, "Really? How did you do it?
00:46:38
Can I do it?" And then some people were
00:46:40
like, "Wait, only 15%." And other people
00:46:42
were like, "Wait, you set it up
00:46:43
yourself. You didn't have somebody help
00:46:44
you."
00:46:46
>> This is the great This is This is
00:46:47
literally the greatest thing ever. Have
00:46:49
you installed it yet, Freedberg?
00:46:51
>> No, I don't want to give
00:46:53
>> Okay.
00:46:53
>> an open source tool access to all my
00:46:55
emails and messages.
00:46:56
>> Have you done it yet?
00:46:58
>> No, because I'm concerned about the
00:46:59
security issues. But I want to I want to
00:47:01
do it, but I'm concerned about the
00:47:02
security.
00:47:02
>> I've spent the last 72 hours doing this.
00:47:05
I'm going to explain to you what we did
00:47:06
at the company. It's mind-blowing. Okay,
00:47:09
so it is basically think like an open-
00:47:12
source Siri or Jarvis, right? You get
00:47:14
into an interface and you can talk to a
00:47:16
virtual assistant and it does things for
00:47:17
you just like chat GPT or XAI might. But
00:47:21
the way it works is you kind of load a
00:47:24
virtual machine if you know what that is
00:47:25
or you can put it on your Mac Mini. You
00:47:27
run this like a server. Then you start
00:47:29
authenticating it with your services.
00:47:31
So, Gmail, Notion, Slack, WhatsApp,
00:47:34
maybe even your password manager. Super
00:47:36
dangerous. Nobody like Claude or XAI or
00:47:40
Microsoft would ever allow you to do
00:47:42
this because it's so dangerous,
00:47:44
obviously.
00:47:46
>> But we did this and we put it to work.
00:47:48
Here's what we did. I created a virtual
00:47:51
podcast producer and we made this
00:47:54
persona
00:47:56
and we created a new Gmail account,
00:47:59
Sachs, a new notion account, a new
00:48:01
WhatsApp account, everything. We created
00:48:03
like basically an a virtual employee. We
00:48:06
put it together and we made it the
00:48:08
producer for my new This Week in AI
00:48:10
podcast. There's a little plug in there
00:48:12
for it. So, we had it start doing
00:48:15
research on guests. So, we said, "Hey,
00:48:17
research guests, right?" just like you
00:48:18
can do in an LLM. Then we connected it
00:48:21
to um the existing feeds like the
00:48:25
podcast feeds and the database of like
00:48:27
people we've had on the show and like
00:48:28
who's booked. Then we said make a CRM
00:48:32
Sachs for all potential guests and
00:48:33
suggest other guests and it vibe coded a
00:48:36
CRM for itself.
00:48:38
>> Can you show this?
00:48:40
>> I said hey do some research on this.
00:48:41
Right. So then we start a thread with it
00:48:43
with the actual producer Oliver who's
00:48:46
working on this. So I said, "Do this
00:48:47
guest research." It did all this guest
00:48:48
research. Now, this is as good as like
00:48:50
Nick or Lisa would do at a first pass,
00:48:53
but it did it instantly. And then I made
00:48:55
a prompt for it to like, and I was using
00:48:57
this prompt, by the way, David, when we
00:48:58
were in Davos for our guests. And so
00:49:01
this like gets the company's name, the
00:49:03
founder information, you know, look for
00:49:05
their competitors, all the stuff that a
00:49:07
producer would do over a day or two. Do
00:49:09
a timeline, give suggested questions,
00:49:11
etc. So I teach it this. So, here's
00:49:13
producer X and it says, "Got it. That's
00:49:17
what uh that's a great guest research."
00:49:20
And I said, "Okay, do the guest research
00:49:21
on this person." It comes back with
00:49:22
that. And it gave me its media
00:49:24
appearances, all the stuff I would want
00:49:25
to do for research. Then I said, "Um,
00:49:28
uh, I think this would be
00:49:30
>> You're interacting with it like an
00:49:31
agent." So, how is it different than an
00:49:32
agent?
00:49:33
>> Watch. Okay, we're getting to it. So,
00:49:35
then I said, um, producer X, email Alex
00:49:39
and let him know that I want to have him
00:49:40
on this weekend in AI. CC me and CC
00:49:42
Oliver. And it says, uh, I couldn't
00:49:45
verify the email, but I'm going to try
00:49:47
this one. And it wrote this email. It
00:49:49
sent the email. Then after it sent it,
00:49:53
>> did he respond?
00:49:54
>> So here's the email to the guest.
00:49:56
>> Oh my god.
00:49:57
>> Hey, Alexis
00:49:58
wanted me to reach out. He'd love to
00:49:59
have you on this weekend to discuss
00:50:01
ExoLabs, your work on distributed
00:50:03
inference on the recent he like did this
00:50:04
like little thing. Let me know if you'd
00:50:06
be interested.
00:50:08
>> So then
00:50:09
>> it's [ __ ] crazy. He says,
00:50:11
>> this guy says, "Sounds great. I'll be in
00:50:13
touch."
00:50:14
>> Then I said, "Here's what I want you to
00:50:16
do. I want you," and we created a group
00:50:19
called replicants. So now we've made
00:50:21
five replicants at the company. And
00:50:23
today is like replicant day or replicant
00:50:25
week. So I told it every time you do
00:50:28
work for me in the replicant room, say
00:50:31
what you did and then also put it on
00:50:33
your calendar. So now it's putting on
00:50:36
its calendar what it did in timestamps.
00:50:39
So he said I did some guest research
00:50:41
here and then yesterday I did this
00:50:42
booking report and that I'm doing this
00:50:45
AI around it that I'm doing this
00:50:47
>> and then I set up a and then producer
00:50:48
lawn had it set up a ticker because we
00:50:50
did this email ticker and did all that
00:50:52
work
00:50:53
>> and the crazy thing is you can start
00:50:55
talking to it directly so I can have a
00:50:58
direct conversation with producer X and
00:51:02
it learns and now it knows everything.
00:51:04
So we started one to be an SDR and then
00:51:08
we are giving it access to our CRM
00:51:10
there. But it built its own CRM, David.
00:51:12
It's building its own SAS tools to solve
00:51:14
its problems. Every time you add it to
00:51:16
something, it gets smarter. So now we've
00:51:19
we're making a LinkedIn for this
00:51:20
persona.
00:51:21
>> The LinkedIn is going to start adding
00:51:23
people. People don't know that these
00:51:26
aren't humans.
00:51:27
>> Yeah,
00:51:28
>> it's insane.
00:51:29
>> All right, let me give you a few
00:51:30
thoughts. So, I was paying close
00:51:32
attention to the whole Claudebot
00:51:33
freakout, you know, over the weekend.
00:51:35
And, you know, again, I wanted to set it
00:51:37
up really badly, but I was too afraid of
00:51:39
the security issues to do it, but I was
00:51:41
watching everyone else do it. And look,
00:51:43
I think the takeaways for me are number
00:51:44
one that this will be the year of the
00:51:47
rise of the personal AI assistant. Until
00:51:50
now, AI has mainly come in the form
00:51:52
factor of a chatbot and it's been used
00:51:55
as a research tool. Some people have
00:51:59
>> search. Yeah,
00:52:00
>> it's better search. There's kind of this
00:52:02
niche case of, you know, fantasy chat
00:52:04
bots where people actually like somehow
00:52:06
talking to a chatbot, but really it's
00:52:08
about better web search and it's about
00:52:10
research. Now, we're moving to a
00:52:11
completely different form factor, which
00:52:14
is again this AI assistant that's an
00:52:16
agent that can do things for you and
00:52:18
it's going to get better and better at
00:52:20
doing different tasks for you.
00:52:21
Everyone's going to have this amazing AI
00:52:23
agent. Now, you know who who is the
00:52:26
beneficiary of this market opportunity?
00:52:28
I mean there's obviously going to be
00:52:29
some startups like the guy who's doing
00:52:31
Claudebot which I guess he renamed it
00:52:32
was it moldbot or
00:52:35
>> it's called Maltbot now because Claude
00:52:37
got upset
00:52:37
>> right so in any event it's Maltbot now
00:52:39
but I think it's a tremendous
00:52:40
opportunity for Google because the
00:52:42
question is who has your data Google has
00:52:44
all my email my calendar my documents
00:52:46
which is exactly the stuff that I would
