00:00:00
all right everybody uh let's get the
00:00:01
show started here wait wait Jason why
00:00:03
you wearing a tux what's going on there
00:00:05
oh well it's time for a very emotional
00:00:08
segment we do hear on the all-in
00:00:11
podcast I just got to get myself
00:00:13
composed for this H Jason are you okay
00:00:17
I'm I'm gonna be okay I think it looks
00:00:19
like you're fighting back at tier what's
00:00:22
yeah this is always a tough one this
00:00:23
year we tragically lost Giants in our
00:00:28
industry these individuals bravely own
00:00:30
their craft at open AI before
00:00:34
departing Ilia sasver he left in he left
00:00:37
us in
00:00:38
[Music]
00:00:39
May Yan Leica also left in
00:00:44
May Sean shman tragically left us in
00:00:49
August wait these are AI employees yes s
00:00:54
left on Wednesday Bob mcru also left on
00:00:58
Wednesday Too Short too short and Mera
00:01:03
moradi also left us tragically on
00:01:05
Wednesday we lost miror too yeah and
00:01:08
Greg Brockman is on extended leave the
00:01:12
enforcer he left too thank you for your
00:01:15
service your memories will live on as
00:01:19
training data and may your memories be a
00:01:24
vesting let your winners
00:01:27
ride Rainman David
00:01:32
and instead we open source it to the
00:01:34
fans and they've just gone crazy with it
00:01:39
[Music]
00:01:41
Queen my goodness all those losses wow
00:01:45
that is three in one day three in one
00:01:47
day my goodness I thought open AI was
00:01:49
nothing without its
00:01:51
[Laughter]
00:01:53
people I well I mean this is a this is a
00:01:56
great whoa we lost something come back
00:01:59
wait oh wait
00:02:02
what this is like the photo it's like
00:02:05
the photo and Back to the Future wow
00:02:07
they're just all gone wait oh no don't
00:02:08
worry he's replacing everybody here we
00:02:10
go he replacing with a g700 a Bugatti
00:02:13
and I guess Sam's got mountains of cash
00:02:15
so don't worry he's got a backup plan
00:02:17
jamas anyway as an industry and as
00:02:19
leaders in the industry the show sends
00:02:21
its regards to Sam and the open AI team
00:02:24
on their tragic losses and
00:02:26
congratulations on the
00:02:27
$150 billion doll value and you're 7%
00:02:31
Sam now just cashed in10 billion
00:02:34
apparently so congratulations to fan of
00:02:36
uh friend of the Pod Sam amen is the
00:02:39
round that's all reported out of some
00:02:41
article right that's not like confirmed
00:02:44
or anything is all of that done I mean
00:02:46
it's reportedly allegedly that he's
00:02:47
going to have 7% of the company and we
00:02:49
can jump right into our first story I
00:02:52
mean what I'm saying is has the money
00:02:53
been wired and the docks been sign
00:02:55
according to reports this round is
00:02:57
contingent on not being
00:03:00
a um nonprofit anymore and sorting that
00:03:03
all out they have to remove the profit
00:03:05
cap and do the seor there's some article
00:03:07
that reported this right none of us
00:03:09
Bloomberg it's not some article it was
00:03:10
Bloomberg and and it got a lot of
00:03:13
traction and it was re-reported by a lot
00:03:15
of places and I don't see anyone
00:03:16
disputing it so is mainstream media we
00:03:20
trust the mainstream media in this case
00:03:23
because think I think that when we could
00:03:25
do a good bit yeah that's mine no I
00:03:27
think that Bloomberg Bloomberg reported
00:03:29
it based on obviously talks that are
00:03:31
ongoing with investors who have
00:03:32
committed to this round yeah and no
00:03:34
one's disputing it has anyone said it's
00:03:36
not true this has been speculated for
00:03:40
months the $150 billion valuation
00:03:44
raising something in the range of six
00:03:46
to7 billion if you do the math on that
00:03:49
and Bloomberg is correct that Sam mman
00:03:52
got his 7% I guess the reality is you
00:03:55
can't raise $6 billion dollar without
00:03:57
probably meeting with a few dozen firms
00:03:59
and some number of Junior people in
00:04:01
those few dozen firms are having a
00:04:03
conversation or two with reporters so
00:04:06
you can kind of see how it gets out all
00:04:08
right and before we get to our uh first
00:04:10
story there about open AI
00:04:12
congratulations to chamath let's pull up
00:04:14
the photo here he was a featured guest
00:04:18
on uh The Alex Jones Show no sorry I'm
00:04:21
sorry uh that would be Joe Rogan
00:04:24
congratulations on coming to Austin and
00:04:26
being on Joe Rogan what was it like to
00:04:27
do a three-hour podcast
00:04:30
with Jo Rogan it's great I mean I loved
00:04:34
it he's really awesome he's super cool
00:04:38
it's good to do long form stuff like
00:04:39
this so that I can actually talk clearly
00:04:41
is the limitation of this podcast is the
00:04:43
other three of us finally you have found
00:04:45
a way to make it about yourself um I saw
00:04:49
a comment somebody commented like oh wow
00:04:51
it's like amazing to hear chth expand on
00:04:53
topics about the constant eruptions by
00:04:56
Jal also known as moderation
00:05:00
someone called me someone called me Fran
00:05:03
or Fran from the 70s Show I thought that
00:05:06
was funny the the amount of trash
00:05:09
talking in Rogan's YouTube comments it's
00:05:11
next level it is it is it is the wild
00:05:15
wildd west uh in in terms of the comment
00:05:18
section on YouTube yeah um a bunch of
00:05:21
comments asking Jak why do you call it
00:05:22
Alex Jones is that cuz it's just a Texas
00:05:25
short podcaster who short and stout and
00:05:27
they look similar just a but I mean it
00:05:31
looks like Alex
00:05:32
Jones uh sort of lifting weights
00:05:34
actually no they're
00:05:35
both the same height and yeah both saw
00:05:38
joean 25 years ago doing standup I have
00:05:40
a photo with him at the club it was like
00:05:42
a small Club in San Francisco and we
00:05:45
hung out with him afterwards he was just
00:05:47
like a nobody back in the day was like a
00:05:48
stand-up guy right now
00:05:50
he's uberstar well you have to go back
00:05:53
pretty far for Joe Rogan to be a nobody
00:05:55
I mean he had a TV show for a long time
00:05:58
and two of them in fact
00:06:00
comic for a while stand up comic yeah
00:06:03
Fear Factor that's right that's also do
00:06:06
Survivor or one of those like so I think
00:06:08
and then the UFC I mean this guy's got
00:06:11
four four blew up UFC yeah yeah well I
00:06:15
mean I think he got the UFC out of Fear
00:06:17
Factor and being a UFC fighter and a
00:06:19
comedian and there's like a famous story
00:06:21
where like Dana White was pursuing him
00:06:25
and he was like I don't know and then
00:06:27
Dana White's like I'll send a plane for
00:06:28
you you can bring your friend he's like
00:06:30
okay fine I'll do it he did it for free
00:06:32
and then Dana White pursued him heavily
00:06:33
to become the voice of the UFC and yeah
00:06:36
obviously it's grown tremendously and
00:06:39
it's you know worth billions of dollars
00:06:42
okay so how is open AI worth 150 billion
00:06:45
dollars can anyone app well why don't we
00:06:49
get into the topic should we make the
00:06:50
bull case and the bear case all right
00:06:51
open AI as we were just joking in the
00:06:54
opening segment is trying to convert
00:06:56
into a for profit benefit Corporation
00:07:00
that's a BC Corp it just means we'll
00:07:01
explain bcorp later Sam wman is
00:07:04
reportedly I thought they're converting
00:07:05
to a C Corp no it's the same thing B
00:07:07
Corp doesn't benefit Corporation is a cc
00:07:10
Corporation variant that is not a
00:07:13
nonprofit but the board of director
00:07:16
Sachs is required not only to be a
00:07:19
fiduciary for all shareholders but also
00:07:22
for the stated mission of the company
00:07:24
that's my understanding of a b Corp am I
00:07:25
right freberg external stakeholders yeah
00:07:28
so like the environment or Society or
00:07:31
whatever but it from all other kind of
00:07:33
legal tax factors it's the same as a CP
00:07:37
and it's a way to I guess signal to
00:07:40
investors the market employees that you
00:07:43
care about something more than just
00:07:45
profit so famous most famous borp I
00:07:48
think is Toms is that the shoe company
00:07:50
Toms that's a famous B Corp somebody
00:07:52
will look it up here pagonia yeah that
00:07:54
falls into that category so for profit
00:07:56
with a mission Reuters has cited
00:07:58
anonymous sources close to the company
00:08:02
that the plan is still being hashed out
00:08:04
with lawyers and shareholders and the
00:08:06
timeline isn't certain but what's being
00:08:08
discussed is that the uh nonprofit will
00:08:11
continue to exist as a minority
00:08:13
shareholder in the new
00:08:15
company how much of a minority
00:08:17
shareholder I guess is uh the devil's in
00:08:19
the detail there do they own 1% or 49%
00:08:22
the uh very much discussed freedberg
00:08:25
100x profit cap for investors will be
00:08:28
removed
00:08:29
that means investors like venod friend
00:08:31
of the Pod and Reed Hoffman also friend
00:08:33
of the Pod could see a 100x turn into
00:08:37
a,x or more according to the Bloomberg
00:08:39
report Sam waltman's going to get his
00:08:41
Equity finally 7% that would put him at
00:08:44
around 10.5 billion if this is all true
00:08:47
and openi could be valued as high as
00:08:50
$150 billion we'll get into all the
00:08:54
shenanigans but let's start with your
00:08:55
question freeberg and since you asked it
00:08:57
I'm going to Boomerang it back to you
00:08:59
make the bull case for $150 $150 billion
00:09:05
valuation the bull case would be that
00:09:08
the moat in the business with respect to
00:09:11
model
00:09:12
performance and infrastructure gets
00:09:15
extended with the large amount of
00:09:17
capital that they're raising they
00:09:18
aggressively deploy it they are very
00:09:20
strategic and tactical with respect to
00:09:23
how they deploy that
00:09:24
infrastructure to continue to improve
00:09:26
model performance and as a result
00:09:28
continue to extend their advantage in
00:09:31
both consumer and Enterprise
00:09:32
applications the API tools and so on
00:09:34
that they offer and so they can maintain
00:09:38
uh both kind of model and application
00:09:40
performance leads that they have today
00:09:43
across the board I would say like the
00:09:45
the 01 model uh their voice application
00:09:48
Sora has not been released publicly but
00:09:50
if it does and it looks like what it's
00:09:51
been demoed to be it's certainly ahead
00:09:53
of the pack so there's a lot of uh
00:09:56
aspects
00:09:57
of of open AI today that kind of makes
00:10:00
them a leader and if they can deploy
00:10:01
infrastructure to maintain that lead and
00:10:02
not let Google Microsoft Amazon and
00:10:05
others catch up then their ability to
00:10:08
use that Capital wisely keeps them ahead
00:10:10
and ultimately as we all know there's a
00:10:12
multi-trillion dollar market to capture
00:10:13
here making lots of verticals lots of
00:10:16
applications lots of products so they
00:10:18
could become a a true kind of global
00:10:20
player here plus the extension into
00:10:22
Computing which I'm excited to talk
00:10:23
about later when we get into this
00:10:25
Computing stuff Sachs here's a chart of
00:10:27
open A's Revenue growth
00:10:30
that has been piec meal together from
00:10:32
various sources at various times but
00:10:35
you'll see here they are
00:10:38
reportedly as of June of 2024 on a $3.4
00:10:42
billion run rate for this year after
00:10:45
hitting two billion in 23 1.3 billion in
00:10:49
October of 23 and then back in 2022 was
00:10:54
reported they only had 28 million in
00:10:56
Revenue so this a pretty pretty hot a
00:10:59
pretty a pretty big shriek here in terms
00:11:00
of Revenue growth I would put it at 50
00:11:02
times Topline Revenue 150 billion
00:11:05
valuation you want to give us the bease
00:11:07
maybe or the bull case well so the The
00:11:09
Whisper numbers I heard was that their
00:11:12
revenue run rate for this year was in
00:11:13
the four to six billion range which is a
00:11:15
little higher than that so you're right
00:11:19
if it's really more like 3.4 this
00:11:21
valuation is about 50 times current
00:11:23
Revenue but if it's more like 5 billion
00:11:26
then it's only 30 times and if it's
00:11:28
growing % year-over-year it's only 15
00:11:31
times next year so depending what the
00:11:33
numbers actually are the $150 billion
00:11:35
valuation could be warranted I don't
00:11:38
think 15
00:11:39
times Ford ARR is a high valuation for a
00:11:44
company that has this kind of strategic
00:11:46
opportunity I think it all comes down to
00:11:49
the
00:11:50
durability of its comparative advantage
00:11:53
here I think there's no question that
00:11:54
open AI is the leader of the pack it has
00:11:57
the most advanced AI model models it's
00:11:59
got the best developer ecosystem the
00:12:02
best apis it keeps rolling out new
00:12:04
products and the question is just how
00:12:06
durable that Advantage is is there
00:12:09
really a moe to any of this for example
00:12:12
meta just announced llama 3.2 which can
00:12:15
do voice and this is roughly at the same
00:12:17
time that open aai just released its
00:12:20
voice API so the open source ecosystem
00:12:23
is kind of hot on open ai's heels the
00:12:27
large companies Google Microsoft so
00:12:30
