00:00:00
all right everybody welcome back to the
00:00:02
Allin podcast the Channel's been active
00:00:06
we're in the Afterglow we're in the
00:00:08
Allin Summit
00:00:09
after it's so glowing that freeberg
00:00:13
couldn't make it he has been writing a
00:00:15
high Nick told me that in the last week
00:00:18
just we uh we've only put out a half the
00:00:20
clips and they've already gotten 20
00:00:22
million views oh my Lord I you know I I
00:00:25
I so we'll be we'll be around 50 million
00:00:27
I think when all the clips are released
00:00:29
and you let it bake for a couple of
00:00:31
months that is an astoundingly large
00:00:34
amount of breach yeah and that's just
00:00:36
YouTube we're not doing it on the
00:00:38
podcast feed right now YouTube and X
00:00:39
well hopefully we get it on the podcast
00:00:41
feed we get another 50 million uh but uh
00:00:44
free Berg's in his Afterglow couldn't
00:00:46
make it but he's very busy right now
00:00:48
look how happy he is the summit went
00:00:51
well is that marijuana I think he's
00:00:54
making potatoes I think that's his farm
00:00:56
but I mean the smile is incredible it's
00:01:00
marijuana it's free BG's version of
00:01:02
founder mode he's in that's freeberg
00:01:05
founder mode he's hitting the bong his
00:01:07
founder mode
00:01:07
[Music]
00:01:10
gives let your winners
00:01:12
[Music]
00:01:17
ride and instead we open sources to the
00:01:20
fans and they've just gone crazy with it
00:01:23
queen
00:01:25
of he is uh he's in the Afterglow um and
00:01:30
he won't be with us this week um but uh
00:01:33
he organized such a great conference
00:01:35
don't you think J he did great uh I mean
00:01:37
he really took charge of that and just
00:01:41
did an amazing J I would like to give
00:01:42
him his flowers absolutely it is like at
00:01:45
least a trillion times better than the
00:01:47
first and at least 50% better than the
00:01:49
second I mean that's how it should go
00:01:52
you know when you create something in
00:01:53
the world chth what you want to do is
00:01:55
you want to H you want to hand it off to
00:01:57
Professional Management to then scale it
00:01:59
right not everybody can do the creative
00:02:01
act of actually forming something you
00:02:03
need to have these operators to go and
00:02:07
then execute your vision and uh I just
00:02:08
want to give free to break his flowers
00:02:10
for executing incredibly well we all
00:02:12
play a role jamat saaks launched a
00:02:15
tequila company I want to say thanks to
00:02:18
uh Friedberg he did all of these great
00:02:20
speakers big thank you to our CEO John
00:02:23
who put together all the operations Nick
00:02:26
did incredible Nick did incredible
00:02:28
incredible job with those opening
00:02:30
Graphics they went viral uh Zach helped
00:02:33
with the graphics you had young
00:02:35
Spielberg chipping in you had Laura did
00:02:37
an amazing job with Stage management and
00:02:39
of course you know I focused on the
00:02:40
moderation I got a lot of great things
00:02:42
so everybody plays a role you got Sachs
00:02:44
with the tequila Friedberg
00:02:47
Laura Zack Spielberg Nick John everybody
00:02:51
brought something to the table are you
00:02:53
congratulations to
00:02:56
everybody you scale through people
00:02:58
that's it scale through people that's it
00:03:01
did anybody nobody got the
00:03:03
joke we chth everybody contributed you
00:03:07
understand Sachs new tequila company
00:03:10
John operations your freeberg content me
00:03:13
with the being the world's greatest
00:03:14
moderator up there what ch's
00:03:16
contribution oh yeah chath showed up
00:03:20
chath chth looked great I showed up
00:03:23
that's just he showed up and looked
00:03:25
great I brought my two votes and I
00:03:27
brought my vision absolutely I would
00:03:29
also say fan favorite you what you
00:03:32
really did that was amazing was you took
00:03:34
a lot of selfies I was very proud of
00:03:36
both of you with the fans service and
00:03:39
fans were very pleased that you guys
00:03:40
took so many selfies you know we got a
00:03:42
lot of feedback too coming in
00:03:46
so it was uh pretty pretty great
00:03:50
feedback do you think that you did
00:03:52
better as moderator because you finally
00:03:55
let go of just the conference
00:03:58
organization what
00:04:01
yeah I think that you were able to focus
00:04:03
on your unique value ad instead of
00:04:05
immersing yourself in a bunch of details
00:04:07
that could be handled by the team I
00:04:10
agree um it was absolutely process to
00:04:13
get you to let go well you know you have
00:04:17
it's it's a fair point I I did people
00:04:19
did say my moderation was dialed in and
00:04:21
I appreciate that positive feedback from
00:04:22
everybody and um yeah there is something
00:04:25
to having people you trust with the
00:04:27
content thought your moderation was
00:04:29
excellent this time it was better than
00:04:31
before because I think that you're
00:04:33
actually exceptional as a moderator and
00:04:35
I think you're mostly average as a
00:04:37
conference
00:04:38
[Laughter]
00:04:41
producer I do think as a moderator
00:04:43
you're excellent I mean like some of the
00:04:45
most memorable moments were you
00:04:49
basically drawing out contrasting
00:04:52
opinions and the way that the people
00:04:55
engaged with them was so healthy and
00:04:57
good that was the I think the curring
00:05:00
theme so I gave you an enormous amount
00:05:02
of credit I think you did an exceptional
00:05:03
job but I also think it's because you
00:05:04
were able to focus on what you great at
00:05:06
I I do agree with that I was talking to
00:05:08
Jade about and she said and Nick also
00:05:10
pointed out you were really dialed in J
00:05:12
what what's up and I said I'm not
00:05:14
worrying about the party and the vendors
00:05:16
and the front desk and the sponsors and
00:05:18
it is actually you're able to to focus
00:05:20
did you have some favorite moments
00:05:22
yourself there saxs any favorite moments
00:05:24
for you or panels or things maybe that
00:05:25
exceeded expectations for you well I
00:05:28
thought the Mir shimer Jeffree Sachs
00:05:30
panel was great I thought it would be
00:05:33
which is why I helped organize it but I
00:05:36
was just glad that the audience so many
00:05:37
people in the audience reacted and said
00:05:39
that was the surprise head of the
00:05:40
conference I would say that was my
00:05:41
favorite of the event one of the best
00:05:43
panels I've ever been part of it's the
00:05:45
most viewed it's like slightly above
00:05:46
elon's one really oh just behind El
00:05:49
elon's slightly ahead but yeah it's
00:05:51
still like growing it's like finding an
00:05:53
audience well I think that I think that
00:05:55
if you if you look at the one from last
00:05:57
year greme Allison where he got a
00:05:58
standing ovation the thing is there are
00:06:00
these Village Elders where they are at a
00:06:04
point in their life where they're
00:06:06
willing to just be a truth teller but
00:06:09
often times they're
00:06:10
deplatformed and we have the ability to
00:06:13
actually bring some of the smartest of
00:06:15
them on and give them a voice and it's
00:06:18
incredible how much they resonate
00:06:20
because what they say is so logical and
00:06:23
sensible that's a that's a really
00:06:26
important thing that we have now at our
00:06:27
disposal and I think that people really
00:06:30
appreciate it you know so we're like a I
00:06:32
think we're doing a really important job
00:06:34
in doing that and now the question is
00:06:37
what village Elders do we get next year
00:06:38
to keep you know being truth tellers
00:06:41
well give us your thoughts you know
00:06:42
there's an all-in um Twitter handle and
00:06:45
he's chamath David saxs and I'm at Jason
00:06:47
and freeberg at freeberg just tell us
00:06:49
who you think would be great but saaks I
00:06:51
know you're super excited and want to
00:06:52
give Biden his flowers the FED just cut
00:06:54
rates 50 bips and the stock market is uh
00:06:58
tearing it up right now on Wednesday fed
00:07:02
cut interest rates by a half a
00:07:04
percentage Point taking them down off of
00:07:07
a 23-year high we've been talking about
00:07:10
this God for two years here on this
00:07:11
podcast First Rate cut since March of
00:07:13
2020 which is about when we started this
00:07:17
podcast jpow basically said the FED
00:07:19
thinks inflation is coming down to
00:07:21
around 2% nicely and they don't want the
00:07:25
job market to soften any further than it
00:07:27
already has he also mentioned mentioned
00:07:30
immigration has helped soften the market
00:07:33
uh the labor market as well obviously
00:07:35
with all those new people looking for
00:07:36
jobs so in the last two months July and
00:07:38
August CPI has been at a two- handle we
00:07:40
talked about that 2.9% in July 2.5% in
00:07:44
August here's the CPI over the last
00:07:47
decade obviously massive boom uh in
00:07:50
interest that you see there from 2021 to
00:07:55
2023 many obviously think we're going to
00:07:58
have more R Cuts probably
00:08:00
every meeting for a little
00:08:02
bit and um Dow's already at an all-time
00:08:06
high surge 300 points on the
00:08:09
news here's um here's uh some
00:08:12
interesting data about the 50 basis
00:08:14
point kickoff Cuts so uh this is where
00:08:17
it gets interesting chth fed only
00:08:19
started publicizing their interest rate
00:08:21
changes in 1994 since '94 fed has
00:08:24
initiated a cutting cycle six times
00:08:26
here's the chart take a good look at
00:08:28
that 95 98 2019 they started with 25
00:08:31
bips
00:08:33
01 and 07 after the great financial
00:08:37
crisis they started with a 50 bip cut so
00:08:41
obviously there was an emergency 50 bip
00:08:43
cut in March of 2020 when Co hit 0107
00:08:47
2020 very severe
00:08:50
situations and the what happened in the
00:08:52
markets is what I want to discuss uh
00:08:54
with both of you today in 20 2001 Market
00:08:57
fell 31% in the two years after that
00:09:01
rate cut in 2007 Market fell 26% after 2
00:09:04
years so and 2020 despite all the fears
00:09:08
Market ripped 44% over two years what's
00:09:11
the more likely scenario chth is this
00:09:14
similar to the Doom great financial
00:09:16
crisis or similar to
00:09:18
2020 well I think 2020 you have to put
00:09:21
it at Big asteris because the question
00:09:23
is what would have happened had there
00:09:25
not been covid and had there not been an
00:09:28
entire global shutdown so if you go back
00:09:30
to that chart you could probably just
00:09:33
extrapolate and cut out that part that's
00:09:37
flat because the part that's flat from
00:09:40
2020 to 2022 was largely artificially
00:09:43
created because on top of that we
00:09:45
injected so much money into the economy
00:09:47
the reality is we probably would have
00:09:49
raised at some rate of change that you
00:09:53
could have predicted from 2016 so what
00:09:55
do you what do you take away from that I
00:09:57
think that you have to like realize we
00:10:00
are at a point in the economy where you
00:10:04
cut rates because there's
00:10:06
tension and there's tension between
00:10:09
employment and unemployment there's
00:10:11
tension between earnings growth and
00:10:15
contraction and so it's a stimulatory
00:10:18
move so if you look through that
00:10:21
stimulatory move why is the Fed doing
00:10:24
this and why will they cut probably all
00:10:26
the way down to two or 3% by the end of
00:10:28
26
00:10:30
it's because we now need to stimulate
00:10:32
the economy again so the reason why
00:10:35
markets tend to fall once the rate cut
00:10:37
cycle starts is because the next couple
00:10:41
of quarters sort of demonstrate what I
00:10:44
think the FED is expecting which is that
00:10:47
there's pressure in the economy we have
00:10:50
not seen that flow through in earnings
00:10:53
or in how companies describe markets on
00:10:56
the field by and large except for a few
00:10:59
so I think this part of the cycle now
00:11:01
will be about all of these companies
00:11:03
telling us whether there's nothing to
00:11:05
see here or whether there is actual real
00:11:07
pressure and if there is real pressure
00:11:09
it'll probably look like the several
00:11:11
times before where you're just going to
00:11:12
have to contract the value of financial
00:11:15