00:52:48
want to integrate with Cloudbot if I
00:52:49
believed it was safe obviously I've
00:52:51
already made the determination that
00:52:53
Google is safe because they already have
00:52:54
all my data so I think Google is in a
00:52:57
tremendous
00:52:58
position to offer a personal AI
00:53:00
assistant that's connected with your
00:53:01
email and calendar for those of us who
00:53:03
are using G Suite and so on. Obviously,
00:53:06
there'll be opportunities for other
00:53:07
companies as well, but I think this is
00:53:09
going to be such a big product that it
00:53:12
may become the dominant form factor of
00:53:14
AI, meaning more popular than the the
00:53:17
sort of researchoriented chatbot.
00:53:19
Speaking from the policy point of view,
00:53:21
I think it's going to change the policy
00:53:22
debate around AI because so much of the
00:53:25
policy debate over AI is about fighting
00:53:29
the last set of battles around social
00:53:31
media
00:53:32
>> question and answer modality.
00:53:33
>> It's sort of treating AI as like a form
00:53:35
of social media. I mean, I hear policy
00:53:38
makers say all the time, what are you
00:53:39
going to do to protect kids against
00:53:40
predators? And it's like, well, wait,
00:53:42
what predators are you talking about?
00:53:43
This is not social media, right? I mean,
00:53:46
I understand if you want to protect kids
00:53:47
against, I don't know, a fantasy chatbot
00:53:49
that's recommending they do bad things,
00:53:51
but it's just it's very different,
00:53:52
right? This is just a completely
00:53:54
different modality. And I think that the
00:53:56
rise of agents will make that clear that
00:53:58
we're dealing with something that's not
00:54:00
social media. Maybe there's some
00:54:02
analogies to the previous set of policy
00:54:04
battles that were fought over social
00:54:06
media, but I think it's quite different.
00:54:08
So, I think Claudebot is a breakthrough.
00:54:11
I think it's a really interesting proof
00:54:13
of concept and we'll also have to see
00:54:15
exactly who who the big winners are
00:54:17
here.
00:54:18
>> I think what's interesting is that if I
00:54:20
said to you, hey, there's going to be an
00:54:22
AI personal assistant. You would have
00:54:25
some point of view in your mind on what
00:54:27
that means. It's like, oh, it can, you
00:54:29
know, change my calendar,
00:54:31
tell me about my travel, read my email
00:54:34
for me, summarize that that kind of
00:54:35
stuff. But what I think people don't
00:54:37
realize, it's almost like the first time
00:54:39
you use FSD in your Tesla or the first
00:54:41
time you used an iPhone, you realize
00:54:43
that it's so much more that it widens
00:54:45
the aperture of what's possible that
00:54:47
it's not just the assistant in the way
00:54:49
that you might otherwise be thinking
00:54:50
about it, but it's like this super
00:54:52
worker. And the super worker, like to
00:54:54
Jal's point, it it does both scheduling,
00:54:56
calendaring, ideation, knowledge work,
00:54:59
creates new code, creates CRM tools,
00:55:02
books your travel, it does everything.
00:55:04
And then if you start thinking about
00:55:06
having an army of super workers, you're
00:55:08
like, "Oh my god, what's possible?" And
00:55:10
now think about what Elon's doing.
00:55:11
Yesterday, Elon shut down the Model S
00:55:14
and Model X production lines in Fremont.
00:55:17
>> And by the way, you know, a year ago,
00:55:18
I'd never owned a Tesla. And I was so
00:55:20
blown away with FSD that in the last
00:55:22
week, we bought two more Model X's. So
00:55:24
now everyone in my family, we're all on
00:55:26
Tesla because the experience is so good.
00:55:29
But they're shutting down the Model S,
00:55:30
Model X lines to make Optimus. So if
00:55:32
you've got the same thing that Jal's
00:55:34
talking about and experiencing, but
00:55:36
you've actually got it in physical form
00:55:38
in addition to digital form, that's the
00:55:40
future where everyone has a super
00:55:43
workforce and you put two optimists in
00:55:44
your garage and they build and run a
00:55:46
business for you and then everyone
00:55:48
becomes an entrepreneur.
00:55:49
>> What was my prediction, Jason? What of
00:55:51
my prediction?
00:55:51
>> Oh, that they would merge all these
00:55:53
companies and then this week
00:55:54
>> it's happening.
00:55:55
>> It's happening apparently. Um or I
00:55:57
shouldn't say it's happening. We don't
00:55:58
know it's happening, but beep
00:56:00
>> boop.
00:56:02
Between this and my copper prediction,
00:56:04
>> I'm about to retire.
00:56:05
>> You know, copper's up 26% in a month.
00:56:10
>> What is that? Is an annualized.
00:56:12
>> Check this out. S
00:56:14
>> I we invited somebody from Craft
00:56:16
Ventures on and I forgot who's from
00:56:17
Craft Ventures who's going to be on this
00:56:19
week in AI. So, I said, "Can you tell me
00:56:22
who's doing it?" And it said, "Oh, I
00:56:24
couldn't find it." I said, "Oh, it's in
00:56:25
the notion database." So, it goes and it
00:56:27
found it. Oh, Mike Robinson from Craft
00:56:28
Ventures is on February 24th. And then
00:56:30
there's also Brian from Craft or
00:56:33
whatever. And he was on a different
00:56:35
liquidity round table. And I said, "Hey,
00:56:37
can you email Mike and I want to do a
00:56:39
pre-show call with him?" Guessed his
00:56:41
email, sent it. I said, "C Heidi, I
00:56:43
haven't given him an access to my
00:56:44
stuff." And it said, "Oh, do you want to
00:56:45
get his top three times, all this
00:56:47
stuff?" It's crazy what's going on here,
00:56:49
folks. I would say out of 50 hours a
00:56:52
producer does a week, this does 40 of
00:56:55
them. And of what an SDR does, this does
00:56:59
95%.
00:57:01
This is going to be crazy because it
00:57:04
keeps learning. Now, your API bill is
00:57:07
going to be nuts. The first day we did
00:57:09
this, we hit like a $100. I think it's
00:57:11
going to be like $1,000 a day in API
00:57:12
calls. So, my team's like, "Hey, this is
00:57:14
a lot of API calls because people are
00:57:16
going crazy inside Slack and making all
00:57:19
of these agents."
00:57:20
>> Good news, Kimmy 2.5. Yeah.
00:57:23
>> Uh we are ordering Mac um studios, the
00:57:27
really powerful studios with like max
00:57:29
memory on and they're putting Kimmy. Is
00:57:31
that the new open source one?
00:57:34
>> Yeah, we're putting Kimmy on them. We're
00:57:36
going to have free
00:57:38
free for and then it according to people
00:57:41
online like 95% of all these queries can
00:57:43
be done for free with Kimmy.
00:57:44
>> You'll save 90%.
00:57:46
>> It's crazy.
00:57:47
>> It's really crazy.
00:57:48
>> This is independent by the way of
00:57:50
whatever LLM you want to use. You can
00:57:52
swap out LLMs. You could route it to
00:57:53
different LLMs. You can do whatever you
00:57:55
want with it. I tried to have it create
00:57:57
Reddit accounts for me. Like, I want a
00:57:59
Reddit account to go do research. And
00:58:00
it's like, I can't do that. It's against
00:58:02
the term of service. I don't think
00:58:03
people understand how important this
00:58:05
Kimmy K2.5 moment is.
00:58:07
>> Wait, what is Kimmy K2?
00:58:08
>> Just to set some context, right? So, I
00:58:11
think the last few years we've all kind
00:58:15
of lived in a world of what I would call
00:58:17
blackbox AI, meaning you go to your
00:58:20
favorite chatbot, basically, you put in
00:58:23
a prompt and you press enter, right? And
00:58:25
all of those things go to proprietary
00:58:27
models and they're excellent. OpenAI has
00:58:31
some, Anthropic has some, Google has
00:58:33
some, and they all give you back an
00:58:35
answer. Super powerful. But the
00:58:37
important thing that we don't know to
00:58:39
care about right now is that all of that
00:58:41
stuff is gated. What does that mean? You
00:58:43
don't own the keys. You don't own the
00:58:45
blueprints. You have no idea what's
00:58:46
actually going on. And what Claudebot
00:58:49
demonstrated this week is you're one
00:58:52