forth they're hot on their heels too
00:12:33
although it seems like they're further
00:12:34
behind where meta is and the question is
00:12:38
just can open AI maintain its lead can
00:12:40
it consolidate its lead can it develop
00:12:41
some Moes if
00:12:43
so it's on track to be the next trillion
00:12:46
dollar big tech company but if not it
00:12:48
could be eroded and you could see the
00:12:51
value of open AI get get commoditized
00:12:53
and we'll look back on it as kind of a
00:12:55
cautionary tail okay chth do us a favor
00:12:57
here if there is a be case what is it
00:13:00
okay let's steeland the be case yes
00:13:03
that's what I'm asking
00:13:04
please so one would just be on
00:13:07
the fundamental technology itself and I
00:13:12
think the version of that story would go
00:13:14
that the
00:13:17
underlying Frameworks that people are
00:13:19
using to make these models great is well
00:13:22
described and available in open source
00:13:25
on top of that there are at least two
00:13:28
viable open source models that are as
00:13:32
good or better at any point in time than
00:13:36
open AI so what that would mean is that
00:13:39
the value of those
00:13:41
models the economic value basically goes
00:13:44
to zero and it's a consumer
00:13:46
surplus for the people that use it so
00:13:49
that's very hard theoretically to
00:13:53
monetize I think the second part of the
00:13:57
case would be that specific Al meta
00:14:00
becomes much more aggressive in
00:14:03
inserting meta AI into all of the
00:14:06
critical apps that they control because
00:14:10
those apps really are the front door to
00:14:12
billions of people on a daily basis so
00:14:15
that would mean WhatsApp Instagram
00:14:18
messenger the Facebook app and threads
00:14:22
gets refactored in a way where instead
00:14:24
of leaving that application to go to a
00:14:26
chat GPT like app you would just stay in
00:14:29
the
00:14:31
app and then the companion to that would
00:14:37
be that Google also does the same thing
00:14:40
with their version in front of search so
00:14:42
those two big front doors to the
00:14:44
internet become much more aggressive in
00:14:47
giving you a reason to not have to go to
00:14:49
chat gbt because a their answers are
00:14:52
just as good and B they're right there
00:14:54
in a few less clicks for you yeah I'm so
00:14:57
that would be the that would be the
00:14:58
second
00:14:59
piece the third piece is that all of
00:15:03
these models basically run out of viable
00:15:07
data to differentiate themselves and it
00:15:09
basically becomes a race around
00:15:11
synthetic information and synthetic data
00:15:14
which is a cost problem meaning if
00:15:16
you're going to invent synthetic data
00:15:18
you're going to have to spend money to
00:15:20
do it and the large companies Facebook
00:15:23
Microsoft Amazon Google Apple have
00:15:26
effectively infinite money compared to
00:15:27
any startup
00:15:30
and then the fourth which is the most
00:15:32
quizzical one is what does the human
00:15:35
capital thing tell you about what's
00:15:38
going on it reads a little bit like a
00:15:40
tel Nolla I have not in my time in
00:15:43
Silicon Valley ever seen a company
00:15:46
that's supposedly on such a straight
00:15:48
line to a rocket ship have so much
00:15:51
highle
00:15:53
churn and I've but and I've also never
00:15:55
seen a company have this much liquidity
00:15:59
and so how are people deciding to leave
00:16:03
if they think it's going to be a
00:16:04
trillion dollar company and why when
00:16:07
things are just starting to cook would
00:16:08
you leave if you are technically
00:16:11
enamored with what you're building so if
00:16:14
you had to construct the barcase I think
00:16:16
those would be the four things open
00:16:18
source front door
00:16:20
competition the move to synthetic data
00:16:23
and all of the executive turnover would
00:16:26
be sort of why you would say May maybe
00:16:29
there's a fire where there's all this
00:16:31
smoke okay I think this is very well put
00:16:33
and I have been
00:16:36
using chat GPT and Claude and Gemini
00:16:40
exclusively I stopped using Google
00:16:42
search and I also stopped Sachs asking
00:16:46
people on my team to do stuff before I
00:16:48
asked chat GPT to do it specifically
00:16:51
freeberg the 01 version and the 01
00:16:54
version is distinctly different have you
00:16:56
gentlemen been using 01 like on a daily
00:17:00
okay so we can have a really interesting
00:17:01
conversation here I did something on my
00:17:04
other podcast this week in startups that
00:17:05
I'll show you right now that was crazy
00:17:08
yesterday is a g is a game changer it's
00:17:10
the first real it's the first real Chain
00:17:14
of Thought production system yes that I
00:17:16
think we've seen are you using o1
00:17:18
preview or ow mini uh I am using o1
00:17:22
preview now let me show you what I did
00:17:24
here just just so the audience can level
00:17:26
set here if you're not watching us go to
00:17:27
YouTube and type in in and and you can
00:17:30
you can watch us we do video here so I
00:17:33
was analyzing you know just some early
00:17:36
stage uh deals and cap tables and I put
00:17:38
in here hey a startup just raised some
00:17:40
money at this valuation here's what the
00:17:42
friends and family invested the
00:17:44
accelerator the seed investor Etc in
00:17:46
other words like the history the
00:17:47
investment history in a company what 01
00:17:50
does distinctly differently than the
00:17:52
previous versions and the previous
00:17:54
version I felt was 3 to 6 months ahead
00:17:56
of competitors this is a year ahead of
00:17:58
competitor s and so here shth if you
00:18:01
look it said it thought for 77 seconds
00:18:04
and if you click the down arrow saxs
00:18:07
what you'll see is it gives you an idea
00:18:09
of what its rationale is for
00:18:12
interpreting and what secondary query is
00:18:14
it's
00:18:15
doing in order this is called Chain of
00:18:18
Thought and this is the underlying Mega
00:18:21
model that sits on top of the llms and
00:18:25
the mega model effectively The Chain of
00:18:27
Thought approach is
00:18:29
the model asks itself the question how
00:18:33
should I answer this question right and
00:18:35
then it comes up with an answer and then
00:18:37
it says now based on that what are the
00:18:39
steps I should take to answer the
00:18:40
question so the model keeps asking
00:18:42
itself questions related to the
00:18:44
structure of the question that you asked
00:18:48
and then it comes up with a series of
00:18:49
steps that it can then call the llm to
00:18:52
do to fill in the blanks link them all
00:18:55
together and come up with the answer
00:18:56
it's the same way that a human train of
00:18:58
thought works and it really is the the
00:19:00
kind of ultimate evolution of what a lot
00:19:03
of people have said these systems need
00:19:04
to become which is a much more call it
00:19:07
intuitive approach to answering
00:19:09
questions rather than just predictive
00:19:11
text based on the single statement you
00:19:13
made and it really has changing the game
00:19:15
and everyone is going to chase this and
00:19:17
follow this it is the new paradigm for
00:19:20
for how these AI kind of systems will
00:19:22
work and by the way what this did was
00:19:25
what prompt Engineers were doing or
00:19:27
prompt engineering websites we're doing
00:19:29
which was trying to help you construct
00:19:30
your question and so if you look to this
00:19:33
one it says listing disparities I'll
00:19:35
compile a cap table with Investments and
00:19:37
valuations building the cap table
00:19:39
accessing the share valuation breaking
00:19:40
down ownership breaking down ownership
00:19:42
Etc evaluating the terms and then it
00:19:44
checks its work a bit it waits
00:19:45
investment options and you can see this
00:19:47
is a this is fired off like two dozen
00:19:51
different queries to as fre correctly
00:19:55
pointed out you know build this chain
00:19:58
and
00:19:59
it got incredible answers explain the
00:20:01
formulas so it's thinking about what
00:20:03
your next question would be and this
00:20:06
when I share this with my team it was
00:20:08
like a super game Cher uh Sach you had
00:20:11
some thoughts here well yeah I mean this
00:20:14
is pretty impressive and just to build
00:20:16
on what freeberg was saying about Chain
00:20:18
of Thought where this All Leads is to
00:20:20
agents where you can actually tell the
00:20:22
AI to do work for you you give it an
00:20:26
objective it can break the objective
00:20:28
down into tasks and then it can work
00:20:30
each of those tasks and open AI at a
00:20:34
recent meeting with investors said that
00:20:36
PhD level reasoning was next on its road
00:20:39
map and then agents weren't far behind
00:20:41
that they've now released the at least
00:20:43
the preview of the PHD level reasoning
00:20:45
with this 01 model so I think we can
00:20:47
expect an announcement pretty soon about
00:20:49
agents yeah and so the and if you think
00:20:52
about if you think about business value
00:20:54
you know we think a lot about this is
00:20:55
like where's the SAS opportunity and all
00:20:57
this the soft off as a service
00:20:59
opportunity it's going to be an agents I
00:21:02
think we'll ultimately look back on
00:21:04
these sort of chat models as a little
00:21:06
bit of a parlor trick compared to what
00:21:08
agents are going to do in the workplace
00:21:11
if you've ever been to a call center or
00:21:13
an operations center they're also called
00:21:15
service factories it's assembly lines of
00:21:18
people doing very complicated knowledge
00:21:21
work but ultimately you can unravel
00:21:23
exactly what the chain is there The
00:21:26
Chain of Thought that goes into their
00:21:28
decisions it's it's very complicated and
00:21:31
that's why you have to have humans doing
00:21:32
it but you could imagine that once
00:21:37
system integrators or or Enterprise SAS
00:21:39
apps go into these places go into these
00:21:42
companies they integrate the data and
00:21:44
then they map out the workflow you could
00:21:47
replace a lot of these steps in the
00:21:49
workflow with agents yeah by the way
00:21:51
it's not just it's not just call centers
00:21:53
I had a conversation with um I'm on the
00:21:56
board of a company with the CEO the
00:21:57
other day and he was like well we're
00:21:59
going to hire an analyst that's going to
00:22:01
sit between our kind of retail sales
00:22:03
operations and do the and you know
00:22:05
figure out what's working to drive
00:22:07
marketing decisions and I'm like no
00:22:08
you're not like I I really think that
00:22:11
that would be a mistake because today
00:22:13
you can use 01 and describe or just feed
00:22:17
it the data and describe the analysis
00:22:19
you want to get out of that data and
00:22:21
within a few minutes and I've now done
00:22:23
this probably a dozen times in the last
00:22:24
week with different projects internally
00:22:26
at my company it gives you the enre
00:22:28
answer that an analyst would have taken
00:22:30
days to put together for you and if you
00:22:32
think about what an analyst's job has
00:22:34
been historically is they take data and
00:22:36
then they manipulate it and the big
00:22:38
evolution in software over the last
00:22:40
decade and a half has been tools that
00:22:42
give that analyst leverage to do that
00:22:45
data manipulation more quickly like
00:22:47
Tableau and you know R and all sorts of
00:22:49
different toolkits that are out there
00:22:51
but now you don't even need the analyst
00:22:53
because the analyst is the Chain of
00:22:54
Thought it's the prompting from the
00:22:57
model and it's completely going to
00:22:59
change how knowledge work is done
00:23:00
everyone that owns a function no longer
00:23:03
needs an analyst the analyst is the
00:23:05
model that's sitting on the computer in
00:23:07
front of you right now and you tell it
00:23:08
what you want and not days later but
00:23:10
minutes later you get your answer it's
00:23:13
completely revolutionary in um uh ad hoc
00:23:18
uh knowledge work as well as kind of
00:23:20
this repetitive structured knowledge
00:23:21
this is such a good point freeberg the
00:23:23
ad hoc piece of it when we're processing
00:23:26
20,000 applications for funding a year
00:23:28
we do a 100 plus meetings a week the
00:23:30
analysts on our team are now putting in
00:23:33
the transcripts um and key questions
00:23:36
about markets and they are getting so
00:23:38
smart so fast that you know when
00:23:41
somebody comes to them with a
00:23:42
Marketplace in diamonds their
00:23:45
understanding of the diamond Marketplace
00:23:47
becomes so rich so fast that we can
00:23:49
evaluate companies faster then we're
00:23:52
also seeing uh chth uh before we call
00:23:55
our lawyers when we have a legal
00:23:57
question about a document we start
00:23:58
putting in you know let's say the the
00:24:00
standard note template or the standard
00:24:02
safe template we put in the new one and
00:24:05
there's a really cool project by Google
00:24:07
called notebook L LM where you can put
00:24:10
in multiple documents and you can start
00:24:11
asking questions so imagine you take
00:24:13
every single legal documents acts that
00:24:16
Yammer had when you had chamath as an
00:24:18
investor I'm not sure if he was on the
00:24:19
board and you can start asking questions
00:24:21