assets because they're just not worth as
00:11:17
much when they're earning less okay saxs
00:11:20
any thoughts here just balls and Strikes
00:11:23
I think a lot of people are commenting
00:11:25
on the fact that the only other two
00:11:28
times where we've had a 50 basis point
00:11:30
rate cut in modern history it has been
00:11:33
just before a recession so I think this
00:11:36
happened in 2001 2007 right before the
00:11:40
recession and the FED had to do a
00:11:42
dramatic rate cut because they could see
00:11:44
in the data that things were weakening
00:11:45
so a lot of people are asking the
00:11:46
question well is that what's going on
00:11:48
here now Pal's comments though are
00:11:52
indicating that the econom is in good
00:11:54
shape he said the economy is in very
00:11:55
good shape that um basically indicating
00:11:58
that they had tamed inflation
00:12:00
and that they would look to cut another
00:12:02
50 basis points this year so Pal's
00:12:06
rhetoric
00:12:08
is uh in a way at odds with the
00:12:11
magnitude of this cut the you know so
00:12:13
why didn't they just cut 25 basis points
00:12:15
I think people are trying to figure that
00:12:17
out reading the tea leaves into 50
00:12:21
because they could just do 25 a month
00:12:23
sure mon the econom is hot yeah if the
00:12:26
economy is hot why would you tiptoe into
00:12:28
rate cuts
00:12:29
uh and just do 25 now that's the key
00:12:32
thing if you look at the The Dot Plot
00:12:34
and if you look at where the Smart
00:12:35
Financial actors are betting where rates
00:12:37
end so it's hard to sort of like look at
00:12:40
any point in time 50 now 25 later what
00:12:44
does it all mean it's very hard to know
00:12:46
but what is much clearer is where do we
00:12:48
think terminal rates will be in even in
00:12:50
the next 18 months and it is
00:12:52
dramatically lower for where they are
00:12:54
now and I think that support sacks your
00:12:57
that argument that you just made which
00:12:58
is if you're going to basically cut this
00:13:02
aggressively over the next year to year
00:13:04
and a half by the estimates of very
00:13:06
smart Financial actors whose job it is
00:13:08
to spend every day observing the FED
00:13:11
then they must see something because
00:13:13
otherwise as you said you could take a
00:13:15
much more gradual approach and so I
00:13:18
think that the Smart Financial actors
00:13:21
are guessing recession or guessing
00:13:24
contraction I think what they're also
00:13:26
guessing is similar to nonfarm payrolls
00:13:30
we're going to go through a couple of
00:13:31
difficult GDP revisions probably
00:13:34
downward and I think that will have an
00:13:36
impact to
00:13:38
people's sense of how the economy is
00:13:40
doing even more than what their sense is
00:13:42
today which is already teetering on it's
00:13:44
at best
00:13:46
okay and I think all of that has to play
00:13:48
itself out so it's going to be a very
00:13:50
complicated and dynamic fall in that
00:13:52
respect yeah and and I think so much of
00:13:55
this has to do with Unemployment uh we
00:13:58
had that period where so many jobs were
00:14:00
available remember we talked about it
00:14:02
here 11 12 million jobs available at the
00:14:04
peak we can debate the numbers of course
00:14:06
but we all saw it where you just
00:14:08
couldn't hire a talent in America there
00:14:10
was so few people available to to take
00:14:13
positions and man has that changed and
00:14:15
you get to see it on the ground in early
00:14:17
stage startups where this whole
00:14:20
narrative I don't know if you saw it in
00:14:22
your board meetings but hey we can't
00:14:23
find a person hey we're looking hey that
00:14:25
search is still going we're still
00:14:27
looking for a director of sales we're
00:14:28
still looking for sales people we're
00:14:29
still looking for developers we're still
00:14:30
looking for operations people now it's
00:14:33
the opposite it's like I I just I'm
00:14:36
hiring producers here in Austin because
00:14:38
I'm building in my inperson studio we
00:14:40
had like I don't know a dozen viable
00:14:43
candidates for this position and I had a
00:14:45
hard time picking between you know the
00:14:48
top three now that's distinctly
00:14:50
different than my experience for the
00:14:52
last five to 10 years where you were
00:14:54
like how do we how do we fill this role
00:14:56
so I think that employment has been
00:14:58
broken
00:14:59
and that's the thing that has me
00:15:00
concerned because with all these people
00:15:02
who came in through the southern border
00:15:04
and then you have people Outsourcing to
00:15:07
other countries I wonder if Americans
00:15:09
are going to lose so many of these midp
00:15:11
paying jobs and this will dovetail into
00:15:13
our next story about Amazon making Cuts
00:15:16
I'm very worried about the the hollowing
00:15:17
out of the upper middle class that Elite
00:15:20
group of
00:15:21
$150,000 jobs that then employ nannies
00:15:24
and spend money in the economy I wonder
00:15:26
I don't know if you're seeing that in
00:15:27
your company sacks
00:15:30
I'm not worried about the hall out of
00:15:31
that that
00:15:33
class you have disain for them but I
00:15:36
mean just in terms of the labor market
00:15:38
what do you see you know in companies
00:15:40
right now you know hiring the talent
00:15:44
pool Etc well I mean in tech things are
00:15:47
are pretty good I mean they they're not
00:15:48
as absurdly frothy as they were during
00:15:51
the bubble of 2020 and 2021 but things
00:15:54
are good you have this huge AI Tailwind
00:15:56
now and there's just a ton of investment
00:15:58
going into AI there's a little bit of A
00:16:00
Tale of Two Cities going on if you're in
00:16:02
AI things are really bubbly and if
00:16:04
you're outside AI they're they've return
00:16:07
to much more normal levels in terms of
00:16:10
valuation and Company operations all
00:16:14
that kind of stuff just to go back to
00:16:16
the state of the economy for a second
00:16:18
the the reason why a lot of people were
00:16:20
predicting a recession including me for
00:16:23
a while is that the yield curve
00:16:25
inverting has been an almost perfect gug
00:16:29
of whether a recession is coming it's
00:16:31
when basically the FED raises short-term
00:16:34
interest rates above long-term interest
00:16:36
rates normally long rates are the ones
00:16:39
that should be higher because investors
00:16:41
demand a higher rate of return to tie up
00:16:43
their money for longer so something's
00:16:44
really off and kind of broken when short
00:16:47
rates go above long rates the yield
00:16:49
curve inverts and it's always been the
00:16:51
Prelude to a recession but the recession
00:16:53
doesn't come when the yield curve
00:16:55
inverts it usually comes when the yield
00:16:57
curve de inverts
00:16:59
and the reason for that is because the
00:17:00
FED now sees weakness and dramatically
00:17:04
cuts the short rates so in other words
00:17:06
it jacks up the short rates to control
00:17:08
inflation that works it trickles through
00:17:10
the economy the economy cools down and
00:17:13
then the FED says oh [ __ ] maybe we've
00:17:14
over corrected they slam on the brakes
00:17:16
and then they cut rates to basically
00:17:18
make up for the effect in the economy so
00:17:20
the yield curve has finally de inverted
00:17:23
and the question is just do we now get
00:17:25
that recession or did the FED manage
00:17:27
this to a soft Landing I don't think we
00:17:29
know I'm not I'm not like calling
00:17:31
recession but this is the the thing that
00:17:32
people are concerned about yeah well
00:17:35
saak we were talking about AI in the
00:17:36
group chat right yeah I think it's now
00:17:38
becoming really clear that call centers
00:17:40
are going to be the first really big
00:17:42
disruption caused by AI yeah I mean all
00:17:46
the level one customer support is going
00:17:47
to get replaced by AI I mean llms plus
00:17:51
voice cuz you know open AI just released
00:17:55
their audio
00:17:56
API you saw that at the All In Summit we
00:17:59
released A Mir shimer AI yeah where we
00:18:03
trained it on all of his work and you
00:18:05
can go to Mir shimmer. and ask a
00:18:07
questions and it will tell you the
00:18:08
answers in his voice because we cloned
00:18:10
his voice using resemble AI anyway so AI
00:18:15
can do voice now and it can be trained
00:18:18
extremely well on large data sets to
00:18:21
give you answers to questions which is
00:18:23
pretty much what customer support is so
00:18:25
I think it's now becoming clear that I I
00:18:28
think within the next two to three years
00:18:29
you're going to see a massive disruption
00:18:31
in that I agree with that massively and
00:18:33
I think there's another underreported
00:18:35
story which is people don't like to call
00:18:39
and talk to a customer service agent
00:18:41
like an actual human if they can avoid
00:18:43
it they would much rather go on YouTube
00:18:45
and say how do I fix this or you know
00:18:47
ask chat GPT how do I fix this it's like
00:18:50
I don't want to waste another person's
00:18:51
time just give me the answer as quick as
00:18:53
possible and AI will give you the answer
00:18:55
quicker YouTube will give you the answer
00:18:57
quicker I've had so many times where I
00:18:59
have people who work for me who are like
00:19:01
I don't know how to do that and I
00:19:03
literally would walk up to their
00:19:04
computer and load YouTube and type in
00:19:07
how do I blank and there's a video there
00:19:10
watch it on two speed you can do it
00:19:12
that's what's you know gonna also kill
00:19:14
this like I I don't want to talk to a
00:19:15
human just change my flight just you
00:19:19
know question yeah I mean you talk about
00:19:21
disruption call centers are a very big
00:19:24
part of the economy in certain
00:19:25
geographies Denver Sal Lake I mean parts
00:19:28
of
00:19:30
yeah exactly it's a really big deal if
00:19:32
like half the cost gets ripped out of
00:19:34
those call centers where would you move
00:19:36
those people if you if you had your
00:19:38
choice could they move to sales well I
00:19:40
think sales will be the one that's
00:19:41
disrupted after customer support but um
00:19:44
but I don't know I think it's going to
00:19:46
be very disruptive one of the reasons I
00:19:48
think this is you know in the early days
00:19:49
of llms people were saying that legal
00:19:52
services would be disrupted and you saw
00:19:55
some very highly valued startups Rock up
00:19:59
based on that I think the problem with
00:20:01
that is the error rate so when you think
00:20:04
about AI applications you have to think
00:20:08
about what is the tolerable error rate
00:20:11
that the industry will allow because we
00:20:13
know that AI get things wrong they can
00:20:15
hallucinate and you're never going to be
00:20:17
able to make it perfect I mean you can
00:20:18
improve the quality but it's still going
00:20:20
to have some errors and when you're
00:20:22
dealing with like legal services for
00:20:23
example you just can't have mistakes
00:20:26
it's just not tolerated however customer
00:20:28
support is different customer support is
00:20:30
already organized into levels level one
00:20:33
level two level three based on
00:20:34
difficulty and there's already in a
00:20:37
sense a mechanism for failover if like
00:20:39
the level one customer support person
00:20:41
can't answer the question they kick it
00:20:43
up to level two so there's a place for
00:20:47
llms to start in customer support which
00:20:50
is replacing all the level one and then
00:20:52
working their way up the chain to level
00:20:54
two as they get better and better and so
00:20:58
what I'm saying is that the level of
00:21:00
accuracy now especially with the new PhD
00:21:02
level reasoning models is good enough
00:21:05
yeah we don't need to wait for like some
00:21:07
perfect llm model and I think this is
00:21:10
why this is going to be a big big
00:21:11
disruption let me people potentially are
00:21:13
going to have their their jobs disrupted
00:21:15
or at least transformed well it could be
00:21:17
the end of the entire career as well
00:21:19
jamath if you were to look at this 4x4
00:21:21
sort of quadron chart that sax is
00:21:24
describing which is the cost of an era
00:21:27
you know and
00:21:29
um the actual complexity of the job
00:21:32
perhaps or the cost of the job how do