terms of service update away from
00:58:54
everything breaking, right? Because at
00:58:55
one point, Anthropic didn't like what
00:58:57
was going on and they said, "No, this is
00:58:59
not allowed." So,
00:59:02
we've talked about this battle between
00:59:04
open source and closed source. So all of
00:59:06
the models that that have been winning,
00:59:08
the blackbox models are closed source
00:59:10
models. But open source is important
00:59:12
because it's transparent. It gives
00:59:15
everybody their own sovereignty whether
00:59:18
you're a company and frankly really more
00:59:21
importantly whether you're a country. It
00:59:23
gives you control of your own speed. It
00:59:25
gives you a lot of execution control.
00:59:28
You can audit the weights of the models.
00:59:30
It allows you to host it on your own
00:59:31
hardware. And the most important thing
00:59:34
is the data never leaves your control.
00:59:35
So that's why open source was really
00:59:37
important, but it was always kind of
00:59:39
like an underdog and it wasn't
00:59:40
particularly good. So this week you wake
00:59:43
up, you go to the office and Kim K 2.5
00:59:45
is important. So this is why it's so
00:59:47
important because it was incredibly
00:59:50
profound. It's a trillion parameter
00:59:53
mixture of expert model.
00:59:57
When you farm out work, the proprietary
01:00:00
models keep that agentic layer kind of
01:00:04
secret.
01:00:05
Kimmy Cape 2.5 was like okay look here's
01:00:08
this thing called agent swarm it's a
01:00:10
technology that we built that's also now
01:00:12
public and it allows you to create all
01:00:14
you know 100 sub aents and what that
01:00:17
allows you to do is basically solve any
01:00:19
complicated
01:00:20
multi-step problem in parallel so I
01:00:23
think this is the moment now if I had to
01:00:26
make a prediction
01:00:28
I think there is the clear shot across
01:00:31
the bow of closed source and I think
01:00:33
open source can
01:00:35
Why?
01:00:35
>> Because when Kimmy K2.5 is accessible,
01:00:40
>> it democratizes something this trillion
01:00:42
parameter reasoning
01:00:44
>> that right now you could not otherwise
01:00:46
get. Now you can do vision to code, you
01:00:50
can do massive context windows. It's
01:00:52
really unbelievable and it's available
01:00:54
to everybody.
01:00:56
>> And to just build on it, you're exactly
01:00:58
correct and I think it's a very big
01:01:01
deal. You can run it. This is the other
01:01:04
thing. What What have the two limiting
01:01:05
factors been here?
01:01:06
>> I'm not a big fan of like these locally
01:01:08
hosted models. I think it's all [ __ ]
01:01:09
and janky. It's all like kids mucking
01:01:11
around.
01:01:14
>> I think running these things locally are
01:01:16
stupid and janky.
01:01:17
>> The whole point of open source is to go
01:01:19
and take them into these huge data
01:01:21
centers is to use nextgen silicon. And
01:01:24
we talked about this last week or the
01:01:26
week before where again post Grock what
01:01:29
I think will happen is you're going to
01:01:31
see an explosion of decode silicon. If
01:01:34
you take these next generation systems
01:01:36
and you marry them to open source you're
01:01:39
going to cut the cost of AI by 90%.
01:01:42
And when you do that you know Jason your
01:01:44
bill is going to be 10 bucks a day.
01:01:47
>> Yeah.
01:01:47
>> And you're just not going to stop.
01:01:49
Here's the interesting thing for me like
01:01:51
back to this like running it like there
01:01:53
are now people who figured out how to
01:01:55
daisy chain the Mac studios. So people
01:01:58
are you can see there's two stack there.
01:02:00
People are starting to stack these. This
01:02:02
is commodity hardware running open
01:02:04
source. The there's two advantages to
01:02:06
this one. We have control of all the
01:02:09
data. We have it on our own hardware. We
01:02:11
can run it infinitely. And these things
01:02:14
are only getting better. the M5 chips
01:02:15
coming to these Mac studios and if you
01:02:17
want more power you stack it up and then
01:02:19
the open source models are getting
01:02:20
easier. So what's going to happen is 90%
01:02:23
or 95% of our jobs are going to go to
01:02:25
this local hardware will control it. We
01:02:28
don't have to worry about our
01:02:29
information going up to Samman I canled
01:02:32
I canled all of my OpenAI accounts.
01:02:36
$25,000 gone. It's a matter of time
01:02:38
until the big model makers create an
01:02:41
incremental revenue stream for guys like
01:02:43
you, Jason, to license back all your
01:02:45
prompt and response data. And you'll
01:02:47
probably make enough to pay for the all
01:02:49
the costs of hosting and running these
01:02:51
models. Anyways,
01:02:52
>> yeah. Anyway, I this is like feels like
01:02:55
>> it's a very big moment.
01:02:56
>> It's a big mo it feels like a very big
01:02:57
moment. And then, you know, let alone if
01:03:00
when do you think freeberg we get it on
01:03:03
here? When do we get it on our phones?
01:03:05
when we'll be running Kimmy.
01:03:07
>> We're not even talking about a lot of
01:03:08
the architectural changes that are
01:03:10
happening that we've talked about in the
01:03:11
past. There papers that indicate we
01:03:13
could probably go down by 70 to 100x in
01:03:16
terms of compute need in how the the
01:03:18
model actually runs. So yeah, ultimately
01:03:21
these things end up on the on the iPhone
01:03:23
running locally and you don't need to go
01:03:25
to the cloud. By the way, just having
01:03:27
our nation's AIS are on the line here, I
01:03:30
think this has a dramatic effect. If you
01:03:32
play all of these paps out on the way a
01:03:35
lot of these states have written their
01:03:36
idiotic legislation where those those
01:03:39
legislative approaches encompass
01:03:42
strictly the view of AI being a chatbot
01:03:44
interface run by a single company
01:03:46
running a model in a data center. And if
01:03:48
all of the models end up running locally
01:03:50
on machines in open source, different
01:03:53
contexts, different use cases, all of
01:03:55
like we said from the beginning, all of
01:03:56
the [ __ ] idiotic, dumbass takes by
01:03:59
legislators on what they think they need
01:04:01
to regulate and how go out the window or
01:04:03
they create a lot of confusion on what's
01:04:05
actually going on in the real world and
01:04:07
it puts us at risk. I strongly endorse
01:04:10
the effort by those in the
01:04:12
administration to pass a federal AI
01:04:14
preeemption law that avoids all of this
01:04:16
nonsense in the states and the local
01:04:17
governments. And I think that this
01:04:18
evolution of AI from being centrally
01:04:22
hosted in data centers by closed models
01:04:24
through chatbot interfaces. All of those
01:04:27
layers break. You start to recognize
01:04:29
very quickly why you need to have
01:04:31
federal preeemption on this stuff
01:04:32
because people get way too ahead of it
01:04:33
and it's going to limit innovation.
01:04:35
>> Yeah. Did you just Well, obviously I
01:04:37
agree I agree with that. And again, just
01:04:39
to build on that point, you can't
01:04:41
underestimate how much the policy debate
01:04:44
in Washington has indexed on a single
01:04:48
use case.
01:04:49
>> That's what I'm saying.
01:04:49
>> Which was a niche fantasy chatbot
01:04:53
application,
01:04:54
>> right?
01:04:54
>> Yeah. That's all that people know. Yeah.
01:04:56
That's all they know.
01:04:57
>> That turned into horror stories. I mean,
01:04:58
legit horror stories. But
01:05:01
>> again, that is just not the predominant
01:05:03
use of AI. And it's certainly not going
01:05:05
to be in a year.
01:05:06
>> One way to think about it is imagine if
01:05:08
the internet came out and the only thing
01:05:09
that happened on the internet for the
01:05:11
first three years was like pornography
01:05:13
websites and like just like and then
01:05:15
people are like, "Okay, we got to
01:05:17
regulate the internet." And it's like,
01:05:18
"Hold on a second. This is an idiotic