about the documents and we have had
00:24:24
people make changes to these documents
00:24:25
and it immediately finds them explains
00:24:27
them and so everybody's just getting so
00:24:30
goddamn smart so fast using these tools
00:24:33
that I insisted that every person on the
00:24:35
team when they hit control tab it opens
00:24:37
a chat gp4 window in 01 and we burned
00:24:40
out our credits immediately like it
00:24:43
stopped us it said you're you're you
00:24:45
have to stop using it for the rest of
00:24:46
the month Cham your thoughts on this
00:24:48
we're seeing it in real
00:24:50
time in 890 what I'll tell you is what
00:24:54
sck said is totally right there's so
00:24:58
many
00:24:59
companies that have very complicated
00:25:02
processes that are a combination of
00:25:05
well- Tred and well-meaning people
00:25:08
and bad software and what I mean by bad
00:25:13
software is that some other third party
00:25:16
came in listened to what your business
00:25:19
process was and then wrote this clunky
00:25:20
deterministic code usually on top of
00:25:23
some system of
00:25:24
record charged you tens or hundreds of
00:25:27
millions of for it and then left and
00:25:30
will support it only if you keep paying
00:25:31
them millions of dollars a year that
00:25:34
whole thing is so nuts because the
00:25:35
ability for people to do work I think
00:25:37
has been very much
00:25:39
constrained and it's constrained by
00:25:41
people trying to do the right thing
00:25:42
using really really terrible software
00:25:45
and all of that will go away the radical
00:25:48
idea that I would put out there is I
00:25:50
think that systems of record no longer
00:25:52
exist because they don't need
00:25:54
to and the reason is because all you
00:25:57
have is data and you have a and just
00:26:00
explain to people what system of record
00:26:01
is just so so inside of a company you'll
00:26:04
have a handful of systems that people
00:26:06
would say are the single source of truth
00:26:08
they're the things
00:26:09
that are used for reporting compliance
00:26:13
and example would be for your general
00:26:17
ledger so toet record your revenues You'
00:26:21
use Nets or you'd use Oracle GL or you'd
00:26:24
use workday financials then you'd have a
00:26:27
different system of record for all of
00:26:29
your Revenue generating activities so
00:26:32
who are all of the people you sell to
00:26:34
how are sales going what is the
00:26:36
PIP there's companies like Salesforce or
00:26:39
Sugar CRM then there's a system of
00:26:42
record for all the employees that work
00:26:44
for you all the benefits they have what
00:26:46
is their salary this is hris so the
00:26:48
point is that the software economy over
00:26:52
the last 20 years and this is trillions
00:26:54
of dollars of market cap and hundreds of
00:26:57
billions of Revenue you have been built
00:26:59
on this premise that we will create this
00:27:03
system of record you will build apps on
00:27:06
top of the system of record and the
00:27:07
knowledge workers will come in and
00:27:10
that's how they will get work done and I
00:27:12
think that saaks is right this totally
00:27:14
flips that on its head instead what will
00:27:16
happen is people will provision an agent
00:27:20
and roughly direct what they want the
00:27:22
outcome to be and they'll be process
00:27:24
independent they won't care how they do
00:27:26
it they just want the answer so I think
00:27:28
two things happen the obvious thing that
00:27:31
happens in that world is systems of
00:27:34
record lose a
00:27:36
grip on the Vault that they had in terms
00:27:40
of the data that runs a company you
00:27:42
don't necessarily need it with in the
00:27:45
same Reliance and Primacy that you did
00:27:47
five and 10 years ago that'll have an
00:27:49
impact to the software
00:27:51
economy and the second thing that I
00:27:54
think is even more important than that
00:27:57
is that then the omic size of companies
00:28:00
changes because each company will get
00:28:03
much more leverage from using software
00:28:05
and few people versus lots of people
00:28:08
with a few pieces of software and so
00:28:09
that inversion I think creates
00:28:12
tremendous potential for operating
00:28:14
leverage all right your thought Sachs
00:28:15
you operate in the SAS space uh with
00:28:17
system of records and investing in these
00:28:19
type of companies give us your take well
00:28:22
it's interesting we were having a
00:28:23
version of this conversation last week
00:28:26
on the Pod and I started getting texts
00:28:28
from Ben off as he was listening to it
00:28:30
and then oh he he called me and I think
00:28:32
he got a little bit triggered by the
00:28:34
idea that systems of record like
00:28:37
Salesforce are going to be obsolete in
00:28:38
this new AI era and he made a very
00:28:41
compelling case to me about why that
00:28:43
wouldn't happen which is yeah well first
00:28:45
of all I think AI models are predictive
00:28:48
I mean at the end of the day they're
00:28:49
predicting the next set of texts and so
00:28:51
forth and when it comes to like your
00:28:53
employee list or your customer list you
00:28:55
just want to have a source of Truth you
00:28:58
want it to be 98% accurate you just want
00:29:00
it to be 100% accurate you want to know
00:29:02
if the federal government asks you for
00:29:04
the tax ID numbers of your employees you
00:29:06
just won't be able to give it to them if
00:29:07
Wall Street analysts asks you for your
00:29:10
customer list and what the Gap revenue
00:29:11
is you just won't be able to provide
00:29:13
that you don't want AI models figuring
00:29:15
it out so you're still going to need a
00:29:16
system of record
00:29:19
furthermore he made the point that you
00:29:21
still need databases you still need
00:29:23
Enterprise security if you're dealing
00:29:24
with Enterprises you still need
00:29:26
compliance you still need sharing models
00:29:28
there's all these aspect all these
00:29:30
things that have been built on top of
00:29:31
the database that SAS company's been
00:29:33
doing for 25 years and then the final
00:29:35
point that I think is compelling is that
00:29:38
Enterprise customers don't want to DIY
00:29:41
it right they don't want to have to
00:29:42
figure out how to put this together and
00:29:44
you can't just hand them an llm and say
00:29:47
here you go there's a lot of work that
00:29:51
is needed in order to make these models
00:29:54
productive and so at a minimum you're
00:29:56
going to need system integrators and
00:29:59
Consultants to come in there connect
00:30:01
hold on just connect all the Enterprise
00:30:02
data to these models map the workflows
00:30:05
you have to do that now how's that
00:30:07
different from how this this clunky
00:30:09
software is sold today I mean look I I
00:30:12
don't want to take away from the quality
00:30:15
of the company that Mark has built and
00:30:17
what he's done for the cloud economy so
00:30:19
let's just put that aside but I wish
00:30:21
this is what we could have actually all
00:30:22
been on stage and talked about I told
00:30:25
him that when he was at the summit I
00:30:26
said that because I agree with basically
00:30:29
every premise of those three things
00:30:31
number one systems integrators exist
00:30:32
today to build apps on top of these
00:30:34
things why do you think you have
00:30:36
companies like Viva how can a 20 billion
00:30:38
dollar plus company get built on top of
00:30:40
Salesforce it's because it doesn't do
00:30:42
what it's meant to do that's
00:30:47
why app stores are a great way to allow
00:30:50
people to build on your platform and and
00:30:52
and cover those Niche cases the point
00:30:54
I'm trying to make is that's no
00:30:55
different than the economy that exists
00:30:57
today is just going to transform to
00:30:58
different groups of people number one
00:30:59
well by the way he said he's willing to
00:31:01
come on the PO and talk about this very
00:31:02
issue but just with
00:31:05
you talk all us now great he'll come on
00:31:09
the Pod and discuss whether AI makes
00:31:10
sass obsolete a lot of people are asking
00:31:12
that question let's talk about it next
00:31:14
year at the summit can you talk about
00:31:16
his uh philanthropy first okay let's get
00:31:18
back to focus here let's get focused
00:31:21
everybody love you mark who's coming to
00:31:24
dream for us raise your hand I want to
00:31:26
make another point the second point is
00:31:28
that when you have agents I think that
00:31:31
we are overestimating what a system of
00:31:33
record did David what you talked about
00:31:35
is actually just an encrypted file or
00:31:37
it's a bunch of rows in some database or
00:31:39
it's in some data Lake somewhere you
00:31:42
don't need to spend tens or hundreds of
00:31:45
millions of dollars to wrap your Revenue
00:31:49
in something that says it's a system of
00:31:51
record you don't need that actually you
00:31:53
can just pipe that stuff directly from
00:31:56
stripe into Snowflake and you can just
00:31:59
transform it and do what you will with
00:32:00
it and then report it I'll tell you
00:32:02
could do that today it's just that
00:32:04
that's an interesting point through
00:32:05
stake dinners and golf outings and all
00:32:07
this stuff we've sold Ci's this idea
00:32:12
that you need to wrap it in something
00:32:13
called a system of record and all I'm
00:32:15
saying is when you confront the total
00:32:18
cost of that versus what the alternative
00:32:21
that is clearly going to happen in the
00:32:23
next 5 or 10 years irrespective of
00:32:25
whether any of us Build It or Not it'll
00:32:27
be not be able to you just won't be able
00:32:29
to justify
00:32:31
it I'll say there's there's probably
00:32:34
also an aspect of this that we can't
00:32:37
predict what is going to work with
00:32:39
respect to data structure so right now
00:32:41
all of um all of the the tooling for AI
00:32:45
is on the front end and we haven't yet
00:32:47
Unleashed AI on the back end which is if
00:32:50
you told the AI here's all the data
00:32:52
ingest I'm going to be doing from all
00:32:54
these different points in my business
00:32:56
figure out what you want to do with all
00:32:58
that data the AI will eventually come up
00:33:01
with its own data structure and data
00:33:04
system nothing that will look nothing
00:33:07
like that's already happening right and
00:33:09
so that's nothing like what we have
00:33:11
today in the same vein that we don't
00:33:13
understand how the translation Works in
00:33:14
a in an in an llm we don't understand
00:33:16
how a lot of the function works a lot of
00:33:18
the data structure and data architecture
00:33:19
we won't understand clearly because it's
00:33:21
going to be opusc by the model driving
00:33:24
the the development there are open
00:33:25
source agentic Frameworks that already
00:33:27
do freeberg what you're saying that so
00:33:29
it's not true that it's not been done
00:33:31
it's already sure so maybe it's being
00:33:33
done right that's it hasn't been fully
00:33:35
implemented to replace this record there
00:33:37
are companies I'll give you an example
00:33:39
of one like mechanical Orchard they'll
00:33:42
go into the most gnarliest of
00:33:44
environments and what they will do is
00:33:45
they will launch these agents that
00:33:47
observe it's sort of what I told you
00:33:49
guys before the io stream of these apps
00:33:51
and then reconstruct everything in the
00:33:53
middle
00:33:55
automatically I don't understand why
00:33:58
we think that there's a world where
00:34:00
customer quality and NPS would not go
00:34:02
Skyhigh for a company that has some old
00:34:05
Legacy forr system and now they can just
00:34:07
pay you know mechanical Orchard a few
00:34:10
million bucks and they'll just replace
00:34:11
it in a matter of months it's G to
00:34:13
happen right yeah that's the very
00:34:15
interesting piece for me is I'm you know
00:34:17
watching startups you know working on
00:34:19
this the AI first ones I think are going
00:34:22
to come to it with a totally different
00:34:24
cost structure the idea of paying for
00:34:25
seats and I mean some of the these seats
00:34:28
are 5,000 per person year you nailed it
00:34:31
a year ago when you were like oh you
00:34:33
mentioned some company that had like
00:34:34
flat pricing at first by the way when
00:34:37
you said that I thought this is nuts but
00:34:40
you're right it actually makes a ton of
00:34:42
sense because if you have a fixed group
00:34:44
of people who can use this tooling to
00:34:48
basically effectively be as productive
00:34:51
as a company that's 10 times as big as
00:34:53
you you can afford to Flat price your
00:34:56
software because you can just work
00:34:58
backwards from what margin structure you
00:35:00
want and it's still meaningfully cheaper
00:35:02
than any other alternative a a lot of
00:35:05
startups now are doing consumption based
00:35:07
pricing so they're saying you know how
00:35:09
many um how many sales calls are you
00:35:12
doing how many are we analyzing as
00:35:14
opposed to how many sales Executives do
00:35:16
you have because when you have agents as
00:35:19
we're talking about those agents are
00:35:21
going to do a lot of the work so we're
00:35:23
going to see the number of people
00:35:24
working at companies become fixed and I
00:35:27
think the St team size that we're seeing
00:35:30