00:21:35
how do you look at this I know you're
00:21:38
working on software that kind of does
00:21:40
this with your startup as
00:21:43
well I mean I'll
00:21:46
preview one use case from 8090 which is
00:21:51
pretty
00:21:53
stunning you know we work with an a very
00:21:56
large regulated highly regulated
00:22:01
company public
00:22:04
company and they have a very complicated
00:22:08
set of people and
00:22:10
processes because of the the field in
00:22:12
which they're
00:22:14
in and David your point is exactly right
00:22:17
it took
00:22:19
us a fairly long
00:22:22
time but we're at a point now where
00:22:24
we've been running AI powered
00:22:27
software versus the old Legacy
00:22:31
deterministic solution and we've been
00:22:33
running it at 100% accuracy now for
00:22:36
about 10
00:22:37
days so this is still very new and it's
00:22:41
an incredible thing because to your
00:22:43
point our first version was like a in
00:22:46
the mid 80s then we were in the mid90s
00:22:49
then we were you know 97 98% but there
00:22:51
was still errors and it just took a lot
00:22:54
of engineering to figure out how to get
00:22:56
to 100 but now it's at 100 and it's been
00:22:58
consistent at 100 and so we're all kind
00:23:00
of like scratching our head because now
00:23:02
the next step is well what do we do to
00:23:04
your point what what do we do do we so
00:23:07
we're we're figuring that out right
00:23:11
now but the art of the possible is that
00:23:15
I think well-crafted AI software is as
00:23:18
good as deterministic software in the
00:23:20
sense that the error rates will be
00:23:23
equivalent in production and at the
00:23:26
level of very highly regulated public
00:23:30
company and I think that's the gold
00:23:31
standard because in those sectors those
00:23:34
companies have zero tolerance it's not a
00:23:37
toy it's not even you know level one
00:23:39
customer support it's system of record
00:23:43
type work yeah but it shows what's
00:23:46
possible and to your point saaks we're
00:23:48
doing that today even though they're the
00:23:50
best models imagine how good those under
00:23:52
the underlying models will get in a year
00:23:54
from now yeah right and we'll be able to
00:23:57
take on more and more work it's it's
00:23:58
very stunning actually it's really have
00:24:00
you guys worked with the 01 preview yet
00:24:03
I I just literally have been using this
00:24:05
new reasoning engine that open AI
00:24:08
released and it is extraordinary and
00:24:10
it's kind of thinking about the next
00:24:12
three or four prompts you would do and I
00:24:14
literally just got this while we're on
00:24:15
the show I've hit the I've hit the limit
00:24:18
for my paid account because this thing
00:24:20
is so intense on compute I guess well
00:24:23
the thing with o1 is that I think it's
00:24:25
starting to add reasoning but the way
00:24:27
that you do reasoning
00:24:29
is sort of this idea that you have this
00:24:30
Chain of Thought and I think that that's
00:24:33
a very powerful but early concept and as
00:24:37
we refine those ways in which these
00:24:41
models get the better answers the
00:24:43
wonderful thing is that open AI will
00:24:46
preview
00:24:47
o1 and then they'll have the actual 01
00:24:50
production build probably in the next
00:24:52
couple of months which will be probably
00:24:53
pretty spectacular but then you'll see
00:24:55
something from Claude you'll see
00:24:57
something from llama and the real art I
00:25:01
think and this is where I do think it's
00:25:02
a little bit of alchemy still which I
00:25:04
think is good because it it keeps humans
00:25:06
involved all of us involved
00:25:08
yes is how do you Stitch all of those
00:25:11
things together to get to a 0% error
00:25:14
rate what what sack said you know how do
00:25:15
you minimize the blast radius and how do
00:25:17
you make sure these things are super
00:25:18
high quality right well and people don't
00:25:22
it's still a very hard technical problem
00:25:24
go ahead SX and then I I'll show you
00:25:25
what so yeah one of the reasons why I'm
00:25:27
bullish on this customer support use
00:25:28
case is because there's a very large
00:25:30
data set to train on you've got all of
00:25:33
the product documentation that companies
00:25:34
very created you've got all the previous
00:25:37
email support you know and calls yeah
00:25:41
the calls have been recorded so you can
00:25:43
out train the AI on that so there's a
00:25:45
very large body of data to train the AI
00:25:48
model on and it's not necessarily the
00:25:50
most proprietary it's not like dealing
00:25:53
with people's medical records or or even
00:25:56
confidential legal documents something
00:25:57
like that so the data is readily
00:25:59
available and then the foundation models
00:26:01
are getting really good I think there's
00:26:02
a big question here about value capture
00:26:05
which is there's a number of startups
00:26:07
now that are becoming very highly valued
00:26:09
that are chasing this disruption this
00:26:12
sort of customer support agent
00:26:14
disruption and they're getting into very
00:26:17
high valuations even unicorn valuations
00:26:20
already and the question is well wait if
00:26:23
if the foundation models are advancing
00:26:24
at such a rate exactly like a year from
00:26:27
now why couldn't a like a developer just
00:26:29
a startup of a few guys take next year's
00:26:32
model train it and then commoditize
00:26:36
the you're you're making such a good
00:26:38
point this so when we were trying to
00:26:40
figure out like what applications we
00:26:42
would build and like which sectors of
00:26:45
the economy we would go after I was like
00:26:47
guys we got to go after the hardest most
00:26:50
regulated places because those are the
00:26:53
things and places and people that have
00:26:55
absolutely zero tolerance for error and
00:26:57
where
00:26:58
you're going to need to do some amount
00:27:00
of customization and and specialization
00:27:03
to actually solve these problems and S
00:27:06
to your point like when you see and I
00:27:07
said you cannot we cannot touch customer
00:27:09
service we cannot touch it because it's
00:27:11
going to
00:27:12
get commoditized and run over by these
00:27:15
foundational models within a year right
00:27:19
you you'll be able to deploy these it's
00:27:21
just too easy you'll be able to do it on
00:27:22
a local computer I mean you'll just
00:27:24
download the entire database of every
00:27:26
call on a MacBook with by the way just
00:27:29
to build on that that the other thing
00:27:31
that's now possible and you saw this
00:27:32
with Clara because Clara put out this
00:27:34
like cryptic tweet press release where I
00:27:37
think maybe it was in their earnings
00:27:39
Nick maybe you can find this where
00:27:40
they're like we've deprecated Salesforce
00:27:42
and worked it that like range how how
00:27:45
can a company that big deprecate those
00:27:48
two systems of record how is that even
00:27:51
it's how is it means they're writing
00:27:53
their own right well I'll tell I'll tell
00:27:54
you how it's possible and so this is
00:27:56
like this next crazy thing that's been
00:27:58
happening we've been doing a version of
00:28:00
this to go after some other sources of
00:28:02
software we haven't had the balls to be
00:28:05
honest to go after s force or or work
00:28:08
day but here's how they do it they write
00:28:10
these agents and these agents can spawn
00:28:13
other agents right so it's very classic
00:28:15
kind of machine that builds a machine
00:28:17
and you start to observe the inputs and
00:28:19
outputs of a system right I'm I'm hyper
00:28:22
simplifying but I'm just it'll make the
00:28:23
point and over time what the agents
00:28:26
start to do is by observing the inputs
00:28:28
and the outputs they start to guess on
00:28:30
what the intervening code is and the
00:28:32
code pths must be in the middle to
00:28:33
generate the outputs based on these
00:28:35
inputs and so over time what happens is
00:28:38
you develop a digital twin and then you
00:28:41
run that against that counterfactual
00:28:45
against workday or Salesforce and then
00:28:47
at some point you're like it's the same
00:28:50
and you you just turn it off and you're
00:28:52
saving yourself tens or hundreds of
00:28:54
millions of dollars so that's it's a
00:28:57
version of what CL did it takes an
00:29:00
enormous amount of technical strength to
00:29:02
do it it also takes tremendous I think
00:29:05
Executive courage and Leadership because
00:29:07
I think that's a very difficult decision
00:29:09
to embark on but if you're an engineer
00:29:12
that must be an unbelievably exciting
00:29:15
technical challenge to be a part of but
00:29:17
but that's the basic premise of what
00:29:19
they were able to
00:29:20
do hopefully they share more and maybe
00:29:23
they even open source what they did
00:29:24
because I think it would just be an
00:29:26
amazing
00:29:28
thing for all of us to look at yeah I
00:29:30
mean to to restate it watch people use a
00:29:33
piece of software and then based on what
00:29:36
they do you could write the code which
00:29:38
you could take a video of a video game
00:29:41
today like Angry Birds and somebody did
00:29:43
this you give the Angry Birds iPad you
00:29:46
know game from 15 years ago to AI it's
00:29:49
going to back into the code just by
00:29:52
watching it so why not just watch people
00:29:54
use Salesforce or workday and those are
00:29:56
very expensive products thousand of
00:29:58
dollars per user right I want to I want
00:30:00
to get Sax's point of view like the
00:30:01
thing in Enterprise software that we
00:30:02
were always told is you cannot touch
00:30:04
these systems of record don't ever start
00:30:06
a systems of record company don't try to
00:30:09
touch these systems of record companies
00:30:10
don't you know try to disrupt them it's
00:30:13
an impossible task but then the question
00:30:15
is if you have these
00:30:18
things why do you necessarily need a
00:30:20
system of record in the way that you
00:30:22
needed to before when you're writing all
00:30:24
this clunky deterministic I don't well I
00:30:27
saw the clar story where they said they
00:30:29
were going to rip out Salesforce and and
00:30:31
work day because they were able to write
00:30:33
their own bespoke code using AI I mean I
00:30:36
have to say I'm a little bit skeptical
00:30:37
of that story for a couple of reasons
00:30:39
one is if that's their goal why wouldn't
00:30:42
they have open- sourced this these
00:30:44
products they created you might as well
00:30:45
get the whole ecosystem working on it
00:30:47
because they're not trying to sell this
00:30:50
product that they've internally created
00:30:52
they're just trying to rip out the cost
00:30:53
so why not let the whole ecosystem see
00:30:55
it the other thing is if it's so easy to
00:30:59
do why hasn't the market already been
00:31:01
flooded with new startups that are
00:31:04
effectively able to reverse engineer I
00:31:06
don't think you're right I don't think
00:31:07
it's easy to do because I don't think
00:31:08
there's a generalization here that's
00:31:11
producti do you know what I mean like I
00:31:13
do think that these are very custom
00:31:15
specific things so maybe there's like
00:31:17
some scaffolding but I don't think that
00:31:19
that scaffolding has a ton of economic
00:31:21
value I think it's really good open
00:31:23
source stuff yeah I think it's what you
00:31:25
build on top of it and so that hasn't
00:31:28
been figured out yet for sure yeah look
00:31:30
I I think that if you're only using a
00:31:32
few use cases of these big complicated
00:31:35
software packages then yeah it's
00:31:37
probably easier than ever to deprecate
00:31:40
them you know eliminate them from your
00:31:42
stack and just have your own internal
00:31:43
engineers build specifically what you
00:31:45
need in a more tightly integrated way I
00:31:47
think that is possible Nick show this
00:31:50
tweet to these
00:31:52
guys here's the tweet this is this was a
00:31:56
crazy one yeah so so look at but look at
00:31:59
the code look at the actual product
00:32:01
itself for a second yeah but the
00:32:03
product's garbage I mean look how
00:32:05
ridiculous this is but that was 600
00:32:08
sorry it was a billion dollars that so
00:32:10
here's
00:32:12
theor paid Oracle 600 million to build
00:32:15