01:05:20
use case. Perhaps we should take a zoom
01:05:22
out and think about what the internet
01:05:23
could enable and all of the other use
01:05:25
cases." And legislation needs to be
01:05:27
crafted with a bigger view. And perhaps
01:05:29
it's a little too early to make those
01:05:31
judgment calls.
01:05:32
>> Yeah. You got to wonder what impact this
01:05:34
is going to have on the valuations of
01:05:37
these companies. I mean, if you can run
01:05:39
this stuff locally on your own hardware,
01:05:42
you can bracket or put it in a virtual
01:05:43
machine and open source wins, it's going
01:05:45
to change the economics of everything.
01:05:47
And the people I'm talking to in
01:05:48
startups, which is where you always see
01:05:50
the most efficient use of technology,
01:05:52
they're all using Kimmy and what was the
01:05:55
other?
01:05:55
>> Let me ask you a question about that
01:05:56
actually, Chimoth. So Kimmy K2 powerful
01:06:00
model but it is a Chinese model or at
01:06:02
least it originates in China obviously
01:06:04
once it's open source people can fork it
01:06:06
and make their own but does that concern
01:06:09
you particularly
01:06:11
around the use case of coding because
01:06:14
what if there was a secret prompt in the
01:06:17
model
01:06:18
>> zero day attack or something
01:06:19
>> well it's like something built into the
01:06:20
model that's secret that could inject
01:06:22
corrupted code and you know so much
01:06:25
code's now being written by these models
01:06:28
in such volume that I mean humans in the
01:06:31
beginning were checking it all. But now
01:06:32
even the founder of Cloudbot said he he
01:06:34
doesn't even check all the code because
01:06:35
he can't personally supervise all of it.
01:06:37
So as AI coding becomes a bigger and
01:06:39
bigger percentage. Right now it's
01:06:41
probably what like 50%. As it goes to
01:06:43
99% no one's even checking it. You have
01:06:45
to really worry about prompt injections.
01:06:48
>> Exactly.
01:06:49
>> Code injections.
01:06:50
>> I think you're bringing up an important
01:06:51
point. Right now we overly rely on eval
01:06:54
to tell us how good a model is and I
01:06:57
don't think we've developed a standard
01:06:59
to I mean the the big model companies do
01:07:01
it internally but the safety teams
01:07:05
who are responsible for red teaming
01:07:07
these models don't really work as a
01:07:10
broad coalition. Everybody has their own
01:07:12
version of what they do to make sure
01:07:13
that their own models are good and
01:07:14
performant. I think that there's an
01:07:17
opportunity because somebody has to be
01:07:19
able to take an open source model. Let's
01:07:21
just say you're France and you're like,
01:07:24
"Wow, where am I in this whole AI race?
01:07:26
I'm nowhere.
01:07:28
We have a bunch of applications that we
01:07:30
want to develop and we need our own
01:07:32
sovereign AI stack." And so, okay, we'll
01:07:35
take Kimmy K 2.5. What do they do to
01:07:38
your point Sachs to get complete
01:07:41
assurance that that model is reliable
01:07:43
and safe under all weather conditions?
01:07:46
Honestly, the answer is I don't know
01:07:48
what they would do right now. But that's
01:07:50
the opportunity because somebody has to
01:07:52
be able to say, "Okay, look, we're going
01:07:54
to sandbox it in the following way.
01:07:55
We're going to run it under all these
01:07:57
race conditions. We're going to get to
01:07:58
all these corner cases so that we can
01:08:00
tell you that it's actually good to go
01:08:01
and then you can use it." But how long
01:08:03
will that take? And then by that time,
01:08:05
are we on Kimmy K3? And then what is
01:08:08
France supposed to do? I don't know. So
01:08:10
these are complicated questions. But
01:08:12
yeah, we do need an entire reimagining
01:08:16
of how you red team some of these open
01:08:17
source models, obviously.
01:08:19
>> And there's open source models being
01:08:21
made here in the US too, by the way.
01:08:23
>> Exactly. No, no. Meta is doing them.
01:08:25
>> Wherever it comes from,
01:08:26
>> you're going to need AI models to look
01:08:28
for corrupted code and to do the
01:08:30
security eval. And there's going to have
01:08:32
to be continuous monitoring. The good
01:08:33
thing about open source is that when one
01:08:35
person discovers a bug or a flaw or
01:08:38
whatever, they share it with the
01:08:39
community and then it gets patched
01:08:42
globally.
01:08:42
>> The the other problem with open source
01:08:44
though right now is you can't really
01:08:45
fork it and make it your own. Why?
01:08:47
Because there's so much investment by
01:08:49
Moonshot. That's the company that makes
01:08:50
Kim K2 in that example that you'll have
01:08:53
so much drift in one version your fork
01:08:55
will be worthless. So why would you do
01:08:57
it? So again, it goes back to sachs. We
01:09:00
have to be able to say okay Moonshot
01:09:01
will provide Kimmy K2 and every update
01:09:03
thereafter but now we need to stress
01:09:05
test it and we need to redte team it and
01:09:07
we need to be able to say that this
01:09:08
thing is bulletproof and right now there
01:09:11
is no clean way of going to a third
01:09:13
party vendor to do that in a quick
01:09:16
reliable cycle and that's a business
01:09:17
opportunity for somebody
01:09:20
just speaking of the business
01:09:20
opportunity I think there are a number
01:09:22
of American
01:09:23
>> AI companies that are working on open
01:09:25
source models it is a gap that we have
01:09:28
>> so and that is an opportunity I mean,
01:09:30
look, if you're running critical
01:09:32
infrastructure, I don't think you want
01:09:33
to be using a Chinese model, period, on
01:09:37
that critical infrastructure.
01:09:38
>> I think if you can get it validated by
01:09:40
an American company that's trustworthy,
01:09:42
then it's no longer really a Chinese
01:09:43
model. Chinese contributed some really
01:09:45
great ideas and now there's a branch
01:09:47
that you can use. I think it's just hard
01:09:48
to ignore how good this stuff is now.
01:09:51
>> It feels like this was a turning point
01:09:54
this week. Okay, I want to get to one
01:09:56
more important story. But dollar has
01:09:59
dropped as gold and silver and copper
01:10:01
have ripped. Dollar index is down 10% in
01:10:04
the last year. Hit its lowest level in
01:10:07
four years on Tuesday. Trump was asked
01:10:09
if the dollar declined too much. His
01:10:11
quote, "No, I think it's great." Wall
01:10:13
Street thinks Trump wants a weaker
01:10:14
dollar to boost US manufacturing and
01:10:16
exports. Obviously, we have a weaker
01:10:17
dollar. That means the stuff from the US
01:10:19
is cheaper. Foreign stuff becomes more
01:10:21
expensive. Uh and um we have a situation
01:10:25
here, Freeberg, that you've talked
01:10:27
about. Money printing has increased to
01:10:29
$2.5 trillion a year. Trump wants to
01:10:31
print an additional 500 billion more.
01:10:34
That would bring us close to 3 trillion
01:10:35
uh for the military and uh money has
01:10:39
poured into gold and silver which have
01:10:42
way outperformed the S&P. Shout out
01:10:44
Vinnie Lingham. Freeberg, your thoughts
01:10:47
on dollar devaluation and what we're
01:10:49
seeing?
01:10:51
Yeah. So, people talk about the market
01:10:53
going up, but I'll use an analogy. If
01:10:55
you live on an island and there's two
01:10:57
huts on the island and there's a bunch
01:10:58
of shells that people are using for
01:11:01
trade, each house is going to be worth a
01:11:03
certain number of shells. And then if
01:11:05
people went and found a whole bunch of
01:11:06
more shells, the price per house would
01:11:08
go up in number of shells. But there's
01:11:10
just more shells in in the supply. And
01:11:13
effectively, you've inflated everything.
01:11:15
And that's effectively what's gone on
01:11:17
with the US fiscal condition. We've
01:11:19
talked about this many times, but I
01:11:20
think it's always worth a rehash. In a
01:11:23