at a lot of large companies is only
00:35:32
going to continue it's going to be down
00:35:33
and to the right and if you think you're
00:35:35
going to get a high-paying job at a big
00:35:36
tech company and you have to beat the
00:35:40
agent you're going to have to beat the
00:35:41
Maestro who has five agents working for
00:35:43
them I think this is going to be a
00:35:45
completely different world uh Cham this
00:35:48
I I want to get back to openi with a
00:35:50
couple other pieces so let's wrap this
00:35:52
up get
00:35:53
next last word for you last word so look
00:35:57
look I I think that on the whole I agree
00:36:00
with Benny off here that there's more
00:36:02
net new opportunity for AI companies
00:36:05
whether they be startups or you know
00:36:07
existing big companies like Salesforce
00:36:09
that are trying to do AI then there is
00:36:12
disruption I think there will be some
00:36:13
disruption it's very hard for us to see
00:36:15
exactly what AI is going to look like in
00:36:17
five or 10 years so I don't want to
00:36:18
Discount the possibility that ex there
00:36:21
will be some disruption of existing
00:36:23
players but I think on the whole there's
00:36:25
more net new opportunity for example the
00:36:27
most highly valued public software
00:36:29
company right now in terms of AR
00:36:31
multiple is paler and I think that's
00:36:35
largely because the market perceives
00:36:37
paler is having a big AI opportunity
00:36:39
what is Pal's approach the first thing
00:36:41
paler does when they go into a customer
00:36:44
is they integrate with all of its
00:36:46
systems and they're dealing with the
00:36:47
largest Enterprises they're dealing with
00:36:48
the government the Pentagon Department
00:36:50
of Defense the first thing they do is go
00:36:52
in and integrate with all of these
00:36:55
Legacy systems and they collect all of
00:36:58
the data in one place they call it
00:37:00
creating a digital twin and once all the
00:37:03
data is in one place with the right
00:37:04
permissions and safeguards now analysts
00:37:07
can start working it and that was their
00:37:08
historical value proposition but in
00:37:11
addition AI can now start working that
00:37:13
problem so anything that the analyst
00:37:15
could work now ai is going to be able to
00:37:16
work and so they're in an ideal position
00:37:19
to master these new AI workflows so what
00:37:22
is the point I'm making it's just that
00:37:24
you can't just throw an llm at these
00:37:26
large enter Rises you have to go in
00:37:29
there and integrate with the existing
00:37:31
systems it's not about ripping out the
00:37:32
existing systems because that's just a
00:37:34
lot of headaches that nobody needs it's
00:37:36
generally an easier approach just to
00:37:38
collect except when the renewal comes
00:37:40
what happens when you have to you know
00:37:42
billion on something yeah and then
00:37:45
you're going to renegotiate or you're
00:37:46
going to spend a billion dollars again
00:37:47
five years from now it just doesn't seem
00:37:49
very likely there's going to be a lot of
00:37:51
Hardcore negotiations going on shama
00:37:53
people are going to ask for 20% off 50%
00:37:55
off and people going to be more
00:37:57
competitive that's all I Su I suspect
00:37:59
palun go to market when they start to
00:38:01
release scale they'll be able to
00:38:02
underprice a bunch of these other
00:38:04
Alternatives and it so I think
00:38:07
that when you look at
00:38:11
the impacts and pricing that all of
00:38:15
these open source and close Source model
00:38:17
companies have now introduced in terms
00:38:19
to the price per token what we've seen
00:38:22
is just a massive step function lower
00:38:25
right so it is incredible
00:38:28
deflationary so the things that sit on
00:38:30
top are going to get priced as a
00:38:33
function of that of that cost which
00:38:35
means it will be an order of magnitude
00:38:37
cheaper than the stuff that it replaces
00:38:39
which means that a company would almost
00:38:42
have to purposely want to keep paying
00:38:45
tens of millions of dollars when they
00:38:47
don't have to they would need to make
00:38:49
that as an explicit decision and I think
00:38:52
that very few companies will be in a
00:38:54
position to be that Cavalier in 5 and 10
00:38:57
10 years so you're either going to
00:38:59
rebase the revenues of a bunch of these
00:39:03
existing deterministic companies or
00:39:05
you're going to create an entire economy
00:39:07
of new ones that have a fraction of the
00:39:09
revenues today but a very different
00:39:11
profitability profile I I think whenever
00:39:15
you're dealing with a whenever you're
00:39:16
dealing with A disruption as big as this
00:39:19
current one I think it's always tempting
00:39:21
to think in terms of the existing Pi
00:39:24
getting disrupted and and shrunk as
00:39:27
oppos to the pie getting so big with new
00:39:29
use cases that on the whole the
00:39:32
ecosystem benefits no no no I agree with
00:39:35
that I suspect That's What's happen no I
00:39:37
I I agree with that my my only point is
00:39:39
that the pie can get bigger while the
00:39:41
slices get much much
00:39:44
smaller well I mean I right between the
00:39:47
two of you I think is um the truth
00:39:50
because what's happening is if you look
00:39:52
at investing it's very hard to get into
00:39:55
these late stage companies because they
00:39:56
don't need as much Capital because to
00:39:58
your point shth they when they do hit
00:40:01
profitability with 10 or 20 people the
00:40:03
revenue per employee is going way up if
00:40:07
you look at Google Uber Airbnb and
00:40:10
Facebook meta they have the same number
00:40:12
or less employees and they did three
00:40:14
years ago but they're all growing in
00:40:15
that 20 to 30% a year which means in but
00:40:19
two to three years each of those
00:40:21
companies has doubled Revenue per
00:40:22
employee so that concept of more
00:40:25
efficiency and then that trickles down
00:40:28
Sachs to the startup investing space
00:40:29
where you and I are I'm a preed seed
00:40:31
investor you're a seed series a investor
00:40:34
if you don't get in in those three four
00:40:35
rounds I think it's going to be really
00:40:37
expensive and the companies are not
00:40:38
going to need as much money Downstream
00:40:41
speaking of of investing in late stage
00:40:43
companies we never Clos the loop on the
00:40:45
whole open AI thing what did we think of
00:40:48
the fact that they're completely
00:40:51
changing the structure of this company
00:40:52
they're changing it into a corporation
00:40:53
from the
00:40:54
nonprofit and Sam's now getting it 101
00:40:57
billion stock
00:40:58
package he's not in for the money he has
00:41:01
health insurance
00:41:03
sex but we never never Congress I don't
00:41:06
need money enough money I just needed
00:41:08
the health insurance pull the clip up
00:41:10
Nick pull the clip up I mean it's the
00:41:11
funniest clip
00:41:13
ever
00:41:14
C no they said Congress watch this this
00:41:17
what in
00:41:18
Congress you make a lot of money do you
00:41:21
I make no I paid enough for health
00:41:23
insurance I have no equity and open AI
00:41:24
really that's interesting you need a
00:41:26
large
00:41:28
I need a what you need a lawyer or an
00:41:30
agent I I'm doing this cuz I love
00:41:33
it thank you it's the greatest look at
00:41:36
me don't believe him I this can I ask
00:41:38
you a question there saxs are you doing
00:41:41
this Venture Capital where you put the
00:41:44
money in the startups cuz you love it or
00:41:48
cuz you're looking to get another Home
00:41:50
In a Coal City and put more jet fuel in
00:41:53
that plane I need an answer for the
00:41:54
people of the sovereign state of
00:41:56
Mississippi no Louisiana that's John
00:41:59
Kennedy from
00:42:00
louisi he's a very smart guy actually
00:42:03
with a lot of you know sort of common
00:42:07
folk wisdom he got that simple talk
00:42:10
simple talk we Shooters
00:42:12
here actually yeah he's very funny but
00:42:16
he's very funny if you listen to him he
00:42:19
knows how to slice and dice you might
00:42:21
need to get yourself uh one of them
00:42:23
fancy Agents from Hollywood or an
00:42:26
attorney from the Wilson Sun cini
00:42:28
Corporation to renegotiate your contract
00:42:31
son because you're worth a lot more from
00:42:32
what I can gather in your performance
00:42:34
today than just some simple healthare
00:42:37
and uh I hope you took the Blue Cross
00:42:39
Blue Shield I would like to make two
00:42:42
semi-serious observations let's go you
00:42:44
please get us back on track I think the
00:42:46
first is that there's going to be a lot
00:42:48
of people that are looking at the
00:42:49
architecture of this conversion because
00:42:52
if it passes muster everybody should do
00:42:54
it think about this model let's just
00:42:57
that you're in a market and you start as
00:42:59
a
00:43:00
nonprofit what that really means is you
00:43:02
pay no income tax so for a long time you
00:43:07
put out a little bit of the percentage
00:43:09
of whatever you
00:43:11
earn but you can now outspend and outco
00:43:14
compete all of your competitors and then
00:43:17
once you win you flip to a corporation
00:43:21
that's a great hack on the tax code and
00:43:24
you let the donators get first bite of
00:43:26
the Apple if you do convert because
00:43:29
remember venod and Hoffman got all their
00:43:31
shares on the conversion the other way
00:43:33
will also work as well because it's
00:43:35
there's no there's nothing that says you
00:43:36
can't go in the other direction so let's
00:43:38
assume that you're already a for-profit
00:43:41
company but you're in a space with a
00:43:42
bunch of competitors can't you just do
00:43:45
this conversion in Reverse become a
00:43:48
nonprofit again you pay no income tax so
00:43:51
now you are economically advantaged
00:43:54
relative to your competitors and then
00:43:57
when they wither and die or you can
00:43:58
outspend them you flip back to a
00:44:01
for-profit again I think the the point
00:44:04
is that there's a lot of people that are
00:44:07
going to watch this closely
00:44:10
and if it's legal and it's allowed I I
00:44:13
just don't understand why everybody
00:44:15
wouldn't do this yeah I mean that was
00:44:17
elon's Point as well and I mean the
00:44:19
second thing which is just more of
00:44:22
like cultural observation is and you
00:44:26
brought up Elon my comment to you guys
00:44:28
yesterday and I'll just make the comment
00:44:30
today
00:44:31
it's a little bit disheartening to see a
00:44:34
situation where Elon built something
00:44:37
absolutely incredible defied every
00:44:40
expectation and then had the justice
00:44:44
system take $ 55 billion away from him
00:44:49
his payment package you're referring to
00:44:50
at payment package the options in Tesla
00:44:54
and then on the other side Sam's going
00:44:56
to pull some like this
00:44:58
off definitely pushing the boundaries
00:45:01
and he's going to make 10 billion and I
00:45:04
just think when you put those two things
00:45:06
in
00:45:07
contrast that's not how the system
00:45:10
should probably work I think is what
00:45:11
most people would say freeberg you've
00:45:13
been a little quiet here any thoughts on
00:45:15
the transaction the nonprofit to
00:45:17
for-profit if you were looking at that
00:45:19
in what you're doing do you see a way
00:45:21
that ohal could take a nonprofit status
00:45:25
raise a bunch of money through donations
00:45:27
for virtuous work then license those
00:45:29
patents to your for-profit it would that
00:45:31
be advantageous to you and and do you
00:45:33
think this could become absolutely zero
00:45:35
idea I have no idea what they're doing I
00:45:38
don't know how they're converting a
00:45:39
nonprofit to a for-profit none of us
00:45:40
have the details on this there's there
00:45:42
may be significant tax implications
00:45:44
payments they need to make I don't think
00:45:45
any of us know I certainly don't I don't
00:45:48
know if there's actually a real benefit
00:45:50
here if there is I'm sure everyone would
00:45:52
do it no one's doing it so there's
00:45:54
probably a reason why it's difficult I
00:45:56
don't know been done a couple times the
00:45:58
the Mozilla Foundation did it we talked
00:45:59
about that in our previous episode Saks
00:46:01
you want to you want to wrap us up here
00:46:02
on the corporate structure any final
00:46:03
thoughts I mean Elon put in 50 million I
00:46:06
think he gets the same as Sam don't you
00:46:09
think he should just chip off 7% for
00:46:11
Elon and not that Elon needs the money
00:46:13
where he's asking but I'm just wondering
00:46:15
why Elon doesn't get the 7% and get or
00:46:18
you know if put in 50 did he put in $50
00:46:21
million he put in 50 million is the
00:46:23
report right in the nonprofit yeah
00:46:25
Hoffman put in 10 look I said on a
00:46:27
previous show that this organizational
00:46:30
chart of open AI was ridiculously
00:46:32
complicated and they should go clean it
00:46:34
up they should open up the books and
00:46:35
straighten everyone out and I also said
00:46:38
that as part of that they could give Sam
00:46:40
Alman a CEO option Grant and they should
00:46:42
also give Elon some Fair compensation
00:46:45
for being the seed investor who put in
00:46:47
the first $50 million and co-founder and
00:46:50