our course management portal it's built
00:32:17
on top of oracle's people soft Suite
00:32:19
which they refuse to customize without
00:32:21
an extra 400 million to hit 1 billion
00:32:23
New Yorkers got the image below and pay
00:32:25
5 million plus a year for hosting look
00:32:28
this this is egregious government waste
00:32:30
I mean that site looks like it's
00:32:33
pathetic I mean honestly this looks like
00:32:35
it a it could have been done with a
00:32:37
SharePoint site and you pay some
00:32:39
consultant to stand it up and for 1% of
00:32:42
the cost and to there are better plat
00:32:44
more modern platforms than that so this
00:32:47
is just incredibly wasteful and
00:32:50
inefficient government spending they're
00:32:53
going for retro they were going for
00:32:54
retro they wanted to hearken back to the
00:32:57
90s
00:32:58
the reason I wanted to I wanted to show
00:33:00
this to you is I think that these kinds
00:33:02
of things will not be possible in the
00:33:03
future I just don't see how one could
00:33:08
spend a billion dollars if one tried to
00:33:10
to enable that feature it be impossible
00:33:13
right but that that that 600 million
00:33:16
that was wasted on that um crappy portal
00:33:18
that shouldn't have happened even
00:33:20
without AI right because there's like
00:33:23
much better ways there you could you
00:33:24
could buy a much better product for 1%
00:33:27
of the cost so or .1% of the cost there
00:33:29
must be some regulatory capture going on
00:33:32
here where somebody's got a
00:33:34
reput Rel that's what I'm saying like a
00:33:36
10year relationship with somebody in
00:33:38
Albany that you know at previous Fraud
00:33:42
and Abuse it's the same thing that's
00:33:44
happening with um rural internet you
00:33:48
paradoxically is our next story so let's
00:33:51
go for it in related news of our
00:33:54
government burning our money we're all
00:33:57
Brad BR band rural broadband and EV
00:34:00
charging 42 billion and 7.5 billion
00:34:03
almost $50 billion combined let's just
00:34:06
go over these two programs real quickly
00:34:08
here both were part of the $1.2 trillion
00:34:11
infrastructure bill in 20121 42 billion
00:34:15
carved out to provide high-speed
00:34:17
internet to people living on farms in
00:34:20
rural locations 7.5 billion carved out
00:34:22
to build 500,000 EV Chargers over 10
00:34:25
years it's been a thousand days days
00:34:27
since the bill was passed so let's check
00:34:29
on the progress zero people have been
00:34:31
connected according to FCC commissioner
00:34:34
Brendan Carr and eight 1 2 3 4 5 6 seven
00:34:39
eight e Chargers have been built as of
00:34:41
May according to Auto Week Magazine
00:34:44
what's even crazier Private Industry
00:34:46
already solved these problems United
00:34:49
Airlines just announced they're putting
00:34:50
starlink on a thousand of their planes
00:34:54
and they're going to offer it for free
00:34:55
and starlink now has 2500 planes under
00:34:59
contract with a bunch of other Airlines
00:35:02
and uh in the second half of 2023 alone
00:35:06
the private sector built over a thousand
00:35:08
charging stations in the US these are
00:35:10
two problems that's have already been
00:35:13
solved sacks why are we burning $50
00:35:17
billion in the
00:35:20
future
00:35:22
with on things that have already been
00:35:24
solved we've solved for this you I own
00:35:26
electric cars I you know the answer you
00:35:28
know the answer say the answer Jason
00:35:31
corruption no come on Jason
00:35:34
incompetence really
00:35:37
graft keep going I mean you tell me
00:35:41
corruption graft buying votes from your
00:35:45
constituents they haven't they haven't
00:35:47
delivered any of it incompetence
00:35:50
yes well there's there's a couple things
00:35:52
going on here so one is typical
00:35:54
government waste Fraud and Abuse where
00:35:58
theyve allocated 42 billion for Rural
00:36:00
internet haven't hooked anyone up and we
00:36:02
could spend a fraction of that giving
00:36:05
people Starling and allowing the private
00:36:08
sector to do its job and why even pay
00:36:10
for it tax why are we paying for it if
00:36:12
it's available for 100 bucks that that's
00:36:14
the Baseline but it's worse than that
00:36:17
because on top of the waste Fraud and
00:36:18
Abuse and the fact that the government
00:36:20
is grossly incompetent inefficient you
00:36:23
also have naked political retaliation
00:36:25
going on here Ah that's the answer yeah
00:36:28
exactly and Brendan Carr who's an FCC
00:36:29
commissioner pointed this out he said
00:36:32
that in
00:36:33
2023 the FCC cancelled or revoked an
00:36:37
$885 million contract with the company
00:36:40
by claiming starlink is not capable of
00:36:42
providing high-speed internet then a
00:36:45
year later yeah of course that was a lie
00:36:48
and then a year later the FCC is now
00:36:50
claiming that Starling provides so much
00:36:52
high-speed internet that the word
00:36:53
monopoly should be uh tossed out yeah so
00:36:57
look this is just it's pure naked
00:37:00
retaliation the the Biden Harris
00:37:02
Administration doesn't want to admit
00:37:04
that Elon has the best solution for
00:37:07
Rural internet just like they couldn't
00:37:09
admit he made the best electric cars
00:37:11
remember when they did that EV Summit
00:37:12
and they didn't invite him that was just
00:37:14
nakedly political um because
00:37:17
he's so look I mean the the Biden Harris
00:37:20
Administration it look it's blue no
00:37:22
matter who and Elon has drifted from
00:37:26
being sort of in independent and
00:37:29
not he was blue what it is Ian voted for
00:37:33
Hillary and Obama he said he's no longer
00:37:35
team blue and so they're punishing him
00:37:36
for this yeah and it's costing taxpayers
00:37:39
a huge amount of money I I think this is
00:37:41
one of the worst Decisions by the
00:37:42
current Administration and if Trump gets
00:37:45
in there he should reverse it on day one
00:37:47
well I wean need to investigate I mean I
00:37:48
think how we got to the point of wasting
00:37:51
50 billion
00:37:53
dollar that requires an investigation I
00:37:56
think chath your thoughts one comment is
00:37:58
and this is so sad but I'm so
00:38:00
desensitized by the amount of waste that
00:38:03
I don't know whether 50 billion is a lot
00:38:05
or a little anymore when it comes to the
00:38:07
United States government isn't that sad
00:38:09
like because now everything I hear is
00:38:11
hundreds hundreds of billions and
00:38:13
trillions but 50 billion is an enormous
00:38:17
amount of money right well that that's
00:38:19
such a good point and I remember you
00:38:20
know back in the day 60 Minutes used to
00:38:24
do these segments on waste Fraud and
00:38:27
Abuse at the Pentagon different parts of
00:38:28
the government $42 billion just spent on
00:38:31
something that really taxpayers could
00:38:33
have for free or without the government
00:38:35
getting involved and you know 42 billion
00:38:37
that was lining someone's pocket when
00:38:39
the service doesn't even work that would
00:38:40
have been a scandal and the media would
00:38:42
have covered it but the media doesn't
00:38:44
even cover it these days and again it's
00:38:45
because the media has become so tribal
00:38:48
that it's better dead than red and blue
00:38:51
no matter who and so because the media
00:38:53
would have to admit that elon's already
00:38:55
solved this problem they just can't go
00:38:57
there they won't even cover this and so
00:38:59
we have no accountability there's no
00:39:01
accountability on the government if I
00:39:03
had to just take a step back and just
00:39:05
generalize going forward do we want to
00:39:08
live in the kind of administrative state
00:39:13
where they will pick
00:39:16
people that they dislike based on
00:39:20
totally random criteria a tweet a meme a
00:39:25
post and then all of a sudden punish a
00:39:28
bunch of the rest of us because of that
00:39:31
they're punishing all of America because
00:39:34
they collect our taxes to waste on it
00:39:36
and then they punish the people that
00:39:38
they actually say they're going to
00:39:40
uplift by not delivering what they
00:39:41
promised and if you take Elon out of it
00:39:44
for a second the the problem was when we
00:39:47
crossed the chasm and did it with the
00:39:49
first guy him but the reality is there's
00:39:52
only one of him and then there's a lot
00:39:53
of the rest of us and what will happen
00:39:55
is people just get added to this list of
00:40:00
folks that certain nameless faceless
00:40:04
people in the administrative State
00:40:06
dislike and what happens is the country
00:40:08
slows down and the country wastes money
00:40:10
and the country pilers it away and that
00:40:13
has to stop and so what really bothers
00:40:15
me about these things is a I don't know
00:40:18
how to UND desensitize myself to the
00:40:20
fact that all of a sudden now because of
00:40:22
just all of this sloppy waste I didn't
00:40:26
react as much as I should have to just
00:40:28
$50 billion being flushed down the
00:40:30
toilet on these two projects and then
00:40:32
two Jason your point it is a solved
00:40:35
problem that you can give incredibly
00:40:39
cheaply and the fact that it's not left
00:40:42
to private Enterprise to solve this and
00:40:44
instead it's just Brazen partisanship
00:40:47
combined with retaliation combined with
00:40:50
incompetence votes by giving this money
00:40:52
to other vendors who are giving them
00:40:55
donations and just to give the Democrat
00:40:58
what happens if then Trump does the same
00:41:00
thing for a solution that you support
00:41:02
and you need and you think should be
00:41:03
everywhere the point is we don't want
00:41:06
any of this stuff under any
00:41:08
Administration andar it's and the minute
00:41:11
that one Administration breaks the seal
00:41:14
and makes it acceptable it becomes part
00:41:17
of the water table and that's the real
00:41:19
problem we broke the seal on this crazy
00:41:22
multi- multi- trillion dollar spending
00:41:24
and it is just never stopped since then
00:41:27
and you know the incentives really
00:41:28
matter uh if you look at a private
00:41:31
company if you were at clar and to our
00:41:33
previous story and you go to the boss
00:41:35
and say I know how to get rid of these
00:41:38
this wasteful spending we're doing here
00:41:39
we can get rid of all tier one calls
00:41:41
with AI and save that money you get a
00:41:43
promotion if you're in the
00:41:45
government you can't if you're a
00:41:48
politician and you cut this program your
00:41:50
constituents get upset you don't have
00:41:53
that stuff being built in your District
00:41:55
there's a perverse incentive that you
00:41:56
can't buy buy the votes which is why
00:41:58
these folks are constantly trying to buy
00:42:00
votes and the good news is the good news
00:42:03
is I I really applaud the people that
00:42:06
have the courage to show the stuff on X
00:42:09
to tweet this stuff out so that the rest
00:42:11
of us know about it and the person that
00:42:13
talked about the NYC thing but then the
00:42:16
next step has to happen which is that we
00:42:18
all need to decide that this stuff needs
00:42:20
to stop other it's going to bankrupt our
00:42:22
country and we have to celebrate it
00:42:24
that's the key if we can celebrate
00:42:26
people save money again like mle is
00:42:29
getting a lot of credit and that's up to
00:42:31
us leadership in podcasting or the media
00:42:34
or influential people who have
00:42:35
followings if you point out hey this is
00:42:37
a waste go save this money and somebody
00:42:39
does save the money well why don't we
00:42:41
start celebrating people saving the
00:42:42
money and doing the right thing here
00:42:44
because this is our children's future is
00:42:47
it true that kamla was the Broadband Zar
00:42:50
that was responsible for this thing I
00:42:51
mean it's who knows it's just no because
00:42:54
I saw it I I saw that a bunch of
00:42:56
senators wrote letter to her and they
00:42:59
claimed that she was the Broadband Zar
00:43:01
but I don't know if that's true or not
00:43:02
true and whether she was remember she
00:43:05
was the AIS are I mean the
00:43:07
administration did put her nomy in
00:43:10
charge of various technology initiatives
00:43:12
here's an idea save money get the get
00:43:15
the best solution at the lowest price