democracy like we have for the past 250
01:11:25
years, without adequate constitutional
01:11:27
constraints, it has always been the case
01:11:29
that over time spending goes up.
01:11:31
Government spending goes up. And this is
01:11:32
because in a democracy, people ask for
01:11:34
their government to do more every year.
01:11:36
And as they ask for their government to
01:11:38
do more every year, the government
01:11:39
agents who are elected say, "Okay, here
01:11:42
you go." And they spend more.
01:11:43
Eventually, when the borrowing capacity
01:11:44
gets unlocked, which is what happened in
01:11:46
the United States when we went off the
01:11:48
gold standard, you borrow like crazy.
01:11:50
You print money to to fund those
01:11:52
borrowing costs and you use that
01:11:55
fundamentally to drive the next voting
01:11:57
cycle, which is to give people more and
01:11:58
more of what they want. But eventually,
01:12:00
the bill comes to you and in the United
01:12:02
States, the bill is coming to you. Let's
01:12:04
start by looking at the money supply
01:12:05
chart. This is the M2 money supply
01:12:06
chart, showing the rapid rise in dollars
01:12:09
in supply as a function of the central
01:12:11
bank of the United States. the Federal
01:12:13
Reserve making loans to banks ultimately
01:12:16
to fund federal spending. I mean really
01:12:18
an extraordinary number. And if you look
01:12:19
at the M2 money supply chart going back
01:12:22
to 1960,
01:12:24
1955
01:12:26
and you can see postcoid we were hoping
01:12:28
that we would have resolved and sort of
01:12:31
reduced the money supply by some amount
01:12:34
but co really created this accelerating
01:12:36
mechanism and you know we're back on
01:12:38
track in the last couple of years to
01:12:40
increasing the money supply. And so over
01:12:43
time the the US dollar gets devalued as
01:12:45
there are simply more dollars in the
01:12:47
market and US treasuries gets
01:12:49
challenged. So if we take a look here
01:12:51
around the world central banks have
01:12:54
decided that they no longer want to hold
01:12:56
US treasuries. And so this is the value
01:12:58
of gold versus treasuries in central
01:13:01
banks in their inventory. So we are now
01:13:04
seeing that for the first time in
01:13:05
history or
01:13:06
>> no no hold on that's not accurate. It's
01:13:08
not like they're selling. This line just
01:13:10
shows it's stable per se, right? It's
01:13:12
more that the incremental buying is in
01:13:13
real assets.
01:13:14
>> Yeah. But this dollar the dollar value
01:13:16
is also adjusted. So fundamentally I
01:13:18
mean one way to think about this is the
01:13:20
relative value of central bank holdings
01:13:22
around the world. We now see gold
01:13:24
eclipsing US treasuries. So now gold is
01:13:26
a larger share of the holdings. Yeah. So
01:13:28
now gold is the largest share of the
01:13:30
holdings of central banks. If you look
01:13:31
at the next chart which is just over the
01:13:33
past year as JCL pointed out this is the
01:13:35
dollar index. So it's the dollar against
01:13:36
a basket of foreign currencies has
01:13:39
declined you know from a index of about
01:13:42
call it 109 down to 96 today. This chart
01:13:46
actually looks at so what is the US
01:13:48
stock market trading at and instead of
01:13:51
trading it in US dollars what if you
01:13:53
just looked at the US stock market the
01:13:55
total value in ounces of gold. And so if
01:13:58
we had the gold standard still and if we
01:14:02
functionally converted stock market
01:14:04
value from dollars back into gold, you
01:14:07
can see that the stock market in the
01:14:09
United States over the past years, so
01:14:12
this is about seven and a half years
01:14:13
going back to the pre-COVID era, is
01:14:15
actually down, down pretty substantively
01:14:18
from the pre-COVID era. So stock markets
01:14:20
are fundamentally down. Everyone's
01:14:21
cheering, clapping, bouncing up and
01:14:23
down. Stock markets are up. Stock
01:14:25
markets are up. And I'm going to tell
01:14:25
you why this is important in a minute.
01:14:27
Uh, and everyone's, you know, jumping up
01:14:28
and down saying, "Great, the stock
01:14:30
market's up." The stock market's up in
01:14:32
dollar denominated terms. But if you
01:14:33
look at the stock market relative to
01:14:34
gold, it's actually down. And the
01:14:38
sell-off is not just in the stock market
01:14:41
relative to gold, but you can actually
01:14:42
look importantly at the metric that we
01:14:44
all should care the most about, which is
01:14:46
US Treasury yield. So, this is the
01:14:47
30-year. So, the 30-year yield is now at
01:14:50
4.9%.
01:14:52
The average US government's cost to
01:14:53
borrow today is 3.3%.
01:14:57
So if we end up needing to roll all of
01:14:59
the US government debt, assuming we take
01:15:01
on no new debt, which we know is not the
01:15:03
case, $39 trillion of debt outstanding,
01:15:07
the federal government level today, and
01:15:09
it had to get refinanced at this rate,
01:15:11
we would have an incremental annual cost
01:15:13
to service the debt, just the interest
01:15:15
on the existing debt, of roughly $700
01:15:18
billion a year. incremental cost to
01:15:21
service existing debt as interest rates
01:15:23
climb from 3.3 to 5%. And so
01:15:27
fundamentally, this is about 70% of the
01:15:29
current defense budget. It's about 10%
01:15:32
of the overall federal budget. It's a
01:15:34
significant percentage of US GDP, about
01:15:36
3% of US GDP. It's a substantial number
01:15:39
and it creates the spiraling problem
01:15:41
that we're in. Now, I just want to make
01:15:42
one final point. So, there's this
01:15:43
dolorization moment. It's always worth
01:15:45
having a reflection on it, but I just
01:15:47
want to tie it back to Minnesota, Donald
01:15:50
Trump, and socialism.
01:15:53
And I think it's important for us to
01:15:55
just highlight that if you own assets
01:15:58
like we do, the four of us, we own
01:16:01
stocks, we own real estate, we own other
01:16:03
assets. As the dollar devalues and
01:16:06
everything inflates in value, our asset
01:16:09
prices go up and we get wealthier and
01:16:12
wealthier and wealthier. The majority of
01:16:15
Americans do not own assets. They are
01:16:17
net asset negative. As a result, they
01:16:21
live off of income
01:16:23
and they do not benefit from the
01:16:26
ddollarization like asset holders do.
01:16:29
And this is what is ultimately fueling
01:16:31
populism in the United States. And the
01:16:34
populism in the United States is what is
01:16:36
driving socialism. And the response to
01:16:39
those behaviors is what Donald Trump
01:16:40
elected to some degree. and the response
01:16:43
to the Donald Trump actions is what's
01:16:46
driving the civil unrest in Minnesota
01:16:49
and other places. And I fundamentally
01:16:51
believe that much of the unrest, the
01:16:53
civil unrest, and ultimately this divide
01:16:56
in this country is driven by the fact
01:16:59
that ddollarization because of excess
01:17:01
government spending ultimately leads a
01:17:04
majority of people in this country to
01:17:05
feeling oppressed and left behind
01:17:08
because they're seeing a few people in
01:17:10
the country accelerate their net worth
01:17:12
like all of us here. And there's no way
01:17:14
for them to catch up because they don't
01:17:15
actually own assets. So, I'll be I'll
01:17:18
I'll be honest with you guys and make a
01:17:20
confession. I was kind of at I was at
01:17:22
the gym this morning on the treadmill.
01:17:23
>> You were at the gym.
01:17:25
>> Yeah. And while I was there, I was
01:17:26
actually thinking about the wealth tax
01:17:28
stuff that's going on in California. And
01:17:31
I I wonder if it may be an inevitability
01:17:35
in order to keep the United States from
01:17:36
going into civil war. I mean that very
01:17:39
wholeheartedly. Like I just don't know
01:17:41
if there's a way of solving this fiscal
01:17:43
problem without a functional
01:17:45
redistribution of wealth. And the
01:17:46