what you're seeing is well they're kind
00:46:52
of doing that they're opening up the
00:46:53
books they're straightening out the
00:46:56
corporate stru
00:46:57
they're giving Sam his option Grant but
00:46:59
they didn't do anything for Elon and I'm
00:47:02
not saying this as elon's friend I'm
00:47:04
just saying that it's not really fair to
00:47:07
basically go fix the original situation
00:47:10
you're making it into a for-profit
00:47:13
you're giving everyone shares but the
00:47:15
guy who puts in the Original Seed Capal
00:47:16
doesn't get anything that's ridiculous
00:47:20
and you know what they're basically
00:47:21
saying to Elon is if you don't like it
00:47:23
just sue us I mean that's basically what
00:47:26
they're doing and I said that they
00:47:28
should go clean this up but they should
00:47:30
make it right with everybody so how do
00:47:32
you not make it right with Elon I
00:47:34
haven't talked him about this but he
00:47:36
reacted on X saying this is really wrong
00:47:39
it appeared to be a surprise to him I
00:47:40
doubt he knew this was coming so the
00:47:42
company apparently made no effort to
00:47:45
make things right with him and I think
00:47:48
that that is a bit ridiculous if you're
00:47:50
going to clean this up if you're going
00:47:51
to change the original purpose of this
00:47:54
organization to being a standard
00:47:57
for-profit company where the CEO who
00:48:00
previously said he wasn't going to get
00:48:02
any compensation is now getting 10
00:48:04
billion of compensation how do you do
00:48:06
that and then not clean it up for the
00:48:08
co-founder who put in the first $50
00:48:10
million yeah that doesn't make sense to
00:48:12
me and you know when Reed was on our pod
00:48:14
he he said well elon's Rich enough well
00:48:17
that that's not a principled excuse I
00:48:19
mean does venod ever act that way does
00:48:21
Reed ever act that way do they ever say
00:48:24
well you know you don't need to do
00:48:25
what's fair for me because I'm already
00:48:27
rich that's that's not a principled
00:48:29
answer the argument that I heard was
00:48:31
that Elon was given the opportunity to
00:48:33
invest along with Reed with VOD and he
00:48:36
he declined to participate in the
00:48:38
for-profit investing side that everyone
00:48:41
else participated we made that argument
00:48:43
and I think it's the best argument the
00:48:44
company has but let's think about that
00:48:46
argument maybe Elon was busy that week
00:48:49
maybe Elon already felt like he had put
00:48:51
all the money that he had allocated for
00:48:53
something like this into it because he
00:48:54
put in a $50 million check whereas reput
00:48:57
in 10 we don't know what Elon was
00:49:00
thinking at that time maybe there was a
00:49:01
crisis at Tesla and he was just really
00:49:03
busy the point is Elon shouldn't have
00:49:05
been obligated to put in more money into
00:49:09
this Venture the fact of the matter is
00:49:11
they're refactoring the whole Venture
00:49:13
Elon had an expectation when he put in
00:49:15
the 50 million that this would be a
00:49:17
nonprofit and stay a nonprofit and
00:49:19
they're changing that and if they change
00:49:21
it they have to make things right with
00:49:22
him it doesn't really matter whether he
00:49:24
had a a subsequent opportunity to to
00:49:26
invest he wasn't obligated to to make
00:49:29
that investment what he had an
00:49:31
expectation of is that his $50 million
00:49:34
to be used for a philanthropic purpose
00:49:36
and clearly it has not been yeah and In
00:49:39
fairness to a node he bought that
00:49:40
incredible beachfront property and
00:49:42
donated it to the public trust so we can
00:49:43
all Surf and have our Halloween party
00:49:45
there so it's all good thank you venod
00:49:47
for giving us that incredible Beach I
00:49:50
want to talk to you guys about
00:49:51
interfaces that came up chth in your
00:49:54
headwinds or your your your four pack of
00:49:56
reasons that you know open AI when you
00:49:59
steal men the barcase could have
00:50:01
challenges obviously we're seeing that
00:50:04
and it is emerging that meta is working
00:50:08
on some AR glasses that are really
00:50:10
impressive additionally I've installed
00:50:13
iOS 18 which is Apple intelligence that
00:50:16
works on 15 phones and 16 phones 18 is
00:50:19
the iOS did any of you installed the
00:50:21
beta of iOS 18 yet and use Siri it's
00:50:24
pretty clear with this new one that
00:50:26
you're going to be able to talk to Siri
00:50:27
as an llm like you do in chat GPT mode
00:50:30
which I think means they will not make
00:50:33
themselves dependent on chat GPT and
00:50:34
they will siphon off half the searches
00:50:37
that would have gone to chat GPT so I
00:50:38
see that as a serious sir is not very
00:50:40
good J kll and you know this because
00:50:42
when you were driving me to the airport
00:50:43
yesterday we tested it it didn't work
00:50:45
yes you tried you try he tried to he
00:50:47
tries to execute this joke where he's
00:50:48
like hey Siri sent choth poly haaa a
00:50:52
message and it was a very off-color
00:50:54
message I'm not gonna say what it is was
00:50:55
a spicy joke and then it's like okay
00:50:57
great sending Linda blah blah blah he
00:51:00
like no stop like no don't send that
00:51:02
joke to her it hallucinates it almost
00:51:04
sends it to some other you know some
00:51:06
woman in his contact would have been
00:51:08
really damaging it's not very good Jason
00:51:10
it's not very good well but what I will
00:51:11
say is there are features of it where if
00:51:14
you squint a little bit you will see
00:51:17
that Siri is going to be conversational
00:51:20
so when I was talking to it with music
00:51:22
and you know you you can have a
00:51:23
conversation with it and do math like
00:51:25
you can do with cat GPT version and you
00:51:28
have Microsoft doing that with their
00:51:30
co-pilot and now met is doing it at the
00:51:33
top of each one so everybody's going to
00:51:35
try to inter intercept the queries and
00:51:37
the voice interface so chat gp4 is now
00:51:40
up against meta Siri Apple and Microsoft
00:51:44
for for that interface it's going to be
00:51:46
challenging but let's talk about these
00:51:47
meta glasses here meta showed off the AR
00:51:50
glasses that Nick will pull up right now
00:51:53
these aren't goggles goggles look like
00:51:55
ski goggles that that's what apple is
00:51:57
doing with their Vision Pro uh or when
00:52:00
you see the meta Quest you know how
00:52:02
those work those are VR with cameras
00:52:05
that will create a version of the world
00:52:07
these are actual chunky sunglasses like
00:52:09
the ones I was wearing earlier when I
00:52:11
was doing the bit so these let you
00:52:14
operate in the real world and are
00:52:18
supposedly extremely expensive they made
00:52:20
a thousand prototypes they were lending
00:52:21
a bunch of influencers and folks like um
00:52:26
Gary vanderchuck uh use them and they're
00:52:29
not ready for prime time but the way
00:52:31
they work freeberg is there's a
00:52:32
wristband that will track your fingers
00:52:35
and your wrist movement so you could be
00:52:36
in a conversation like we are here on
00:52:38
the Pod and below the desk you could be
00:52:40
you know moving your arm and hand around
00:52:42
to be doing replies to I don't know
00:52:44
incoming messages or whatever it is what
00:52:47
do you think of this AR vision of the
00:52:49
world and meta making this progress well
00:52:52
I think it ties in a lot to the the AI
00:52:56
discussion because I I think we're
00:52:58
really witnessing this big shift from
00:53:02
and this big transition in Computing
00:53:04
probably the biggest transition since
00:53:05
mobile you know we moved from Main
00:53:08
frames to desktop computers everyone had
00:53:11
kind of this computer on their desktop
00:53:12
but used a mouse and a keyboard to
00:53:13
control it to mobile where you had a
00:53:15
keyboard and clicking and touching on
00:53:17
the screen to do things on it and now to
00:53:19
what I would call this kind of ambient
00:53:21
Computing method and you know I think
00:53:25
the big difference is control and
00:53:27
response in directed Computing you're
00:53:30
kind of telling the computer what to do
00:53:32
you're controlling it you're using your
00:53:33
mouse or your keyboard to to go to this
00:53:36
website so you type in a website address
00:53:38
then you click on the thing that you
00:53:40
want to click on and you kind of keep
00:53:42
doing a series of work to get the
00:53:43
computer to go access the information
00:53:46
that you ultimately want to achieve your
00:53:47
objective but with ambient Computing you
00:53:50
can more kind of cleanly state your
00:53:52
objective without this kind of directive
00:53:54
process you can say hey I I want to say
00:53:57
I want to have dinner in New York next
00:53:58
Thursday at the at a Michelin star
00:54:00
restaurant at 5:30 book me something and
00:54:02
it's done and I think that there are
00:54:04
kind of five core things that are needed
00:54:06
for this to work um both in control and
00:54:09
response it's voice control gesture
00:54:12
control and eye control are kind of the
00:54:14
control pieces that replace you know
00:54:16
mice and clicking and touching and
00:54:18
keyboards and then response is audio and
00:54:21
kind of integrated visual which is the
00:54:23
idea of the goggles voice control works
00:54:25
have you guys used the open AI voice
00:54:27
control system lately I mean it is
00:54:29
really incredible I had my earphones in
00:54:31
and I was like doing this exercise I was
00:54:33
trying to learn something so I told open
00:54:35
AI to start quizzing me on this thing
00:54:38
and I just did a 30 minute walk and
00:54:39
while I was walking it was asking me
00:54:41
quiz questions and I would answer it
00:54:42
tell me I was right or wrong it was
00:54:43
really this incredible dialogue
00:54:45
experience so I think the voice control
00:54:46
is there I don't know if you guys have
00:54:48
used Apple Vision Pro but gesture
00:54:49
control is here today you can do single
00:54:52
finger movements with apple Vision Pro
00:54:54
it triggers actions and I control is
00:54:56
incredible you look at the letters you
00:54:58
want to have kind of spelled out or you
00:54:59
look at the thing you want to activate
00:55:00
and it does it so all of the control
00:55:03
systems for this ambient Computing are
00:55:04
there and then the AI enables this kind
00:55:06
of audio response where it speaks to you
00:55:09
and the big breakthrough that's needed
00:55:11
that I don't think we're quite there yet
00:55:12
but maybe Zuck is highlighting that
00:55:14
we're almost there and apple Vision Pro
00:55:15
feels like it's almost there except it's
00:55:17
big and bulky and expensive is
00:55:19
integrated visual where the ambient
00:55:20
visual interface is always there and you
00:55:23
can kind of engage with it so there's
00:55:24
this big change I don't think that
00:55:26
mobile handsets are going to be around
00:55:27
in 10 years I don't think we're going to
00:55:29
have this like phone in our pocket that
00:55:30
we're like pressing buttons on and
00:55:32
touching and telling it where on the
00:55:34
browser to go to the browser interface
00:55:36
is going to go away I think so much of
00:55:38
how Computing is done how much how we
00:55:40
integrate with data in the world and how
00:55:42
the computer ultimately fetches that
00:55:44
data and does stuff with it for us is
00:55:46
going to completely change to this
00:55:47
ambient model so I'm um I'm pretty
00:55:50
excited about this Evolution but I think
00:55:52
that what we're seeing with Zuck what we
00:55:53
saw with apple Vision Pro and all of the
00:55:55
open AI demos they all kind of converge
00:55:58
on this very incredible shift in
00:56:00
Computing um that will kind of become
00:56:02
this ambient system that exists
00:56:04
everywhere all the time and I know folks
00:56:05
have kind of mentioned this in the past
00:56:07
but I think we're really seeing it kind
00:56:08
of all come together now with these five
00:56:10
key
00:56:11
things jamath any thoughts on Facebook's
00:56:15
progress with ar and how that might
00:56:20
impact Computing and interfaces when
00:56:23
paired with language models I think
00:56:26
think David's right that there's
00:56:28
something that's going to be totally new
00:56:31
and
00:56:32
unexpected so I agree with that part of
00:56:36
what free Brook says I am still not sure
00:56:40
that glasses are the perfect form factor
00:56:42
to be ubiquitous when you look at a
00:56:46
phone a phone makes complete sense for
00:56:50
literally everybody right man woman old
00:56:54
young every
00:56:58
race every country of the world it's
00:57:01
such a ubiquitously obvious form
00:57:06
factor but the thing is like that
00:57:08
initial form factor was so different
00:57:10
than what it replaced even if you looked
00:57:13
at like flip phones versus that first
00:57:14
generation iPhone so I I do think
00:57:18
freeberg you're right that there's like
00:57:20
this new way of interacting that is
00:57:23
ready to explode onto the scene
00:57:26
and I think that these guys have done a
00:57:29
really good job with these classes I
00:57:30
mean like I give them a lot of credit
00:57:32
for sticking with it and iterating
00:57:34
through it and getting it to this place