00:43:17
and then re-evaluate that as you go and
00:43:19
I just want to point out with the it's a
00:43:21
this is a a subtle point but Elon also
00:43:25
open sourced his patents for the
00:43:27
superchargers and let anybody do them
00:43:29
and he opened up the superchargers to
00:43:31
other vehicles which he didn't have to
00:43:33
do and when they gave him a loan back in
00:43:37
the cindra days and the Fisker days
00:43:39
remember they gave these incentives in
00:43:40
the form of loans he's the only guy who
00:43:42
paid it back everybody else failed so
00:43:44
now you're punishing the guy who
00:43:46
actually built the infrastructure for
00:43:48
both of these projects so the reward for
00:43:51
actually doing the right thing which
00:43:52
starlink did SpaceX did and Tesla did is
00:43:56
to be punished and then you're giving a
00:43:58
leg up to somebody else who's building
00:43:59
these Char who's more qualified to build
00:44:01
these charges at scale or a satellite
00:44:04
Network at scale the person who's
00:44:05
already done it he's already done it I
00:44:07
do worry that there's a growing version
00:44:10
of the Elon derangement syndrome that's
00:44:13
also kind of like festering yeah for
00:44:16
sure which just it just stops people
00:44:18
from thinking rationally of course I
00:44:21
mean we're talking about laying fiber
00:44:23
lines cable modems to people who are
00:44:26
hundreds of miles into the countryside
00:44:29
that makes no sense when you can just
00:44:31
beep put a satellite dish up
00:44:33
today what are we even talking about I
00:44:36
mean government has never been
00:44:37
particularly efficient but there was a
00:44:40
period of time where people would at
00:44:41
least care about wanting to make it more
00:44:44
efficient and it would be a scandal if
00:44:47
there was political corruption to try
00:44:48
and bias the result in a way that
00:44:51
actually deprived the intended
00:44:53
recipients of the program from getting
00:44:55
the services they were supposed to get
00:44:56
and cost cost the government way more
00:44:58
money than it needed to we're so far
00:45:00
beyond being that country anymore where
00:45:04
we actually debate the best policy we're
00:45:07
now it's just like we're Waring
00:45:09
political tribes and the objective of
00:45:12
the party is to punish its political
00:45:16
opponents to engage in retaliation and
00:45:18
to basically loot the public coffers as
00:45:21
much as possible on behalf of their
00:45:22
constituents and that's what's basically
00:45:24
happening you know it's completely
00:45:26
dysfunctional well let's use this
00:45:27
podcast if you see government ways tell
00:45:29
us no really cares because the media
00:45:31
doesn't really shine a light on it
00:45:33
because they're they're completely
00:45:34
tribalize as well I agree with
00:45:36
everything you're saying except the last
00:45:37
part I don't think it's on behalf of
00:45:38
their constituents I don't think any of
00:45:40
us see any benefit from any of this
00:45:42
spending no no I meant they're donors
00:45:44
the they donor constituents they not not
00:45:47
the citizens of the country but who but
00:45:50
who's winning in this it's not like this
00:45:52
42 is is 42 billion lining the pockets
00:45:55
of I don't know name me how do you think
00:45:58
all those fiber companies that are going
00:45:59
to lay that fiber are going to get that
00:46:00
money and then and then they're going to
00:46:03
cont it's been it's been three or four
00:46:05
years they they haven't done a single
00:46:06
thing I mean I still think they're cash
00:46:09
the checks yeah it seems like we're at
00:46:11
the stage of just pure incompetence and
00:46:13
retaliation we're not even at the stage
00:46:14
of actually even giving it to anybody
00:46:16
else I mean that would be so they're
00:46:19
giving the money away and they're so
00:46:21
incompetent they're not getting the
00:46:22
political benefit from it no they're so
00:46:24
incompetent they can't get out of their
00:46:25
own way but somebody's getting that call
00:46:27
it 50 billion that we don't need to
00:46:30
spend and the way that money is awarded
00:46:33
is going to be political we're going to
00:46:35
think that they're going to turn around
00:46:35
and give big political contributions of
00:46:38
course they well I think I think that I
00:46:39
think the good news is that the more of
00:46:41
these things we shine a light on the
00:46:43
harder it'll be
00:46:44
to hide when these grants are actually
00:46:47
given or what the execution is and
00:46:50
running list let's start a running list
00:46:52
no to your point Sak maybe like you know
00:46:54
we need a Revival of the 60 minutes you
00:46:57
know waste Fraud and Abuse on this
00:46:59
program we'll do it at the end of the
00:47:00
show every time we have a running list
00:47:02
at allin.com of just every of one of
00:47:05
these scandals and we'll feature it so
00:47:06
leak it to us first send it to us my DMs
00:47:09
are open all right listen early stage
00:47:11
investing has always been hard there was
00:47:13
a tweet storm this week that y
00:47:14
combinator might be having a hard time
00:47:17
replicating their early success we'll
00:47:19
discuss it now a thread this week from
00:47:21
ex user molsen Hart caught a couple
00:47:23
people's eyes he made the case that it's
00:47:25
been a rough decade for y based on the
00:47:27
accelerator top companies page YC list
00:47:30
as top companies by 2023 Revenue there
00:47:34
and uh you'll notice there's not a lot
00:47:35
of companies from the recent cohorts out
00:47:37
of the 50 companies featured only three
00:47:38
are from the classes after 2020 most of
00:47:41
them being from the early 20110 10
00:47:44
obviously that's because they've been
00:47:45
around longer but it sparked a big
00:47:48
discussion that there were so many
00:47:49
winners from the 2009 to 2016 era and
00:47:53
that uh maybe the class size at YC has
00:47:56
expanded a whole bunch and maybe that's
00:47:59
part of the problem but there's a bigger
00:48:00
problem in VC that we've talked about
00:48:02
here here's a chart from Carta that just
00:48:05
shows the percentage of VC funds that
00:48:07
have made a distribution since
00:48:09
2017 over 40% of 2018 vintage funds have
00:48:13
not made a single distribution yet uh
00:48:16
and it's getting to the point year five
00:48:18
six or seven where you probably should
00:48:19
have had some distributions occur
00:48:22
obviously a lot of this has to do with
00:48:24
maybe m&a and those early winds being
00:48:26
taken off the table we've talked about
00:48:28
that a whole bunch but here's the chart
00:48:32
that kind of gets really interesting an
00:48:35
explosion in fund managers occurred as
00:48:37
we all know and this chart shows from
00:48:39
pitchbook the first time first time VC
00:48:42
managers that raised a second VC fund as
00:48:44
a share of all firsttime VC managers and
00:48:47
it's now down from above 50% to
00:48:51
below gosh 15% so what are your thoughts
00:48:55
here Chima my gosh Venture is a really
00:49:00
really tough business every year for the
00:49:04
last seven six years seven years I have
00:49:07
published my returns which most VCS
00:49:10
don't want to
00:49:11
[Music]
00:49:12
do I do it because I go back and I look
00:49:15
at it and I think
00:49:17
having public accountability actually
00:49:19
drives some good decisions they they may
00:49:22
seem suboptimal in the moment but they
00:49:26
in in the long run turn out to be good
00:49:28
decisions and the biggest one has been
00:49:31
generating liquidity so Nick you can
00:49:34
throw up this thing but I'm sure there
00:49:36
are funds in each of these vintages that
00:49:38
have done way better than me so I'm not
00:49:40
I'm not saying you know it is what it is
00:49:42
but what I want to point out is if I go
00:49:45
and look inside of these funds and tell
00:49:47
you how hard it has been to generate
00:49:49
this DPI it's like it's like dragging an
00:49:55
entire just
00:49:58
sack of potatoes over the Finish Line
00:50:01
like like a truck of dead bodies over a
00:50:03
Finish Line it is super super hard and
00:50:08
the things that we have fought are
00:50:10
two one is that the gation of companies
00:50:14
has totally blown out we used to be in a
00:50:17
world where by year five six or seven
00:50:19
you could return money you just can't do
00:50:21
that anymore unless you get
00:50:22
extraordinarily lucky which by the way I
00:50:25
got when saaks was running Yammer
00:50:28
it was an enormous win for all of us but
00:50:31
that is just exceptionally rare and that
00:50:33
was m in year what five or six sacks
00:50:36
there's so few there's so few
00:50:37
entrepreneurs capable of that he's one
00:50:39
of maybe five or 10 so other than that
00:50:43
I've never really had a company that has
00:50:46
generated liquidity in year five six or
00:50:48
seven they've always generated if
00:50:50
they've generated it at all in years 11
00:50:53
12 and 13 and so the problem with that
00:50:57
is that at some point you have these
00:50:59
paper marks that say you're winning and
00:51:02
things are working but there's no path
00:51:05
to liquidity so then I what I did was I
00:51:08
stepped in to the secondary markets and
00:51:10
I would
00:51:11
sell and it would really upset certain
00:51:16
Founders but I was very clear that when
00:51:19
I was running outside capital and I was
00:51:21
running outside capital on behalf of
00:51:23
really organizations that I believed in
00:51:26
the broad Foundation the Mayo clinics
00:51:28
Memorial slone ketering my job was to
00:51:30
get them money back you know these were
00:51:33
their Pension funds these were the
00:51:34
things that they use to build facilities
00:51:36
cancer research cancer research I didn't
00:51:39
have the you know ability to just sit on
00:51:42
my hands and say oh you know what you're
00:51:44
15 don't worry so it it's just meant to
00:51:47
say that the the tactics of generating
00:51:50
liquidity and Venture are very
00:51:53
misunderstood and very underappreciated
00:51:57
and even then you sell something that
00:51:59
are just absolute winners that had you
00:52:01
waited another five or six years would
00:52:04
have turned another you know one or two
00:52:05
turns but that's not the job the job is
00:52:09
not to maximize absolute every single
00:52:12
win the job is to return capital in a
00:52:14
reasonable time period so that your
00:52:17
investors don't run out of money to give
00:52:19
you yeah it's a it's a tough game man it
00:52:22
is really really really tough yeah and
00:52:24
the the insight and and Sorry by the way
00:52:26
and I feel this now because you know the
00:52:28
last five or six years has been entirely
00:52:29
my own capital and my gosh it's hard
00:52:32
yeah managing liquidity is it's
00:52:35
impossible especially when you can't
00:52:36
rely on anybody else so well and thank
00:52:39
God for the secondary markets even
00:52:41
emerging because at the same time that
00:52:43
the secondary markets emerged and people
00:52:45
were willing to buy Venture assets you
00:52:48
know going into their second decade I
00:52:50
would have been in real trouble without
00:52:51
the without reasonably liquid myself
00:52:54
included I mean my numbers my numbers
00:52:55
would be a quarter of what they are yeah
00:52:58
and I I took advantage of almost every
00:53:00
time I had one of those opportunities to
00:53:02
sell some shares pair some positions and
00:53:04
that's how we got our DPI as well
00:53:06
because let's face it Lina con and the
00:53:08
anti-tech sentiment has led to these
00:53:12
large companies not buying startups and
00:53:15
instead they compete with them they just
00:53:16
say We'll build it inhouse because
00:53:18
you're not letting us buy it and it's
00:53:20
broken the entire ecosystem now that's
00:53:23
broken the the IPO process is broken
00:53:27
I
00:53:28
tried to flip that on its head with
00:53:32
backs you know some worked some didn't
00:53:34
many didn't in the end many of mine
00:53:36
didn't work out at the end there was a
00:53:38
period where it looked like it was
00:53:39
working but these are all attempts at
00:53:42
changing the liquidity cycle yeah of
00:53:45
these companies because the way that
00:53:47
things stand today we are not in a
00:53:50
sustainable industry it is if you raise