question is can you do it violently or
01:17:48
nonviolently? And if there's a
01:17:49
nonviolent path, I think that's probably
01:17:51
the preferable path.
01:17:53
>> Did you ever think about do you ever
01:17:54
think about violently picking up some of
01:17:56
those weights?
01:17:58
We can leave that that in or take it
01:17:59
out.
01:18:00
>> The problem with that is look, you know,
01:18:01
where this California wealth tax is
01:18:04
going, right? It's not going to the
01:18:06
quote unquote people. It's going to
01:18:07
these special interests who've been
01:18:09
looting the state for decades.
01:18:12
>> Audit everything before you raise taxes.
01:18:13
It's very simple, folks. Audit
01:18:15
everything.
01:18:16
>> I mean, if the money is going to waste,
01:18:17
foreign abuse, and special interest,
01:18:19
then how do you solve the divide
01:18:21
problem? I mean, I guess the ones that
01:18:23
the special interests are capable of
01:18:25
organizing are able to extract, but
01:18:27
doesn't actually solve the problem. In
01:18:29
fact, everything gets worse because
01:18:32
>> yeah,
01:18:32
>> those government special interests
01:18:34
generally rig the system in their favor
01:18:36
in a way that actually raises the cost
01:18:38
for everybody.
01:18:39
>> So you look at California, everything's
01:18:40
performing worse.
01:18:42
>> Can we pour one out for David Freeberg's
01:18:45
favorite
01:18:47
government program?
01:18:49
>> Free beer.
01:18:50
>> Free beer.
01:18:50
>> Free homeless for 55 homeless people
01:18:54
suffering from the shakes. And they
01:18:56
spent $6 million on it, but it's over. 6
01:18:59
million a year. 6 million a year for 60
01:19:01
bo hobos to get beer because they had
01:19:04
the shakes. The dream is over. The dream
01:19:07
is over. Freeberg. No more free beer.
01:19:09
>> By the way, I think Sax's point is the
01:19:11
right point, which is the resolution to
01:19:13
this isn't a fair
01:19:15
and reasonable redistribution of assets.
01:19:18
It's fundamentally a moment of
01:19:20
extraordinary theft. When there's this
01:19:22
massive movement of capital like this
01:19:24
through a centralized system like the
01:19:26
government, there's no free market
01:19:27
transition of capital and as a result
01:19:29
you end up most seeing a large
01:19:31
percentage of it go into into theft back
01:19:33
into the hands of a few who were really
01:19:35
good at capturing that money as it comes
01:19:36
out of the government's coffers. That's
01:19:38
a good point.
01:19:39
>> You've diagnosed this many times. I
01:19:40
mean, look, Texas and Florida
01:19:43
do a better job for their population
01:19:47
collecting half as much in taxes per
01:19:50
capita as California does. And having no
01:19:52
income tax or capital gains tax,
01:19:55
>> that's because if you commit fraud here,
01:19:57
we put you in a firing squad. That's how
01:19:59
it works. Right to the firing squad.
01:20:01
We're going to go to the range. Let's go
01:20:02
to the Come by Sachs. Come by the ranch.
01:20:04
I have a shooting range.
01:20:06
>> Matt Matt Mayan entered the California
01:20:08
guminatorial race. He's running for
01:20:10
governor.
01:20:11
>> He's running California.
01:20:13
>> Yeah.
01:20:13
>> Uh what does it mean? Who is he? Explain
01:20:15
to the audience why this is important.
01:20:16
>> Well, he's the mayor of San Jose.
01:20:18
>> Chats are on fire.
01:20:19
>> Yeah. He's much more of a moderate and
01:20:21
he's not a union captured candidate.
01:20:24
>> Oh, here's your poly.
01:20:25
>> Yeah, here's poly market. So, Matt Mayan
01:20:27
announced this morning very late entry
01:20:29
to the gubernatorial race. Katie Porter
01:20:32
is kind of the, you know, the output of
01:20:35
the Democratic machine. Tommy Styer is
01:20:36
the billionaire climate change advocate.
01:20:39
But Swallwell, you know, the congressman
01:20:41
from the east,
01:20:42
>> hasn't met a virtue. He doesn't want a
01:20:44
signal.
01:20:46
>> He's literally like Christmas lights
01:20:48
blinking.
01:20:49
>> Tom St.
01:20:50
>> Well, look, if if if Maym, is that the
01:20:52
right way to pronounce his name?
01:20:53
>> Yeah. Matt Mayan.
01:20:54
>> Matt Mayan. California has a jungle
01:20:56
primary, right? So, everyone's running
01:20:58
at the same time. There's no separate
01:20:59
lanes for Democrats or Republicans. If
01:21:01
he ends up top two, let's say it's him
01:21:04
and Swallwell, I think he'll win because
01:21:06
all the Republicans will vote for him.
01:21:08
They'll go for the more centrist
01:21:10
candidate.
01:21:11
>> We had a political strategist over for
01:21:13
dinner. There's a version where the top
01:21:14
two people could be both Republicans
01:21:16
actually right now. Current course and
01:21:17
speed, the top two vote getters are
01:21:19
trending to be both Republican.
01:21:22
>> Yeah.
01:21:22
>> I like the Karen. The Karen's the most
01:21:24
entertaining. She's the one who said,
01:21:25
"Get out of my shot." That's the one.
01:21:27
>> Katy Porter.
01:21:28
>> The get out of my shot one.
01:21:29
>> Yeah. She's tumbling. She's tumbling.
01:21:31
Yeah, she's not doing well. She's coming
01:21:32
down in the she's not likable and she
01:21:34
seems and I think everyone sees right
01:21:36
through her as being, you know,
01:21:37
effectively captured by California
01:21:39
institutions,
01:21:41
>> but Maym and Swallwell much more
01:21:43
independent, but I think it's going to
01:21:45
be a battle between the two of them.
01:21:47
>> Just to go back to your point, Jamas, so
01:21:48
what's interesting is there are two
01:21:50
Republicans running for governor of
01:21:52
California. Chad Biano, who's a sheriff,
01:21:55
and then Steve
01:21:56
>> Steve Hilton,
01:21:56
>> who's a political commentator, used to
01:21:58
be on Fox News. They're both at about
01:22:00
15%. Which actually puts them in the
01:22:03
lead or close to the lead because the
01:22:05
field's so fragmented right now.
01:22:06
Exactly. The problem that they have is
01:22:08
so I guess if the field stayed
01:22:10
permanently fragmented, yes, they could
01:22:11
be in the
01:22:13
>> Could you imagine the melting mines?
01:22:15
>> Yeah. But that's not going to happen as
01:22:17
much as I would like it to because the
01:22:20
Democratic field's going to coalesce and
01:22:21
if it doesn't coales the the party
01:22:23
machine will get together and they'll
01:22:25
tell a bunch of people to drop out and
01:22:27
get real
01:22:28
>> and they'll they'll basically shift it.
01:22:29
So the problem with the Republicans is
01:22:31
and it actually be better if there was
01:22:33
one Republican instead of two because
01:22:34
they need to be at like 30%. Right.
01:22:37
>> Right.
01:22:37
>> But you know they're each going to get
01:22:39
15%. They they have more of a ceiling is
01:22:41
what I'm trying to say. Right.
01:22:42
>> Yeah. That's true. That is true.
01:22:43
>> Yeah. It's only the final two that make
01:22:45
it into the general. So, if the rest of
01:22:47
the Dems literally stay equal and
01:22:49
there's five Dems that are at like 5 to
01:22:51
10% each, you could end up seeing the
01:22:53
two Republicans in the top two spots.
01:22:54
>> I don't think that's going to happen. I
01:22:55
don't think you're going to end up with
01:22:56
five Dems at at 10%. I think what could
01:22:59
happen though actually is it's probably
01:23:01
better that you have two Republicans
01:23:02
rather than one because the disaster
01:23:05
would be let's say you end up with
01:23:08
Swallwell and one of the Republicans in
01:23:11
the runoff, then Swallwell definitely
01:23:13
wins. That's what always happens, right?
01:23:16
>> The best chance for Matt Mayan is that
01:23:18
it's him versus either a Democrat or a
01:23:21
Republican actually.
01:23:23
>> And then if he's up against a Democrat,
01:23:24
a more liberal Democrat like Swallwell,
01:23:27
the Republicans will support Maym and
01:23:28
he'll win. And if he's up against a
01:23:30
Republican, then he'll also win because
01:23:31