00:57:36
it looks meaningfully better than the
00:57:37
Vision Pro to be totally
00:57:40
honest but I'm still not convinced that
00:57:42
we've explored the best of our
00:57:45
creativity in terms of the devices that
00:57:47
we want to use with these AI models you
00:57:49
need some visual interface I think the
00:57:51
question is where is the visual
00:57:52
interface is it in the is in in the wall
00:57:55
but do I mean well when you're asking
00:57:57
like I want to watch chamath on Rogan
00:57:59
like I don't just want to hear I want to
00:58:01
see like when I want to visualize stuff
00:58:03
I want to visualize it I want to look at
00:58:05
the food I'm buying online I want to
00:58:06
look at pictures of the restaurant I'm
00:58:08
GNA go to but how much of that time when
00:58:10
you say those things are you not near
00:58:13
some screen that you can just project
00:58:15
and broadcast that on I Me Maybe the
00:58:17
model if the use case is I'm walking in
00:58:20
the park and I need to watch TV at the
00:58:22
same time I don't think that's a real
00:58:24
use case I I think you're on this one
00:58:25
wrong WR shth because I saw this
00:58:29
revolution in Japan maybe 20 years ago
00:58:32
they got obsessed with augmented reality
00:58:34
there were a ton of startups right as
00:58:36
they started getting into the mobile
00:58:37
phones and the use cases were really
00:58:40
very compelling and we're starting to
00:58:41
see them now in education and when
00:58:42
you're at dinner with a bunch of friends
00:58:45
how often does picking up your phone and
00:58:49
you know looking at a message disturb
00:58:50
the flow well people will have glasses
00:58:52
on they'll be going for walks they'll be
00:58:54
driving they'll be at a dinner party
00:58:55
they'll be with their kids and you'll
00:58:57
have something on like Focus mode you
00:58:59
know whatever the equivalent is in apple
00:59:01
and a message will come in from your
00:59:02
spouse or from your child but you won't
00:59:05
have to take your phone out of your
00:59:06
pocket and I think once these things
00:59:08
weigh a lot less you're going to have
00:59:10
four different ways to interact with
00:59:12
your computer in your pocket your phone
00:59:13
your watch your airpods whatever you
00:59:16
have in your ears and the glasses and I
00:59:17
bet you glasses are going to take like a
00:59:19
third of the tasks you do I mean what is
00:59:22
the point of taking out your phone and
00:59:23
watching The Uber come to you but seeing
00:59:25
that little strip that tells you the
00:59:27
Uber is 20 minutes away 15 minutes away
00:59:29
or what the gate number is I don't have
00:59:31
that anxiety well I don't know if it's
00:59:33
anxiety but I just think it's ease of
00:59:35
use all those 15 minutes 10 minutes
00:59:37
that's that's the definition I think it
00:59:38
adds up I think it taking your phone out
00:59:40
of your pocket 50 times those are all
00:59:43
useless notifications the whole thing is
00:59:45
to train yourself to realize that it'll
00:59:47
come when it comes okay sack do you have
00:59:48
any thoughts on this uh impressive demo
00:59:51
or you know the demo that people who've
00:59:52
seen have said is pretty pretty darn
00:59:54
compelling I think it it does look
00:59:56
pretty impressive I mean you can wear
00:59:58
these meta Orion glasses
01:00:01
around you know and look and look like a
01:00:03
human I mean you might look like Eugene
01:00:05
Levy but you'll still look like a
01:00:07
semi-normal person whereas you can't
01:00:10
wear the Apple Vision Pro I mean you
01:00:12
can't wear that around what they don't
01:00:13
look good you don't like them Nick can
01:00:16
you please find a picture of Eugene
01:00:19
Levy har so I mean it seems like a major
01:00:23
advancement certainly compared to apple
01:00:24
Vision Pro I mean you don't hear you
01:00:26
don't hear about the Apple Vision Pros
01:00:28
anymore at all I mean those things came
01:00:30
and
01:00:31
[Laughter]
01:00:36
went it's pretty funny it seems to me
01:00:39
that who's that meta is executing
01:00:42
extremely well I mean you had the very
01:00:44
successful cost cutting which Wall
01:00:45
Street
01:00:46
liked Zuck published that letter which I
01:00:49
give him credit for regretting the
01:00:50
censorship that meta did which was at
01:00:52
the beest of the deep State yeah
01:00:56
they made huge advancements in AI I
01:00:58
don't think they were initially on The
01:00:59
Cutting Edge of that but they've caught
01:01:01
up and now they're leading the open
01:01:03
source yeah with with llama 3.2 and now
01:01:06
it seems to me that they're ahead on
01:01:08
augmented reality ever since uh Zuck
01:01:11
grew out the hair yeah you know don't
01:01:13
ever don't ever cut the hair it's like
01:01:14
Samson I mean you've been it's like
01:01:16
Samson Bas based Zuck is the best
01:01:20
Zuck I want to be clear I I think these
01:01:23
glasses are are going to be successful
01:01:25
my only comment is that I think that
01:01:28
when you look back 25 and 30 years from
01:01:30
now and say that was the killer AI
01:01:33
device I don't think it's going to look
01:01:35
like something we know today that's my
01:01:38
only point and maybe it's going to be
01:01:40
this thing that Sam Alman and Johnny IV
01:01:42
are baking up that's supposed to be this
01:01:45
AI infused iPhone Killer maybe it's that
01:01:48
thing I doubt that will be a pair of
01:01:51
glasses or a phone or a pin if you think
01:01:55
about like so so take the constraints on
01:01:57
I don't need a keyboard because I'm not
01:02:00
going to be typing stuff I don't need a
01:02:02
normal browser interface you could see a
01:02:05
device come out that's almost like
01:02:08
smaller than the palm of your hand that
01:02:10
gives you enough of the visuals and all
01:02:11
it is is a screen with maybe two buttons
01:02:13
on the side and it's all audio driven
01:02:15
you put a headset in and you're
01:02:17
basically just talking or using gesture
01:02:20
or looking at it to kind of describe
01:02:21
where you want things to go and it can
01:02:22
create an entirely new Computing
01:02:24
interface because AI does all of these
01:02:26
things withed text with gesture control
01:02:29
with eye control and with audio control
01:02:32
and then it can just give you what you
01:02:33
want on a screen and all you're getting
01:02:34
is a simple interface so jamat you may
01:02:36
be right it might be a big watch or a
01:02:38
handheld thing that's much smaller than
01:02:39
an iPhone and just all it is is a screen
01:02:42
with nothing I really resonate when you
01:02:45
talk about voice only because I
01:02:47
think it's like it you I think there's
01:02:50
like a part of like social
01:02:52
decorum that all of these goggles and
01:02:55
glasses
01:02:57
violate and I think we're going to have
01:02:58
to decide as a society whether that's
01:03:00
going to be okay and then I think like
01:03:03
are when you like go trekking in Nepal
01:03:06
are you going to encounter somebody
01:03:08
wearing AR glasses I think the odds are
01:03:10
pretty low but you do see people today
01:03:12
with a phone so what do they replace it
01:03:14
with and I think voice as a modality is
01:03:18
is I think it's more incredible that
01:03:20
that could be used by 8 billion people I
01:03:22
think social Fabric's more affected by
01:03:24
people staring at their phones all the
01:03:25
time you sit on a bus you sit at a
01:03:26
restaurant you go to dinner with someone
01:03:27
and they're staring at their phone like
01:03:29
you know spes friends we all deal with
01:03:31
it where you feel like you're not
01:03:33
getting attention from the person that
01:03:35
you're interfacing with in the real
01:03:36
world because they're so connected to
01:03:38
the phone if we can disconnect the phone
01:03:40
but still get take away this kind of
01:03:42
addictive feedback loop system but still
01:03:44
give you this Computing ability in a
01:03:46
more ambient way that allows you to
01:03:48
remain engaged in the physical world I
01:03:49
think everyone feel better hurt your
01:03:51
feeling when he's playing chess and not
01:03:53
paying attention yeah I'll be playing
01:03:54
chess on my our glasses while pretending
01:03:57
to listen to to you you
01:03:59
idiot oh F's he's bu them he got he got
01:04:03
version one what one point I want to
01:04:05
just hit on is the reason why these
01:04:07
glasses have a chance of working is
01:04:09
because of AI I mean Facebook initially
01:04:12
made these that's exactly my point it's
01:04:14
exactly my point Facebook or meta made
01:04:16
these huge investments before AI was a
01:04:18
thing and in a way I think they've kind
01:04:20
of gotten lucky because what AI gives
01:04:22
you is uh voice and audio so you can
01:04:26
talk to the glasses or whatever the
01:04:27
wearable is it can talk to you perfect
01:04:30
natural language and computer vision
01:04:33
allows it to understand the world around
01:04:35
you so whatever this device is it can be
01:04:37
a true personal digital assistant in the
01:04:40
real world and that's telling you if you
01:04:42
guys play with apple Vision Pro have any
01:04:44
of you actually like used it to any
01:04:46
extent if you actually used it I used it
01:04:49
for a day or a night when we were
01:04:51
playing poker and I've never used it
01:04:54
again since right which I get but I do
01:04:56
think that it has these tools in it
01:04:58
similar to like the original Macintosh
01:05:00
had these incredible Graphics uh editors
01:05:02
like mac paint and all these things that
01:05:04
like people didn't like get addicted to
01:05:06
at the time but they became this like
01:05:09
tool that completely revolutionized
01:05:11
everything in Computing later and fonts
01:05:14
and so on but like this I think has
01:05:16
these tools Apple Vision Pro with
01:05:18
gesture control and the keyboard and the
01:05:21
eye control those aspects of that device
01:05:24
highlight where this could all go which
01:05:26
is this these systems can kind of be
01:05:28
driven without keyboards without typing
01:05:31
without like you know moving your finger
01:05:33
around without CLI I think that's the I
01:05:35
think that's the key observation I
01:05:36
really agree with what you just said
01:05:38
it's this it's this idea that you're
01:05:40
just you're liberated from the hunting
01:05:43
and pecking the controlling it's you
01:05:45
don't need to control the computer
01:05:47
anymore the computer now knows what you
01:05:49
want and then the computer can just go
01:05:51
and do the work and they can respond so
01:05:53
now this is the behavior change that I
01:05:55
don't think we're fully giving enough
01:05:57
credit to so today part of what Jason
01:06:00
talked about what I called anxiety is
01:06:02
because of the information architecture
01:06:04
of these apps that is totally broken and
01:06:06
the reason why it's broken is when you
01:06:08
tell an AI agent get me the cheapest car
01:06:11
right now to go to XYZ place it will go
01:06:15
and look at lft and Uber and whatever
01:06:17
it'll provision the car and then it'll
01:06:19
just tell you when it's coming and it
01:06:21
will break this cycle that people have
01:06:24
of having to check these apps for what
01:06:26
is useless filler information and when
01:06:28
you strip a lot of that notification
01:06:30
traffic away I think you'll find that
01:06:32
people start looking at each other in
01:06:34
the face more often and I think that
01:06:36
that's a net positive so will will meta
01:06:39
sell hundreds of millions of these
01:06:41
things I suspect probably but all I'm
01:06:44
saying is if you look backwards 30 years
01:06:47
from now what is the device that sells
01:06:49
in the
01:06:50
billions it's probably not a form factor
01:06:52
that we understand today I just want to
01:06:54
point out like the the form factor
01:06:56
you're seeing now is going to get
01:06:58
greatly reduced these were um some of
01:07:00
the early Apple um I don't know if you
01:07:02
guys remember these but frog design made
01:07:05
these crazy tablets in the
01:07:08
80s that were the advenal inspiration
01:07:10
for the iPad you know 25 years later I
01:07:14
guess exactly uh and so that's the
01:07:16
journey we're on here right now this
01:07:18
clunky and and these are not functional
01:07:20
protot dude the Apple Newton is like and
01:07:24
then it turns out hey you throw away the
01:07:25
sty and you got an iPhone right and
01:07:28
everything gets a million x better the
01:07:29
other subtle thing that's happening
01:07:31
which I don't think we should sleep on
01:07:32
is that the airpods are probably going
01:07:36
to become much more socially acceptable
01:07:39
to wear on a 24x7 basis because of this
01:07:41
feature that allows it to become a
01:07:43
useful hearing aid and I think as it
01:07:46
starts being worn in more and more
01:07:48
social environments and as the form
01:07:51
factor of that shrinks that's when I
01:07:53
really do think we're going to find some
01:07:55
very novel use case which
01:07:57
is you know very unobtrusive it kind of
01:08:00
Blends into your own
01:08:02
physical makeup as a person without it
01:08:04
really sticking out I think that's when
01:08:06
you'll have a really killer feature but
01:08:08
I think that the the airpods as hearing
01:08:10
aids will also add a lot so meta's doing
01:08:13