00:53:53
funds and think about fee generation but
00:53:55
it is not if you think about returning
00:53:57
money to Founders LPS getting employees
00:54:00
compensated for many years of you know
00:54:02
toil that they put in it's very tough
00:54:05
game right now well Sachs right now
00:54:06
we're seeing people do things like
00:54:09
selling you know their early SpaceX or
00:54:11
their early stripe whatever it is to
00:54:14
other VCS to later stage funds a lot of
00:54:17
ways to try to secure DPI what's your
00:54:19
thoughts on the state of venture today
00:54:22
given all this data that we're looking
00:54:24
at today well two two points so first I
00:54:27
agree with chamath that the amount of
00:54:29
time it takes to generate an outcome for
00:54:32
I'd say most startups is longer than the
00:54:34
10-year period of these funds and these
00:54:37
funds can be extended up to 12 years
00:54:39
usually but then what do you do after
00:54:40
that I this takes a lot longer than that
00:54:43
in a lot of cases to generate a
00:54:44
meaningful outcome I just had two
00:54:46
companies that I invested in in my
00:54:49
second fund so in 2019 and 2020 so four
00:54:52
years ago and five years ago just got
00:54:55
marked up and it was a big markup the
00:54:57
company's doing well I call them late
00:54:59
bloomers it took four to five years for
00:55:02
them to accomplish what they wanted to
00:55:04
in terms of like building out the tech I
00:55:05
mean I invested at like the earliest
00:55:06
stage so that's how long it took and now
00:55:09
they just did growth rounds and they're
00:55:10
kind of Off to the Races but you know I
00:55:13
could easily be 10 years from here to
00:55:15
get to yeah a liquidity event so you're
00:55:17
talking about more like 15year funds so
00:55:19
I agree with that point the second thing
00:55:21
though is that the big thing that's
00:55:24
happened in our industry is we had a
00:55:25
bubble
00:55:26
in 2020 and especially
00:55:29
2021 and we just had a ton of capital
00:55:32
come into the industry because the fed
00:55:34
and the the um federal government aird
00:55:37
dropped 10 trillion do liquidity onto
00:55:39
the economy in reaction to covid and not
00:55:43
all that money went into VC it went into
00:55:44
a lot of places but the VC industry was
00:55:47
flooded with cash and you see this in
00:55:49
the deployments I mean in those bubble
00:55:51
years there was something like 200
00:55:53
billion a year of capital deployment
00:55:55
when normally it's 60 to 100 billion so
00:55:58
if twice the amount of money is going
00:56:00
into the industry and is being deployed
00:56:02
and rounds are now twice as big and
00:56:04
valuations are twice as big that has a
00:56:06
huge outcome a huge effect on returns so
00:56:10
for example the average Venture fund is
00:56:12
like a 2X return but if the entry prices
00:56:15
were artificially double then there goes
00:56:18
your return right there you get 2X 1X so
00:56:21
I think we're just in The Hangover of
00:56:23
this massive liquidity bubble that
00:56:26
didn't originate in the Venture Capital
00:56:28
industry it came from frankly the
00:56:30
federal government but we're just
00:56:32
Downstream of that now what I would say
00:56:34
is I I do think we're at the tail end of
00:56:36
working that out and the good news is
00:56:39
that we now have maybe the most exciting
00:56:41
Tech wave ever which is AI definitely
00:56:44
the most exciting Tech wave since the
00:56:45
internet came along in the mid to late
00:56:48
90s so the hope is we're finally going
00:56:51
to have like really exciting things to
00:56:52
invest in again but but yeah look I
00:56:56
think think we're at the tail end of the
00:56:57
last cycle and the beginning of a of a
00:56:59
new cycle and vintage Distortion is so
00:57:02
real you know it's very hard to
00:57:04
understand how each of these vintages
00:57:06
with your late bloomers or overpriced
00:57:08
things company's getting hundred million
00:57:11
rounds totally at a billion dollar
00:57:13
valuation before they have product
00:57:14
Market fit and those
00:57:16
distortions were just so pronounced the
00:57:18
last five to 10 years that we're now
00:57:20
sorting them out like a like a house of
00:57:23
mirrors where you don't know who's tall
00:57:24
who's fat who's skinny what the reality
00:57:26
is here and the other big thing is this
00:57:28
peanut butter effect that you know I
00:57:30
tweeted about today you know during Peak
00:57:32
Zer you had all these exceptional team
00:57:35
members you know the number two three
00:57:38
four five person at a company that was
00:57:40
doing great they would leave to start
00:57:42
their own company so the talent got
00:57:43
spread then you had so many of these
00:57:46
Founders rushing into the same vertical
00:57:48
so you'd have 20 startups because there
00:57:50
was too much Capital pursuing the same
00:57:51
opportunity you pursu the same
00:57:53
opportunity what happens to earnings
00:57:56
they get spread then what happens to
00:57:58
customers they get spread across 20
00:58:00
different products competing for the
00:58:02
same customer and then what happens with
00:58:04
you know ownership Stakes for us as GPS
00:58:07
and LPS chamath the ownership Stakes
00:58:10
because the valuations went up so much
00:58:12
they got spread like peanut butter and
00:58:14
instead of a series a getting you 20% of
00:58:16
a company it got you 10 instead of a c
00:58:18
check getting you 5% it got you one
00:58:20
there's no DPI possible you nailed it
00:58:23
and saaks nailed it but but and the
00:58:24
thing to remember is both of those two
00:58:26
things now work together to erode the
00:58:28
return stream for the general partner
00:58:31
but really most importantly for The
00:58:33
Limited partner so I I do think that we
00:58:36
are in a situation where the average
00:58:38
returns are going to Decay by 50 to 100%
00:58:41
because of what sack said and because of
00:58:43
what you said on top of that I don't
00:58:46
think we know what the actual cap
00:58:48
structure needs to be for a successful
00:58:51
AI company is it 20 people that does the
00:58:54
work of 2,000 now because they all of
00:58:56
these agents and systems that work on
00:58:58
their behalf if that's true giving that
00:59:01
company hundreds of millions of dollars
00:59:03
is actually the opposite of what you
00:59:04
want to do you want to give that company
00:59:06
10 or 15 and then let them cook and so
00:59:10
we have a we have a right sizing of
00:59:12
capital problem that needs to happen the
00:59:14
data would tell you though that the
00:59:15
industry understands that so the fact
00:59:17
that we've gone from 50% of people being
00:59:19
able to raise a fund to 12% means that a
00:59:23
lot of people will get washed out of the
00:59:24
industry less Capital being raised which
00:59:28
probably is foreshadowing the fact that
00:59:30
these companies will need a lot less
00:59:32
Capital but you know that has a lot of
00:59:34
implications as it ripples through our
00:59:35
economy it has I think it's very good
00:59:37
for the early stage I think you know you
00:59:39
guys are very good there you've talked
00:59:41
about how it's good for you it's very
00:59:43
complicated I think for the expansion
00:59:44
and growth stage capital and then I
00:59:47
think it's going to be there's going to
00:59:49
be another turn on what happens on the
00:59:51
IPO markets because you can't have so
00:59:54
many companies waiting
00:59:57
with very very few ways of accessing
01:00:00
Public Market capital and exposure I
01:00:02
just think this is that is that is
01:00:04
fundamentally broken and we're going to
01:00:06
have to reinvent we tried once with
01:00:08
spaxs we're going to have to go back to
01:00:10
the drawing board and try again listings
01:00:12
secondary markets that are more fluid I
01:00:14
don't know what it is but we need to do
01:00:15
something because the status quo doesn't
01:00:17
work I think there's a lot so many good
01:00:19
points that we're hitting here I I'll
01:00:21
just say the the other thing to build on
01:00:24
your point about hey these take less
01:00:26
Capital you have to look at what does
01:00:28
your ownership after you've been diluted
01:00:30
half by 50% as a seed or series a
01:00:34
investor you're going to be down to half
01:00:35
so if you own 10% you own five if you
01:00:37
own seven like YC or we do in a company
01:00:40
you're going to own three you're going
01:00:42
to really have to model out is the
01:00:44
valuation you're looking at what does it
01:00:46
pencil out to for an outcome and when I
01:00:48
did this with our investments I saw a
01:00:50
leak in my game which was hey I'm
01:00:52
putting a 100K into a $25 million round
01:00:54
or a $50 million round as a followon
01:00:56
investment you know to support the
01:00:58
founder okay what does that do for my
01:01:00
LPS well that 100K would need to hit
01:01:03
some extraordinary outcome 5 10 2040
01:01:07
billion doar in order for us to return
01:01:09
the fund so now my team understands hey
01:01:12
take that 125k that 250k that 500k do
01:01:15
more four do four more accelerator
01:01:17
companies with it because those could
01:01:19
return the fund and that's that fund
01:01:22
math people stop doing I think all these
01:01:24
fund managers who are getting wiped out
01:01:27
they never penciled out what does this
01:01:29
company I'm giving a million dollars
01:01:31
need to hit in order for me to return my
01:01:33
fund and now they're finding out that
01:01:36
look just tweeted look at that let me
01:01:39
see you know everybody's course
01:01:41
correcting I mean it's basically the
01:01:42
capital deployments gone back to where
01:01:45
it was in 2019 let's call it so again we
01:01:48
had this bubble the foam started
01:01:50
building in 2020 but you had Co people
01:01:52
didn't know what to think so there was
01:01:54
some restraint I guess and then 2021 it
01:01:57
just went
01:01:58
wild that was nuts man midle vintages
01:02:02
are just going to be garbanzo beans 21
01:02:04
22 well you know that's such an
01:02:06
interesting point if you could return
01:02:08
Capital you're gonna look like a
01:02:09
Euro also chth I remember I don't know
01:02:13
if it was Michael Moritz or or Doug
01:02:15
Leone but I was talking to seoa about
01:02:18
the time dispersion of your fund like
01:02:20
over what period of time are you
01:02:22
deploying a fund and man people started
01:02:24
deploying funds in 18 months because
01:02:27
they could raise the next fund so quick
01:02:28
so like screw it I'm going to deploy
01:02:30
this Fund in 18 months 24 months and LPS
01:02:33
were saying to me like how what period
01:02:35
are you going to deploy this and I said
01:02:36
well you know I was taught by Fred
01:02:37
Wilson and this person 36 months 48
01:02:40
months would be a good window to deploy
01:02:43
Capital because you know it SMS it out I
01:02:46
think you're saying the the Dirty Little
01:02:48
Secret of the Venture business which is
01:02:50
at some point people get to a fork in
01:02:51
the road if they hyper optimize for
01:02:54
returns I'll put benchmark I'll put Fred
01:02:57
Wilson in USV I'll put sequoia's early
01:03:00
stage fund they have to introduce time
01:03:03
diversity they keep the funds small and
01:03:06
they look to hit Grand
01:03:08
Slams but there are many other people
01:03:11
and I would say the most of the set
01:03:13
outside of that take the road more
01:03:17
traveled which is then you optimize for
01:03:19
size which then becomes a fee game and
01:03:21
so you optimize for velocity get the
01:03:23
funds out as quick as possible raise a
01:03:25
new f they have no intention of
01:03:27
generating returns because they have no
01:03:29
ability to when you have absolutely no
01:03:31
time diversity in this business in a
01:03:33
pool of capital you're giving away one
01:03:35
of your best edges David just talked
01:03:36
about it as a smart practitioner he was
01:03:39
able to nurture these companies and all
01:03:40
of a sudden they start to win if you've
01:03:42
all of a sudden flushed all your money
01:03:44
in fund one then you go to fund two fund
01:03:46
three by the time something in fund one
01:03:48
hits what are you going to do you're
01:03:49
going to cross the funds or you're going
01:03:52
to justify taking money from the left
01:03:54
hand to pay the right hand or just going