all the Democrats will support him. And
01:23:33
remember, this is like a plus 20 or 30
01:23:36
blue state. So,
01:23:37
>> and by the way, he is such a good guy.
01:23:39
He's been such an effective mayor of San
01:23:41
Jose. you know, he cleared up
01:23:42
homelessness. I mean, his polic actual
01:23:45
work running something stands out
01:23:47
amongst the rest of these folks. So, I
01:23:49
think he's got a real shot, Sachs, at
01:23:51
moving up real fast, even though he's
01:23:52
coming.
01:23:52
>> Well, yeah. Look, if he plays his cards
01:23:54
right, he could win. I mean, look, I
01:23:55
think there is a a path here for
01:23:57
California restoration. You get someone
01:23:59
like Matt Mayan as governor, maybe Rick
01:24:02
Caruso takes another shot at running for
01:24:04
mayor in LA.
01:24:05
>> I think everyone recognizes that.
01:24:07
>> Get in the game.
01:24:08
>> I think everyone recognizes that Karen
01:24:10
Bass has been a disaster. She, you know,
01:24:12
aided and abetted the
01:24:13
>> burning down of the Palisades. And then
01:24:16
if you defeat the wealth tax so that the
01:24:18
>> Wait, did Trump take over,
01:24:20
>> you could have a path?
01:24:21
>> Did Trump just take over the rebuild in
01:24:23
Pacific Palisades with an executive
01:24:25
order? Did I see that in my feed?
01:24:26
>> He did. Yeah. Well, you know, there's
01:24:28
been a very small number of permits
01:24:29
granted. I mean, how long has it been
01:24:32
like 18 months now?
01:24:33
>> Sound [ __ ]
01:24:34
>> Wait, so Sax, would you move back if
01:24:36
there was a great restoration in
01:24:37
California?
01:24:41
I feel like somehow it's just not going
01:24:42
to happen. You know, it's like
01:24:44
>> it's over. Not in our lifetime.
01:24:46
>> I feel like the fact that you're telling
01:24:48
me that Ma May is this great candidate
01:24:50
and, you know, the whole tech world's
01:24:51
going to get behind him. I'm sure that's
01:24:52
true. And then somehow I think it's the
01:24:54
reason he's going to lose and I think,
01:24:57
you know, we'll end up with Swallwell,
01:24:59
who I think
01:24:59
>> you're so jaded, Sax. I think that
01:25:01
somehow
01:25:03
>> well Swall is considered to be kind of a
01:25:06
lightweight and a pretty dim bulb.
01:25:11
>> Friend of the pond, Eric Swall, come
01:25:13
back anytime. Who interviewed him with
01:25:14
me?
01:25:15
>> Jim, was that you? Did you interview him
01:25:17
with me?
01:25:18
>> Yeah, I did. Yeah, you did. Yeah. Yeah.
01:25:20
Yeah.
01:25:20
>> Yeah.
01:25:21
>> I mean, just manifestly not up to the
01:25:23
challenges that are going to beset
01:25:26
California in the coming years. You need
01:25:28
someone really stellar to clean things
01:25:30
up after the fiscal insanity of the
01:25:32
Newsome years.
01:25:33
>> Well, look, I'll restate this on the
01:25:35
show very clearly. California has a
01:25:38
trillion dollar fiscal cliff coming up
01:25:40
because of the pension obligations. I
01:25:41
don't know if you guys know this. I'll
01:25:42
make this one point very important to
01:25:44
know. They once tried to change the
01:25:47
benefits, the pension benefits, and they
01:25:49
lost a court case. And so there's
01:25:51
precedent in California state court that
01:25:54
you cannot change pension benefits on
01:25:56
the date that someone was hired. You can
01:25:58
never change their future benefits.
01:26:00
Really important to know. So all of the
01:26:03
benefits that have been given to every
01:26:04
California public employee up to this
01:26:06
day or that they've been promised, they
01:26:08
are promised for life and you're not
01:26:10
allowed to take them out and they cannot
01:26:11
be disposed of except for some form of
01:26:14
bankruptcy. And there is no mechanism by
01:26:17
which the state can declare bankruptcy.
01:26:19
So fundamentally and functionally, there
01:26:21
are two ways that California can be
01:26:22
saved. Number one, you pass an amendment
01:26:25
to the Constitution to fix this pension
01:26:27
liability problem. And number two is the
01:26:29
state has the ability to declare
01:26:31
bankruptcy. Everything else is all about
01:26:32
how long are you keeping the state alive
01:26:34
for.
01:26:34
>> Well, I I I support that that change to
01:26:37
bankruptcy law. I just think it'll be
01:26:38
hard to get. But, you know, we could
01:26:39
try. And look, if you had a governor of
01:26:41
California
01:26:43
come to the federal government and say,
01:26:44
"We would like this," then there's a
01:26:45
much higher chance of it happening. But
01:26:47
actually, I think one question I think
01:26:49
one question
01:26:50
>> for you to ask in evaluating this is
01:26:52
what does Gavin Newsome want? Because
01:26:54
obviously he's running for president in
01:26:55
2028, you know, who would he like to
01:26:58
replace him as governor? You know, and
01:27:00
obviously he doesn't want a Republican
01:27:02
because that'd be a rebuke. You know,
01:27:04
does he want Swall? Does he want someone
01:27:07
who will appear even weaker and stupider
01:27:11
than him? or does he want someone like
01:27:13
Maym who will be appearing
01:27:16
>> to clean up the state? I think he might
01:27:18
want the idiot
01:27:20
>> although it's a it's a tough it's a
01:27:21
tough tough choice for him.
01:27:23
>> She'll just scream. I like
01:27:24
>> Well, here's the thing is he also
01:27:25
doesn't want the state to fall apart
01:27:27
while he's running for president cuz he
01:27:28
might be blamed for that too.
01:27:30
>> So, does he want someone like Compton
01:27:31
like Maym who's criticized him because
01:27:33
Maym has criticized Newsome on things
01:27:36
like homelessness, right?
01:27:38
>> So, Gavin is super thin skin.
01:27:40
>> They can get past that.
01:27:41
>> I don't know. But Gavin is super thin
01:27:42
skinned, so I'm sure he resents that. So
01:27:45
probably the machine gets behind
01:27:46
Swallwell,
01:27:48
even though it probably means a train
01:27:49
wreck for the state. But we'll see.
01:27:51
>> Yeah, that's doesn't seem like a good
01:27:53
idea. Yeah, I think he would want.
01:27:56
>> But this is a big opportunity for tech
01:27:57
to flex his political muscles to see.
01:27:59
>> Yeah, let's see. Let's see.
01:28:00
>> Let's see if they can do it. Let's see
01:28:01
if they can do it.
01:28:02
>> It's not just tech. It's anyone that
01:28:04
doesn't want an establishment governor.
01:28:06
And I think that includes Hollywood. It
01:28:08
includes agriculture. It includes large
01:28:10
swats of the state's economy. So, let's
01:28:12
see what happens.
01:28:12
>> All right, everybody. The AllIn event
01:28:15
series continues. All-in Summit Los
01:28:17
Angeles September 13th to 15th. Tickets
01:28:22
are now taking applications. So, join us
01:28:24
in September. And if you are a venture
01:28:27
capitalist, LP in funds, sovereign
01:28:30
wealth fund, um, endowment, we're going
01:28:33
to be having our first all-in liquidity
01:28:35
event May 31st to June 3rd. Go to
01:28:39
allin.com/events
01:28:40
to apply for a ticket to liquidity as
01:28:43
well. And we'll see you all next time on
01:28:46
the world's greatest podcast, the All-In
01:28:49
podcast. Go ahead and subscribe to our 1
01:28:53
million subscriber channel on YouTube.
01:28:55
That's right. YouTube just broke 1
01:28:58
million. All right. Love you, besties.
01:28:59
Bye-bye.
01:29:01
>> Bye-bye. Love you, Bruce.
01:29:04
>> We'll let your winners ride.
01:29:06
Rainman David
01:29:11
>> and it said
01:29:11
>> we open sourced it to the fans and
01:29:13
they've just gone crazy with it.
01:29:24
>> Besties are
01:29:27
my dog taking your driveways.
01:29:31
>> Oh man, my habitasher will meet the
01:29:33
ugly. We should all just get a room and
01:29:35
just have one big huge orgy because
01:29:37
they're all just useless. It's like this
01:29:38
like sexual tension that they just need
01:29:40
to release somehow.
01:29:45
>> Your feet.
01:29:47
>> We need to get merch.
01:29:49
>> I'm going all in.
01:29:57
I'm going all in.