a lot Apple's doing a lot but I don't
01:08:15
think we've yet seen the super killer
01:08:17
Hardware device yet and there was an
01:08:19
interesting Waypoint Microsoft had the
01:08:22
first tablets here's the the Microsoft
01:08:24
tablet for those of you watching
01:08:26
that came you know I don't know this was
01:08:27
the late 90s or early 2000s freedberg if
01:08:30
you remember it these like incredibly oh
01:08:33
bulky tablets that Bill Gates was
01:08:35
bringing to all the events 99 2000 that
01:08:39
era so you get a lot of false starts
01:08:41
they're spending I think close to 20
01:08:43
billion doar a year on this arv or
01:08:45
anyway we're definitely on this path to
01:08:47
ambient Computing I don't think I don't
01:08:48
think this whole like hey you got to
01:08:49
control a computer thing is anything my
01:08:51
kids are going to be doing in 20 years
01:08:52
this is uh this is the convergence of
01:08:54
like three or four really interesting
01:08:56
technological waves all right just uh
01:08:58
dovetailing with tech jobs and the
01:09:00
static team size there is a report of a
01:09:03
blue collar Boom the tool belt
01:09:06
generation is what gen Z is being
01:09:09
referred to as a report in the Wall
01:09:11
Street Journal reports hey tech jobs
01:09:13
have dried up we all we're all seeing
01:09:15
that and according to indeed Developer
01:09:17
jobs down more than 30% since February
01:09:19
of 2020 pre-co of course if you look at
01:09:23
layoffs.fyi you'll see all
01:09:26
you know tech jobs that have been
01:09:27
eliminated since 2022 over a half
01:09:30
million of them bunch of things at play
01:09:32
here and the Wall Street Journal notes
01:09:35
that entry level Tech workers are
01:09:37
getting hit the hardest especially all
01:09:40
these recent college graduates and if
01:09:42
you look at a historical College
01:09:45
enrollment let's pull up that chart Nick
01:09:47
you can see here undergraduate graduate
01:09:48
and total with the red line we peaked at
01:09:51
21 million people in either graduate
01:09:54
school or under graduate in 2010 and
01:09:57
that's come down to 8.6 million at the
01:09:59
same time obviously in the last 12 years
01:10:01
you've had uh the population has grown
01:10:03
so this is even you know if it was a
01:10:05
percentage basis would be even more
01:10:08
dramatic so what's behind this a poll of
01:10:11
a000 teens this summer found that about
01:10:13
half believe a high school degree trade
01:10:15
program or two-year degree best meets
01:10:18
their career needs and 56% said Real
01:10:20
World on the job experience is more
01:10:22
valuable than obtaining a college degree
01:10:24
something you've talked about with your
01:10:26
own personal experience jamath at waterl
01:10:28
doing apprenticeships essentially your
01:10:31
thoughts on generation tool belt such a
01:10:33
positive trend I mean there's so many
01:10:37
reasons why this is good I'll I'll just
01:10:39
list a handful that come to the top of
01:10:41
my mind the first and probably the most
01:10:43
important is that it breaks this
01:10:46
strangle hold that the university
01:10:49
education system has on America's kids
01:10:54
we have
01:10:56
tricked millions and millions of people
01:11:00
into getting trillions of dollars in
01:11:02
debt on this idea that you're learning
01:11:04
something in University that's somehow
01:11:06
going to give you
01:11:09
economic stability and ideally freedom
01:11:13
and it has turned out for so many people
01:11:15
to not be true it's just so absurd and
01:11:18
unfair that that has happened so if you
01:11:21
can go and get a trade degree and live a
01:11:24
econom ially productive life where you
01:11:26
can get married and have kids and take
01:11:28
care of your family and do all the
01:11:29
things you want to do that's going to
01:11:30
put an enormous amount of pressure on
01:11:33
higher ed why does it charge so much
01:11:36
what does it give in return that's one
01:11:38
thought the second thought which is much
01:11:40
more narrow Peter teal has that famous
01:11:42
saying where if you have to put the word
01:11:45
science behind it it's not really a
01:11:47
thing and what we are going to find out
01:11:50
is that that was true for a whole bunch
01:11:52
of things where people went to school
01:11:54
like political science
01:11:55
social science yeah social science but I
01:11:58
always thought that computer science
01:12:00
would be immune but I think he's going
01:12:02
to be right about that too because you
01:12:04
can spend two or $300,000 getting in
01:12:07
debt to get a computer science degree
01:12:09
but you're probably better off learning
01:12:10
JavaScript and learning these tools in
01:12:12
some kind of a boot camp for far far
01:12:14
less and graduating in a position to
01:12:16
make money right away so those are just
01:12:18
two ideas I think that it allows us to
01:12:21
be a better functioning Society so I am
01:12:26
really supportive of this trend sax your
01:12:28
thoughts on this generation tool belt
01:12:30
we're we're reading about and you know
01:12:33
the sort of combination with static team
01:12:36
size that we're seeing in technology
01:12:38
companies keeping the number of
01:12:40
employees the same or trending down
01:12:42
while they grow 30% year-over-year
01:12:44
oh my God I'm like so sick of this this
01:12:46
topic of of job loss or job disruption
01:12:50
you I got in so much trouble last week
01:12:51
you asked a question about whether the
01:12:53
upper middle class is going to suffer
01:12:55
because they're all going to be put out
01:12:56
of work by Ai and I just kind of brush
01:12:58
it off not because I'm advocating for
01:13:01
that but just because I don't think it's
01:13:02
going to happen this whole thing about
01:13:04
job loss is so overdone there's going to
01:13:07
be a lot of job disruption but in the
01:13:09
case of coders just as an example I
01:13:12
think we can say that coders depending
01:13:14
on who you talk to are 10 20 30% more
01:13:17
productive as a result of these coding
01:13:19
assistant tools but we still need coders
01:13:22
you can't automate 100% of it and the
01:13:24
world needs needs so many of them the
01:13:26
need for software is unlimited we can't
01:13:29
hire enough of them at Glue by the way
01:13:31
shout out if you're looking if here a
01:13:33
coder who is afraid of not being able to
01:13:34
get a job apply for one at glue believe
01:13:36
me we're hiring I just think that this
01:13:38
is so overdone there's going to be a lot
01:13:41
of disruption in the knowledge worker
01:13:44
space like we talked about the workflow
01:13:46
at call centers and to service factories
01:13:50
there's going to be a lot of change but
01:13:52
at the end of the day I think there's
01:13:54
going to be plenty of work for humans to
01:13:56
do and some of the work will be more in
01:13:58
the blue collar space and I agree with
01:14:00
jamath that this is a good thing I think
01:14:02
there's been perhaps an overemphasis on
01:14:05
the idea that the only way to get ahead
01:14:07
in life is to get uh like a fancy degree
01:14:10
from one of these universities and we've
01:14:12
seen that many of the universities
01:14:14
they're just not that great they're
01:14:16
overpriced you end up graduating with a
01:14:18
mountain of debt and you get a degree
01:14:20
that is you know maybe even far worse
01:14:23
than computer science that's just
01:14:24
completely worthless so if people learn
01:14:27
more vocational skills if they skip
01:14:30
College because they have a proclivity
01:14:32
to do something that doesn't need that
01:14:33
degree I think that's a good thing and
01:14:35
that's healthy for the economy freedberg
01:14:36
is this like uh just the pendulum swung
01:14:38
too much and education got too expensive
01:14:40
spending 200k to make $50,000 a year
01:14:43
distinctly different than our childhoods
01:14:45
or I'm sorry our adolescence when we
01:14:46
were able to go to college for 10K a
01:14:48
year 20K a year graduate with you know
01:14:51
some low tens of thousands in debt if
01:14:53
you did take debt and then you're your
01:14:55
entry level job was 5050 60 70k coming
01:14:57
out of college what are your thoughts
01:14:58
here is this a value issue with
01:15:00
college well yeah I think the Market's
01:15:02
definitely correcting itself I think for
01:15:04
years as chamat said there was kind of
01:15:05
this belief that if you went to college
01:15:07
there was regardless of the college
01:15:10
there was this outcome where you would
01:15:12
make enough money to justify the debt
01:15:15
you're taking
01:15:16
on and I think folks have woken up to
01:15:19
the fact that that's not reality again
01:15:22
if there was a free market remember most
01:15:24
people go to college with student loans
01:15:27
and all student loans are funded by the
01:15:29
federal government so the the cost of
01:15:32
education has ballooned and the
01:15:34
underwriting criteria necessary for this
01:15:36
free market to work has been completely
01:15:39
destroyed because of the federal
01:15:41
spending in the student loan program
01:15:44
there's no discri discrimination between
01:15:46
one school or another you could go to
01:15:48
Trump University or you could go to
01:15:51
Harvard it doesn't matter you still get
01:15:52
a student loan even if at the end of the
01:15:55
process you don't have a a degree that's
01:15:57
valuable and so I think folks are now
01:15:59
waking up to this fact and the market is
01:16:01
correcting itself which is good I'll
01:16:03
also say that I think that there's this
01:16:06
premium with generally mass production
01:16:10
and
01:16:11
industrialization of the human touch and
01:16:15
what I mean is if you think about hey
01:16:18
you could go to the store and buy a
01:16:20
bunch of cheap food off of the store
01:16:22
shelves you could buy a bunch of
01:16:23
Hershey's chocolate bars
01:16:25
or you can go to a Swiss choler in
01:16:27
downtown San Francisco pay $20 for a box
01:16:31
of Handmade Chocolates you'll pay that
01:16:32
premium for that better product same
01:16:35
with clothes there's this big Trend in
01:16:36
kind of handmade clothes and high-end
01:16:38
luxury goods spoke artisanal Aral
01:16:43
handmade and and similarly I think that
01:16:46
there's a premium in Human Service in
01:16:50
the partnership with a human it's not
01:16:51
just about blue collar jobs it's about
01:16:54
having a
01:16:55
talk to you and serve you if you go to a
01:16:58
restaurant instead of having a machine
01:17:00
spit out the food to you there's an
01:17:02
experience associated with that that
01:17:04
you'll pay a premium for there's
01:17:06
hundreds and hundreds of micro breweries
01:17:08
in the United States that in aggregate
01:17:09
out sell Budweiser and Miller and even
01:17:12
Modello today and that's because they're
01:17:14
handcrafted by local people and there's
01:17:16
a there's an artisan craftsmanship so
01:17:19
while technology and AI are going to
01:17:21
completely reduce the cost of a lot of
01:17:24
things and increase the production and
01:17:26
productivity of those things one of the
01:17:29
complimentary consequences of that is
01:17:31
that there will be an emerging premium
01:17:33
for Human Service and I think that there
01:17:35
will be an absolute burgeoning and
01:17:37
blossoming in the salaries and the
01:17:40
availability and demand for Human
01:17:42
Service in a lot of walks of life
01:17:44
certainly there's all the work at home
01:17:46
the electricians and the plumbers and so
01:17:48
on but also Fitness classes
01:17:51
food personal service around tutoring
01:17:54
and learning and developing oneself
01:17:56
there's going to be an incredible
01:17:57
blossoming I think in human service jobs
01:17:58
and they don't need to have a degree in
01:18:00
poai to be performed I think that there
01:18:03
will be a lot of people that would be
01:18:04
very happy in that world how do you see
01:18:05
the differentiation the person makes
01:18:08
free ber in doing that job versus the
01:18:10
agent or the AI or whatever well these
01:18:12
are iners human jobs so if I want to do
01:18:15
a fitness class do I want to stare at
01:18:16
the tonal or this what I'm this is what
01:18:18
I'm asking you yeah like what I think I
01:18:21
I think that there's an aspect of um it
01:18:23
look it's like your Laura like you you
01:18:26
talk about the story of Laura piano
01:18:29
where is theun coming from how's it made
01:18:31
Who's involved in it like yes look God
01:18:36
go don't stop Reber I could give you
01:18:39
truffle flavoring out of a can but you
01:18:41
you love the white truffles you want to
01:18:42
go to Italy you want the storytelling
01:18:44
there's an aspect of it right like yeah
01:18:46
and I think that there's an aspect of
01:18:48
humanity that we pay a premium that we
01:18:51
do and will look Etsy crushes I don't
01:18:52
know how much stuff you guys buy an Etsy
01:18:54
I love buying from Etsy I love finding
01:18:56
handmade stuff on Etsy for no you don't
01:18:59
do you really yes I do yeah hand crating
01:19:01
I yeah handmade so I think that there's
01:19:03
an aspect of this that um that in a lot
01:19:06
of walks of life I mean I have so many
01:19:07
jokes right now
01:19:09
chewing what I've never used that se but
01:19:12
I'm gon to try it now after this have
01:19:14
you guys taken music lessons lately you
01:19:16