01:03:56
to let your ownership Wayne because you
01:03:58
frittered all the money away these are
01:04:00
all the problems that most of these
01:04:01
folks have encumbered themselves with
01:04:04
it's very difficult to get out of it's
01:04:05
going to take now look In fairness to
01:04:07
them they probably you know got good
01:04:10
while the getting's good so they'll make
01:04:11
a ton of money and fees but they will
01:04:14
not be able to raise funds and those
01:04:15
fees are not clawed back folks for those
01:04:17
of you playing along at home just just
01:04:18
by the way I feel better about those
01:04:20
late bloomers in my portfolio because I
01:04:23
know the marks are real because if
01:04:24
they're getting marked up now now then
01:04:26
it's very very solid compared to frankly
01:04:30
some of those marks that we got in the
01:04:32
bubble year like 20201 I call them tiger
01:04:34
marks whether it was tiger or not it's
01:04:37
just less real quite frankly and a lot
01:04:39
of those companies are retrenching and
01:04:41
have issues so a mark now it just means
01:04:43
something different than a mark then but
01:04:46
look I want to you know just so we're
01:04:48
not like totally beating up on VC there
01:04:51
was you remember that in this bubble
01:04:53
period of September 2021
01:04:57
everybody thought that this party would
01:04:59
just continue forever and this is a good
01:05:01
example from The Wall Street Journal
01:05:03
where I was talking about how University
01:05:04
endowments were minting billions in
01:05:06
Golden Era Venture Capital so the bubble
01:05:09
wasn't just in VC it was in the public
01:05:11
markets too because we had Zer right
01:05:13
like interest rates were zero liquidity
01:05:15
was just flowing and so it was very easy
01:05:18
for companies to get liquid they ipoed
01:05:21
and then the valuations were
01:05:23
stratospheric so the distributions LPS
01:05:26
were massive in 2021 and then that led
01:05:29
to again more funds be able to raise
01:05:32
bigger funds everyone was just kind of
01:05:33
paying it forward and thought the party
01:05:35
would just keep going so this is what
01:05:38
happens in a bubble is everybody thinks
01:05:40
that it's just going to keep going like
01:05:42
that this is why it's so important as a
01:05:45
fund manager or an entrepreneur for you
01:05:47
to get great advice from people who've
01:05:49
been at this for a long time and focus
01:05:51
on the process you cannot control all
01:05:54
these outcomes you cannot control all
01:05:56
these meta events what you can control
01:05:58
is your relationship with your customers
01:06:01
building a team making great bets
01:06:04
supporting late bloomers that's the
01:06:06
critical part of all this is the process
01:06:08
and you can make your process better and
01:06:11
so with my team internally I'm
01:06:13
constantly talking to them about our
01:06:15
selection of companies how we help
01:06:17
companies get pulled through and get
01:06:19
Downstream funding how we literally our
01:06:21
big effort this year is how do we
01:06:23
introduce our companies to the top VC
01:06:25
fir terms and we've been working on that
01:06:28
as a internal project right of just
01:06:31
getting our great breakout companies to
01:06:34
the best investors to increase our pull
01:06:37
through it is a process and you have to
01:06:40
trust and focus on your process yeah
01:06:42
well look ironically just I mean just to
01:06:46
end on sort of a positive note if these
01:06:48
interest rate cuts are real like if we
01:06:50
we just got 50 if we get another 50 this
01:06:52
year if inflation's really Tamed and are
01:06:56
never going to go to zero but if they go
01:06:58
down
01:06:59
substantially and we have this new AI
01:07:02
disruption this new AI Tailwind we could
01:07:05
be back in another Golden Era it's not
01:07:08
going to be a bubble but it's could be
01:07:09
another Golden Era so we'll see start
01:07:12
companies from your lips to God's
01:07:15
ears love you guys I gotta go love you
01:07:18
all right shth had to go do work
01:07:20
apparently he's starting this New
01:07:22
Concept Sachs which jth is actually
01:07:23
going to work and uh uh at a company uh
01:07:27
we never got to talk about the uh debate
01:07:29
because we were busy doing the summit
01:07:31
and we took the week off from a new
01:07:32
episode uh people wanted to hear your
01:07:34
take what did you think of kamla and
01:07:37
Trump the one and only debate we're
01:07:39
going to hear
01:07:40
apparently any any thoughts I think
01:07:44
that KLA Harris performed better than
01:07:47
expected she did that I think mostly
01:07:50
through having canned answers to topics
01:07:55
and she was able to kind of memorize
01:07:56
those answers and and say them and she
01:08:00
was never knocked out of her preparation
01:08:02
she was well prepared yeah I think she
01:08:05
was well prepared however we now know
01:08:06
that these were canned answers because
01:08:08
in subsequent press interviews she gives
01:08:10
us the exact same thing it's like a
01:08:11
jukebox where you just push the button
01:08:13
get the same answer exactly so she's
01:08:16
she's memorized a certain number of
01:08:19
talking points and that's all she's
01:08:20
going to give you no matter what the
01:08:22
question is and if you saw that it's
01:08:24
become a meme now where
01:08:26
if you saw that question when she was
01:08:27
asked about inflation there's a pause
01:08:29
when she's figuring out which greatest
01:08:31
hit she's going to play and then you
01:08:33
know she I guess pushes b26 in her head
01:08:36
and then it begins so I was born in the
01:08:39
middle class and it's working apparently
01:08:42
right it seems like it's it's helping
01:08:44
her yeah I think what you saw is that
01:08:46
she got a bounce out of the debate but
01:08:48
now it's sort of like a lot of these um
01:08:51
bounces there's been kind of a
01:08:53
effervescence to it and then it kind of
01:08:55
Settles down back to their occurring
01:08:57
pattern and so I think the election is
01:09:00
extremely close but I don't think oh
01:09:01
yeah I mean every day it's like a poll
01:09:04
going one way or the other I mean this
01:09:05
is the closest of our lifetime maybe or
01:09:08
that I can remember I mean it's nuts how
01:09:10
this thing has flipped over and over
01:09:12
again what did you think of Trump's
01:09:13
performance were you disappointed there
01:09:15
were some rumors uh people were a little
01:09:17
upset that he doesn't prep as much as he
01:09:19
should what what's your what's your
01:09:21
advice there you know well I mean I
01:09:24
think that uh he was in a very difficult
01:09:26
situation you basically had a
01:09:27
three-on-one situation where he was up
01:09:30
against not just KLA Harris but the two
01:09:32
debate moderators it turns out that
01:09:34
lindsy Davis is a comma sorority sister
01:09:38
David Mur was factchecking him
01:09:41
constantly and some of those fact checks
01:09:44
weren't even correct um for example we
01:09:46
now know that the Springfield City
01:09:49
manager has acknowledged complaints
01:09:51
about pets being eaten far I was
01:09:54
wondering if we were going to get
01:09:55
through
01:09:55
it's as far as far back as March there
01:09:58
are videos of him talking about the
01:10:01
complaints at city council meeting now
01:10:03
you can you can say that you don't
01:10:06
believe those stories or whatever but
01:10:08
those reports were real but David Mir
01:10:11
fact checked in real time saying that
01:10:14
Trump was wrong and there was like this
01:10:17
effort to kind of Gaslight and make him
01:10:19
sound crazy during the debate when there
01:10:21
are in fact sources for what he was
01:10:23
saying and it might have thrown him off
01:10:24
a little bit I noticed like it was like
01:10:26
he I I agree they going into it I think
01:10:30
they need to negotiate in the future you
01:10:31
know how they're
01:10:33
negotiating the microphones on or off
01:10:36
audience on or off I think they should
01:10:38
negotiate are we factchecking in real
01:10:40
time or are we not factchecking and
01:10:42
who's doing that fact and they only fact
01:10:44
check one candidate for example when KLA
01:10:46
Harris repeated numerous hoaxes like the
01:10:48
very fine people hoax the bloodbath hoax
01:10:52
the suckers and losers hoax I mean these
01:10:54
are things that were already addressed
01:10:56
in the last debate and you know even
01:10:58
leftwing sites like Snopes have said the
01:11:00
whole veryify people thing is for people
01:11:03
who know that they there's been
01:11:04
selective edits and I mean there's been
01:11:06
selective edits forever but that one is
01:11:08
particularly egregious it's really
01:11:09
egregious the blood bath one is really
01:11:10
egregious too because because he was
01:11:12
talking about the blood
01:11:13
bath yeah just make it into a January
01:11:17
6th extension which it's not right so
01:11:20
she was able to say these things and
01:11:21
never got fact checked once which meant
01:11:23
she never got knocked out of her
01:11:25
and let's also be honest like Trump is
01:11:27
hyperbolic so if you are going to say
01:11:31
you know oh we're going to fact check
01:11:32
Trump like there's a lot of material
01:11:34
there and he's just he's a hyperbolic
01:11:37
guy that's kind of his shtick right I
01:11:39
mean but but here's the thing is that in
01:11:41
the wake of that debate look I I think a
01:11:43
lot of people scoring the debate on like
01:11:45
technical Debaters points would award
01:11:47
her the the the win for for that night I
01:11:50
don't clearly she won yeah I don't deny
01:11:52
that however what I think has been
01:11:54
surprising is that in the wake of the
01:11:57
debate you're seeing her support sort of
01:12:00
return more to its previous level and so
01:12:04
what I'm saying is the effect of that's
01:12:05
wearing off and I think one of the
01:12:06
reasons why that's wearing off is
01:12:09
because Trump still has the killer
01:12:10
issues in this election he's got the
01:12:12
border and he's got inflation and the
01:12:15
economy and Harris may have done well
01:12:18
again on debers points but what
01:12:20
substantive answer did she give in that
01:12:23
debate except to say I'm not Joe Biden
01:12:25
which is I guess true however what
01:12:28
you're basically saying is you won't
01:12:30
defend your own administration's record
01:12:32
you are the incumbent you're not the
01:12:33
change candidate and you're saying that
01:12:36
people should vote for you because
01:12:37
you're not Joe Biden well what is it
01:12:38
about Joe Biden's record that what is it
01:12:42
about Joe Biden's policies that you
01:12:43
don't agree with I mean after all you
01:12:45
cast the tiebreaking vote for the uh
01:12:48
inflation reduction act you cast it for
01:12:51
the 2 trillion American Rescue plan that
01:12:52
set off the inflation so the debate
01:12:54
Moder never asked Harris well what is it
01:12:57
about you that is different than Joe
01:12:59
Biden on a policy level other than the
01:13:02
fact I thought that was like a great
01:13:04
moment for her objectively I think you
01:13:08
and I've said this forever here on this
01:13:09
show uh putting our feelings aside about
01:13:11
the candidates I think whoever comes
01:13:14
across as the most normal or the most
01:13:15
moderate is going to win and I think
01:13:18
she's done a great job of like P
01:13:21
convincing those moderates that she's
01:13:24
not crazy and he is what what are your
01:13:26
thoughts on that because people looked
01:13:27
at this very podcast and they've said to
01:13:29
me my God that's the Trump I want to
01:13:31
vote for that Trump 2.0 the Allin Trump
01:13:34
and then people are like ah he's going
01:13:36
back to the insult comic Trump but I
01:13:38
don't want the chaos what are your
01:13:39
thoughts on moderates specifically in
01:13:42
the swing States and and this sort of
01:13:44
strategy let's talk about let's talk
01:13:46
about the teamsters so Biden when he was
01:13:48
still in the race was plus eight among
01:13:51
the teamsters Rank and file and now that
01:13:54
uh Harris is the the candidate Trump has
01:13:57
up something like plus 26 with the
01:13:59
teamsters yeah why is that cuz she's
01:14:02
isn't she pro-union as well he was Union
01:14:04
Joe so I mean it was like in the name I