Podspun Insights

In this episode of the All-In podcast, the original quartet reunites, bringing a lively banter that dances between the absurd and the profound. The crew dives into their recent experiences at Davos, where the atmosphere was thick with anticipation over Donald Trump's speech, leading to a mix of chaos and intrigue. They share humorous anecdotes about their accommodations and the peculiarities of the event, all while dissecting the political landscape that seems to be shifting under their feet.

As the conversation unfolds, they tackle the tragic events in Minneapolis involving federal agents and the community's response. The discussion is both intense and thought-provoking, as they navigate the complexities of immigration policy and public safety, all while keeping the tone light and engaging. The hosts don’t shy away from the heavy topics, but they manage to weave in humor and camaraderie, making for a captivating listen.

Amidst the serious discussions, the episode takes a turn towards the innovative with the introduction of Claudebot, an open-source personal assistant that has taken the tech world by storm. The hosts share their experiences with this new technology, highlighting its potential to revolutionize how we interact with AI. The excitement is palpable as they envision a future where personal AI assistants become commonplace, leading to a fascinating exploration of the implications for society and the economy.

With a blend of humor, insight, and a touch of chaos, this episode captures the essence of the All-In podcast: a space where serious discussions are balanced with lighthearted banter, making complex topics accessible and engaging for all listeners.

Badges

This episode stands out for the following:

  • 90
    Most intense
  • 85
    Most heartbreaking
  • 85
    Most chaotic
  • 85
    Best concept / idea

Episode Highlights

  • Davos Takeover
    This year's Davos was marked by a business-centric atmosphere and a notable Trump presence.
    “It was a business takeover and a Trump takeover.”
    @ 02m 14s
    January 31, 2026
  • Tragic Deaths in Minneapolis
    Two tragic deaths involving federal agents raise questions about immigration enforcement in Minnesota.
    “The deaths of Renee Good and Alex Prey are regrettable and tragic.”
    @ 12m 27s
    January 31, 2026
  • The Will of the Majority
    Polls indicate a majority of Americans support deporting illegal immigrants. 'There's no arguing with these different numbers.'
    @ 18m 42s
    January 31, 2026
  • Concerns Over Law Enforcement Tactics
    Debate on the actions of federal law enforcement and the need for accountability. 'Federal law enforcement agents should not wear masks.'
    @ 26m 08s
    January 31, 2026
  • Compassion for Immigrants
    We should be compassionate to those who've been here for years and paid taxes.
    “They should be given a path.”
    @ 36m 27s
    January 31, 2026
  • The Role of Business Owners
    Finding and penalizing business owners hiring illegal aliens could reduce immigration issues.
    “Wouldn't that be a better policy?”
    @ 44m 14s
    January 31, 2026
  • Claudebot Goes Viral
    An open-source personal assistant project created by Peter Steinberger is gaining attention.
    “I have been claed. I am all in on this.”
    @ 45m 58s
    January 31, 2026
  • The Future of Work
    Imagine having an army of super workers at your disposal, revolutionizing entrepreneurship.
    “Now think about what Elon’s doing... everyone becomes an entrepreneur.”
    @ 55m 49s
    January 31, 2026
  • Open Source AI Revolution
    The emergence of open source AI models democratizes access and control over technology.
    “This is why it’s so important because it was incredibly profound.”
    @ 59m 47s
    January 31, 2026
  • Trump's Take on Dollar Decline
    Trump believes the weaker dollar is beneficial for US manufacturing and exports.
    “"No, I think it's great."”
    @ 01h 10m 11s
    January 31, 2026
  • California's Free Beer Program Ends
    A controversial government program providing beer to homeless individuals has been terminated.
    “"The dream is over."”
    @ 01h 19m 07s
    January 31, 2026
  • California's Fiscal Crisis
    California faces a looming trillion dollar fiscal cliff due to pension obligations.
    “California has a trillion dollar fiscal cliff coming up.”
    @ 01h 25m 38s
    January 31, 2026

Episode Quotes

Key Moments

  • Davos Impressions01:25
  • Tragic Events10:19
  • Law Enforcement Accountability26:08
  • Emotional Activation29:57
  • Open Source Impact59:47
  • Dollar Devaluation1:10:11
  • California Politics1:20:10
  • Pension Crisis1:25:38

Words per Minute Over Time

Vibes Breakdown