01:14:06
understand why they loved him there's
01:14:08
something about her policies and I think
01:14:12
her the the look I think within the
01:14:14
Democratic party think it's her
01:14:16
personality I think I think it's partly
01:14:18
personality but I also think it's its
01:14:19
policies and cultural issues so within
01:14:21
the Democratic party there have always
01:14:23
been two tracks there's the beer track
01:14:25
and there's the wine track and so you
01:14:28
know Bill Clinton was classic beer track
01:14:30
guy right your summit with Obama right
01:14:34
and I think Joe Biden was was beer track
01:14:36
then there's kind of the wine track
01:14:37
which is the more it's the part of the
01:14:39
party that cares about these Boutique
01:14:41
cultural issues starting with Dei and
01:14:44
equity and uh trans and things like that
01:14:47
limousine liberals is what they used to
01:14:49
be called but I like yours wine liberals
01:14:51
or yeah the woke wine basically the
01:14:54
entire California Democratic party is
01:14:56
very wi track I mean Gavin is very wine
01:14:58
track Comm Harris is very wi track you
01:15:00
can understand why a blue collar worker
01:15:03
it doesn't appeal to that they want more
01:15:05
of that lunch paale traditional Democrat
01:15:08
but that Democratic party doesn't really
01:15:11
exists anymore I mean the Democratic
01:15:12
party has evolved to be the party of the
01:15:14
professional class whereas the
01:15:16
Republicans are more the party of the
01:15:18
working class and you're now starting to
01:15:20
see it I think Biden was the democrat's
01:15:23
last vestage of this working-class party
01:15:26
he really worked at being appealing to
01:15:27
those voters you know that whole
01:15:28
Scranton Joe image yeah and un Joe yeah
01:15:31
exactly whereas kamla when you get her
01:15:34
talking in an unguarded moment and is
01:15:37
not a canned answer she's going to talk
01:15:38
about diversity equity and inclusion and
01:15:41
that's not what your typical Teamster
01:15:42
wants to hear let me ask you challenging
01:15:44
question because one it's likes when I
01:15:46
ask you a challenge a bit if Trump
01:15:49
loses what do you think will be the
01:15:51
cause of the
01:15:53
loss if if he loses like strategically
01:15:56
when we look back on the last six months
01:15:58
what do you think you would change what
01:16:01
would cause it well look I mean the the
01:16:04
the the great asset that KLA Harris has
01:16:06
is not her likability it's not her track
01:16:09
record it's not her policies it's the
01:16:12
fact that she's got the media behind her
01:16:15
and if you look at like for example ABC
01:16:17
News 100% of the coverage by ABC News is
01:16:21
positive whereas something like 93% of
01:16:24
the of their coverage on Trump is
01:16:26
negative and you saw this that before
01:16:30
Harris replaced Biden's the nominee she
01:16:32
had very low favorability ratings and
01:16:35
then the media basically reinvented her
01:16:36
as this transformative candidate so look
01:16:38
when you've got the media willing to
01:16:41
operate as day facto members of your
01:16:43
campaign that's tremendously powerful if
01:16:46
we had a fair media this election
01:16:47
wouldn't be close so that is the
01:16:50
advantage the Democrats had now look
01:16:52
should Trump have done the debate with
01:16:55
ABC News no I think he should have
01:16:57
chosen more fair moderators I mean to
01:16:59
their credit I think CNN played the
01:17:02
Biden Trump debate pretty fair and down
01:17:04
the middle but ABC I mean it was
01:17:06
predictable that like I said I mean one
01:17:09
of the hosts was her sari sister they
01:17:11
friends so you know I I think that if
01:17:15
Trump loses you could say that his
01:17:17
willingness to walk into the liance den
01:17:19
take on all comers do every interview
01:17:22
you could say maybe that wasn't as
01:17:23
strategic as what she did but at the end
01:17:25
of the day I think that voters will
01:17:28
appreciate that both Trump and JD are
01:17:30
willing to do basically every podcast
01:17:32
every interview they're not afraid to
01:17:34
answer questions and when they do answer
01:17:37
questions you can see them thinking and
01:17:39
they don't give you the same canned
01:17:40
answer they've given 10 times before
01:17:42
including at the debate so yeah I mean
01:17:45
that's my take what what's yours JK on
01:17:48
which aspect be more specific give me a
01:17:49
give me a specific what do you think
01:17:51
what do you if if if she ends up winning
01:17:54
what do you think the reason will
01:17:56
be yeah that's a good that's a good
01:17:58
question if she ends up winning I think
01:18:02
it will be that people
01:18:04
believe that they I think it will be
01:18:07
that moderates in those swing States and
01:18:10
women believe that it's too much chaos
01:18:13
and that Trump will be too much they
01:18:16
want a calmer same thing reason Biden W
01:18:19
right like that there's this like
01:18:20
concept that the adults are in the room
01:18:22
and it will be calm and it won't be
01:18:23
chaotic and I think people just still
01:18:26
see Trump as a bit chaotic and I I think
01:18:28
that's the big fear and I think they've
01:18:30
played the abortion card and the right
01:18:33
to choose really well even though Trump
01:18:35
said it here I'm not going to sign the
01:18:36
abortion ban I'm Pro IVF I think they
01:18:38
have that really great win of saying hey
01:18:41
you bragged about overturning roie Wade
01:18:43
probably wasn't smart to brag about that
01:18:46
and they have that clip that they can
01:18:47
keep reinforcing so if he does lose and
01:18:49
I don't know that he's going to lose I
01:18:51
think there's a lot of people
01:18:54
who are going to go in there and vote
01:18:57
for him but not say it to pulsers and
01:19:00
not say it to their family and friends
01:19:02
because they're embarrassed because of
01:19:05
the pressure against orange Hitler or
01:19:07
you know this whole rhetoric that he's
01:19:09
going to you know overturn democracy so
01:19:14
I think it's a pretty good chance that
01:19:16
he's going to win actually I don't think
01:19:18
that this yeah I mean look I I think in
01:19:20
a close race right they say the
01:19:21
statistics in a close race favor him
01:19:24
yeah look I mean maybe we're asking the
01:19:26
wrong question here which is why would
01:19:27
he lose I mean I think maybe the real
01:19:29
question is why is he favored to win
01:19:30
because I think the polls including Nate
01:19:32
still ver still show him favor to win
01:19:34
and I think that when you look at what
01:19:36
the big issues are in this campaign and
01:19:39
what has people agitated and upset why
01:19:41
they think the country is on the wrong
01:19:43
track something like 65% it has to do
01:19:45
with the economy it has to do with
01:19:46
inflation it has to do with the Border I
01:19:48
think the on the cultural issues the
01:19:50
trans stuff drives parents crazy they
01:19:52
don't want the government telling them
01:19:53
what to do with their kids so it's hard
01:19:55
to think of a killer issue other than
01:19:58
maybe abortion that Harris has on her
01:20:01
side it feels like all the issues cut
01:20:03
Trump's way but the again the thing that
01:20:05
Trump doesn't have and there's no way to
01:20:07
for him to fix this is the media is just
01:20:10
so in the tank for for Harris now you
01:20:13
raise a good point look could Trump be
01:20:15
more disciplined yeah absolutely however
01:20:19
you know I think that what amplifies
01:20:21
that is the fact that the media is quick
01:20:23
to jump on every little thing he says
01:20:25
and distorts it and he sets himself up
01:20:27
for it you know like part of what makes
01:20:29
him activate the base is that erratic
01:20:33
Behavior his stick you know the comedy
01:20:35
and then I do believe that it gets
01:20:37
weaponized by the Press because it's
01:20:39
like such so easy for them I agree with
01:20:41
you that Trump could be more disciplined
01:20:43
however I don't think it's as bad as
01:20:45
what you're saying because if it were
01:20:47
there'd be no need to make up these
01:20:48
obvious hoaxes there'd be no need to you
01:20:51
know lie about the very fine people or
01:20:53
or what he said about blood bath so if
01:20:56
he was really saying that many
01:20:57
outrageous things why would you need to
01:20:59
keep inventing things that he didn't say
01:21:02
and if they're just stacking them yeah
01:21:04
the answer to that question is just
01:21:05
throw everything you got at him yeah
01:21:06
they're throwing everything at him but
01:21:07
look at look at kamla's interviews I
01:21:09
mean she hasn't given very many but I
01:21:12
mean her answers are just I mean just
01:21:14
watch them I'm not going to characterize
01:21:15
them but just just watch her actual said
01:21:18
it I mean Megan KY thinks she's stupid
01:21:20
and not bright I mean she's not the most
01:21:23
dynamic speaker that's for sure um and
01:21:26
she doesn't seem to be able
01:21:28
to uh have a dynamic debate with
01:21:31
intelligent people who are experts in
01:21:34
their field let's say you know she can't
01:21:36
hold her own in the way you can see JD
01:21:39
can right and and Trump Canen uh so here
01:21:42
we go and just on the um on the second
01:21:45
assassination attempt I don't know if
01:21:46
you even want to go there but I mean
01:21:48
gosh I'm so glad that he didn't get shot
01:21:52
at again this is scary stuff folks uh
01:21:55
this rhetoric's got to come down I keep
01:21:56
saying it nobody wants to listen to me
01:22:00
but man be well let's look at the
01:22:02
rhetoric that Ryan Ruth was literally
01:22:05
quoting on his Twitter was saying that
01:22:07
Trump is basically existential threat to
01:22:09
democracy he was quoting what Joe Biden
01:22:13
and KLA Harris and the mainstream media
01:22:15
have been saying chapter and verse uh so
01:22:17
I think that you know if you want to
01:22:20
ascribe motivation there where did Ruth
01:22:24
get get these ideas they've been
01:22:26
endlessly Amplified by the mainstream
01:22:27
media and it's not like a one-off
01:22:29
comment it's been the central narrative
01:22:31
for the last several years is that
01:22:32
somehow Trump represents this
01:22:34
existential threat to democracy and one
01:22:36
way or another that threat must be
01:22:38
eliminated and I think Ryan Ruth simply
01:22:41
took literally what the mainstream media
01:22:43
has been saying 1% of your followers is
01:22:46
what I tell everybody high-profile
01:22:47
people you and I both know is 1% of
01:22:51
people in your following and we all have
01:22:53
large followings here and and there
01:22:56
certainly people who have extremely
01:22:57
large followings 1% are mentally ill
01:22:59
like when I say mentally ill I mean
01:23:01
severely mentally ill and if it's but 1%
01:23:04
of your following if it's 0.1% this
01:23:06
could be thousands of people and this is
01:23:08
what happened to John lennin and and and
01:23:10
other famous people who've been killed
01:23:12
tragically is those mentally ill people
01:23:15
interpret things in a very different way
01:23:17
and when you say you know a phrase that
01:23:21
has triggers in it threat to democracy
01:23:23
fight like how whatever it is they
01:23:25
interpret it differently and so just
01:23:27
please folks when you call the guy
01:23:28
Hitler for years and again you
01:23:32
create millions or billions of
01:23:34
Impressions around that and it's not
01:23:35
like a one-off statement but it's
01:23:37
something that's drummed into the public
01:23:40
over and over again it seems to me
01:23:41
you're asking for trouble stay safe
01:23:44
please turnone down the rhetoric
01:23:45
everybody and we will see you next time
01:23:47
on the all in podcast
01:23:50
byebye let your winners
01:23:53
ride Rain Man
01:23:57
David and instead we open source it to
01:24:00
the fans and they've just gone crazy
01:24:02
with
01:24:02
[Music]
01:24:10
it besties
01:24:13
are that's my dog taking
01:24:16
[Music]
01:24:18
driveway oh
01:24:21
Manet we should all just get a room and
01:24:23
just have one big huge cuz they're all
01:24:25
this useless it's like this like sexual
01:24:26
tension that they just need to release
01:24:34
someh we need to get mer
01:24:38
[Music]
01:24:44
our going
01:24:46
[Music]