know I start my kids do piano lessons
01:19:17
and so last year I started ducking in to
01:19:19
do a 45-minute piano lesson with the
01:19:21
piano teacher there's just like a great
01:19:23
aspect to paying for for these services
01:19:25
to getting fascinating bring oh here we
01:19:29
go you can play the harmonica really
01:19:32
I've just playing some I want to play
01:19:34
some Zack Bryan songs and he's got a
01:19:36
couple songs I like with a harmonica in
01:19:37
them so I just got harmonica my daughter
01:19:39
and I have been playing harmonica yeah
01:19:42
are you teaching yourself let's hear it
01:19:43
let's hear it let's hear it I'll play it
01:19:45
next week I'm deep in the laboratory not
01:19:48
a bit it could be a bit but could be a
01:19:50
I'll write I'll write your song next
01:19:52
week the uh a little shy he's a little
01:19:54
shy no no I'll do I'll I'll write a
01:19:56
trump song for you I'll do the uh the
01:19:58
the trials and tribulations of Donald
01:20:01
Trump and I'll I'll do a little Bob
01:20:02
Dylan send up song
01:20:05
for you see that interview with um with
01:20:08
Bob Dylan I don't know when it was
01:20:10
recently about how andc that clip
01:20:15
Bradley amazing Bradley clip about magic
01:20:18
yeah some of those songs I don't know
01:20:20
how I wrote them they just came out and
01:20:25
do that anymore he's like no but I did
01:20:27
it once no but I did it once and what an
01:20:30
incredible that means something yeah
01:20:34
eclipes ground it's really grounding you
01:20:37
understand too soon there is no chance
01:20:39
of ding yeah that's an incredible
01:20:42
clip guys want to wrap or you want to
01:20:43
keep talking about more stuff we we're
01:20:45
at 90 minutes let me just say tell you
01:20:47
something I think there's G to be a big
01:20:48
war I think by the time the show airs
01:20:51
Israel's incursion into Lebanon is going
01:20:52
to get bigger it's going to escalate
01:20:55
and by next week we could be in a
01:20:57
full-blown multinational war in the
01:21:00
Middle East and if I am you know a
01:21:04
betting man I would bet that the odds
01:21:05
are you know more than 30 40% that this
01:21:08
happens before the election that this
01:21:10
this conflict in the Middle East
01:21:11
escalates thank you for bringing this up
01:21:14
I
01:21:15
am not asking anybody to go listen to
01:21:18
what I my interview with Rogan uh but I
01:21:22
will say this part of why I was so
01:21:27
excited to go and talk to him in a long
01:21:29
form format was this issue of war is I
01:21:33
think the existential crisis of this
01:21:37
election and of this moment and I really
01:21:40
do agree with you freeberg there is a
01:21:43
non-trivially high probability the
01:21:45
highest it's ever been that we are just
01:21:48
bumbling and sleepwalking into a really
01:21:51
bad situation we can't walk back from I
01:21:54
really hope you're wrong and here's the
01:21:57
here's the situation I really hope
01:21:59
you're wrong if Israel incurs further in
01:22:01
into Lebanon going after Hezbollah and
01:22:05
Iran ends up getting involved in a more
01:22:08
active way does Russia start to provide
01:22:11
supplies to Iran like we are supplying
01:22:14
to Ukraine today does this sort of bring
01:22:17
everyone to a line just to give you a
01:22:19
sense you know of the scale of what
01:22:21
Israel could then respond with Iran has
01:22:24
600 ,000 active udy military another
01:22:26
350,000 in reserve they have dozens of
01:22:29
ships they have 19 submarines they have
01:22:31
a 600 km range missile system Israel has
01:22:35
90,000 has sorry 170,000 active duty and
01:22:38
half a million Reserve Personnel 15
01:22:39
warships five submarines potentially up
01:22:42
to 400 nuclear weapons including a very
01:22:45
wide range of tactical sub one kiloton
01:22:48
nuclear weapons small small payload you
01:22:51
could see that if Israel starts to feel
01:22:55
incurred upon further they could respond
01:22:59
in a more aggressive way with what is
01:23:01
you know by far in
01:23:03
away you know the the most significantly
01:23:06
stock Arsenal and military force in the
01:23:09
Middle East again we've talked about
01:23:12
what are these other countries going to
01:23:13
do what is Jordan going to do in this
01:23:15
situation how is how the Saud going to
01:23:18
respond what is Russia going to do
01:23:20
Russia well the Russia Ukraine thing
01:23:22
meanwhile still goes on and we saw in
01:23:24
our group chat one of our friends posted
01:23:27
but Russia basically said any more
01:23:29
attacks on our land you know we reserve
01:23:31
All rights including nuclear response
01:23:34
that is insane well you know so just to
01:23:36
give you a sense um how how are we here
01:23:40
yeah so the the nuclear bombs that were
01:23:43
set off during World War II I just want
01:23:46
to show you how crazy this this is do
01:23:50
you see that image on the
01:23:52
left that all all the way over on the
01:23:55
left that's a bunker Buster you guys
01:23:56
remember those from Afghanistan and and
01:23:58
the Damage that those bunker Buster
01:24:00
bombs caused Hiroshima is a 15 kiloton
01:24:04
nuclear and you can see the size of it
01:24:06
there on the left that's a zoom in of
01:24:09
the image on the right and the image on
01:24:12
the right starts to show the biggest
01:24:14
ever tested was Zar bomba by the Soviets
01:24:18
this was a 50 Megaton bomb it it caused
01:24:21
shock waves that went around the Earth
01:24:25
three times they could be felt as
01:24:26
seismic shock waves around the earth
01:24:28
three times from this one detonation
01:24:30
today there are a lot of .1 to1 kiloton
01:24:35
nuclear bombs that are kind of
01:24:36
considered these tactical nuclear
01:24:38
weapons that kind of fall closer to
01:24:41
between the bunker Buster and the
01:24:43
Hiroshima and that's really where a lot
01:24:45
of folks get concerned that if Israel or
01:24:47
Russia or others get cornered in a way
01:24:51
and there's no other tactical response
01:24:53
that that is what that get gets pulled
01:24:55
out now if someone detonates A.1 or 1
01:24:58
kiloton nuclear bomb which is going to
01:25:00
look like a mega bunker Buster what is
01:25:02
the other side and what's the world
01:25:03
going to respond with that's how on the
01:25:06
brink we are and there's
01:25:08
12,000 nuclear weapons with an average
01:25:10
payload of 100
01:25:12
kilotons around the world the US has a
01:25:15
large stockpile Russia has the largest
01:25:18
many of these are hair trigger alert
01:25:20
systems China has the third largest and
01:25:22
then Israel and India and and so on it
01:25:27
is a very concerning situation because
01:25:29
if anyone does get pushed to the brink
01:25:31
that has a nuclear weapon and they pull
01:25:33
out a tactical nuke does that mean that
01:25:35
game is on and that's why I'm so nervous
01:25:37
about where this all leads to if we
01:25:38
can't decelerate it's very scary because
01:25:40
you could very quickly see this thing
01:25:42
Accel I am the most objectively scared
01:25:44
I've ever
01:25:45
been and I think that people grossly
01:25:49
underestimate how quickly this could
01:25:51
just spin up out of control and right
01:25:53
now
01:25:55
not enough of us are spending the time
01:25:57
to really understand why that's possible
01:26:00
and then also try to figure out what's
01:26:01
the offering and I think it's I think
01:26:04
it's just incredibly important that
01:26:06
people take the time to figure out that
01:26:08
this is a nonzero probability and this
01:26:11
is probably for many of us the First
01:26:12
Time In Our Lifetime where you could
01:26:14
really say that well I think freeberg is
01:26:16
right that we're at the beginning stages
01:26:18
of I think what will soon be referred to
01:26:20
as the third Lebanon war the first one
01:26:23
was in
01:26:24
1982 uh Israel went into Lebanon and
01:26:27
occupied it until 2000 then it went back
01:26:30
in 2006 left after about a month and now
01:26:33
we're in the third war it's hard to say
01:26:36
exactly how much this will escalate the
01:26:40
IDF is exhausted after the war in Gaza
01:26:44
there's significant opposition within
01:26:47
Israel and within the Armed Forces to a
01:26:49
big ground invasion of Lebanon so far
01:26:53
most of the fighting has has been Israel
01:26:55
using its air superiority overwhelming
01:26:58
Firepower against Southern Lebanon and I
01:27:02
I I think that if Israel makes a ground
01:27:05
Invasion they're giving hisbah the war
01:27:08
that hisbah wants I mean hisb would love
01:27:11
for this to turn into a Guerilla war in
01:27:13
southern Lebanon so I think there's
01:27:14
still some question about whether
01:27:17
Netanyahu will do that or not at the
01:27:20
same time it's also possible that his
01:27:23
Bella will
01:27:24
attack Northern Israel what nzala has
01:27:28
threatened to invade the
01:27:32
Galilee in response to what Israel is
01:27:34
doing so there's multiple ways this
01:27:36
could escalate and if Hezbollah and
01:27:41
Israel are in a fullscale war with
01:27:44
ground forces it could be very easy for
01:27:46
Iran to get pulled into it on hisb side
01:27:49
and if that happens I think it's just
01:27:51
inevitable that the United States will
01:27:53
be pulled into this war so yeah look I
01:27:55
think we are drifting and and we have
01:27:58
been drifting into a regional war in the
01:28:01
Middle East that you know ideally would
01:28:04
not pull in the US I I think the US
01:28:05
should try to avoid being pulled in but
01:28:08
I think very likely will be pulled in if
01:28:10
it escalates and then meanwhile in terms
01:28:14
of the war in Ukraine I mean I've been
01:28:15
warning about this for two and a half
01:28:17
years how dangerous the situation was
01:28:19
and that's why we should have availed
01:28:22
ourselves of every diplomatic
01:28:23
opportunity to make peace and we now
01:28:26
know because there's been such Universal
01:28:28
reporting that in Istanbul in the first
01:28:31
month of the Ukraine war there was an
01:28:33
opportunity to make a deal with Russia
01:28:35
where Ukraine would get all of its
01:28:37
territory back is just that Ukraine
01:28:39
would have to agree not to be part of
01:28:40
NATO it would have to agree to be
01:28:42
neutral and not part of the western
01:28:44
military block that was so threatening
01:28:46
to Russia the Biden Administration
01:28:48
refused to make that deal they sent in
01:28:50
Boris Johnson to Scuttle it they threw
01:28:51
cold water on it they blocked it they
01:28:53
told zinsky will give you all the
01:28:55
weapons you need to fight Russia
01:28:57
zalinsky believed in that it has not
01:28:59
worked out that way Ukraine is getting
01:29:02
destroyed it's very hard to get honest
01:29:04
reporting on this from the mainstream
01:29:06
media but the sources I've read suggest
01:29:10
that the ukrainians are losing about
01:29:12
30,000 troops per month and that's just
01:29:15
Kia I don't even think that's wounded
01:29:17
that on a bad day they're suffering
01:29:20
1,200 casualties it's more than even
01:29:23
during that counter offensive last
01:29:25
summer that Ukraine had during that time
01:29:27
they were losing about 20,000 troops a
01:29:28
month so the level of Carnage is
01:29:31
escalating Russia has has more of
01:29:35
everything more weapons more Firepower
01:29:38
air superiority and they are destroying
01:29:40
Ukraine and it's very clear I think that
01:29:44
Ukraine within it could be in the next
01:29:46
month it could be in the next two months
01:29:47
it could be in the next six months I
01:29:48
think they're eventually going to
01:29:50
collapse they're getting close to being
01:29:52
combat incapable and in a way that poses
01:29:54
the biggest danger
01:29:56
because the closer Ukraine gets to
01:29:59
collapse the more the West is going to
01:30:00
be tempted to to intervene directly in
01:30:02
order to save them and that is what
01:30:05
zalinski was here in the US doing over
01:30:07
the past week is arguing for direct
01:30:10
involvement by America in the Ukraine
01:30:13
war to save him how did he propose this
01:30:15
he said we want to be directly admitted
01:30:17
to Nato immediately that was his request
01:30:21
and he called this the the victory plan
01:30:23
so in other words his plan for victory
01:30:25
is to get America involved in the war
01:30:26
and fighting it for him but that is the
01:30:29
only chance Ukraine has and it is
01:30:31
possible that the Biden Harris
01:30:33
Administration will agree to do that or
01:30:35
at least agree to some significant
01:30:37
escalation so far I think Biden to his
01:30:39
credit has resisted another zalinsky
01:30:42
demand which is the ability to use
01:30:44
America's long-range missiles and
01:30:47
British long- range missiles the storm
01:30:49
Shadows against Russian cities that is
01:30:51
what zalinski is asking for if zalinsky
01:30:52
wants a major escal of the war because
01:30:54
that is the only thing that's going to
01:30:56
save him save his side and maybe even
01:30:59
his neck personally and so we're one
01:31:02
mistake away from the very dangerous
01:31:05
situation that chth and freeberg have
01:31:07
described if a President Biden who is
01:31:11
basically scile or a president Harris
01:31:14
agree to one of these zalinsky requests
01:31:17
we could very easily find ourselves in a
01:31:19
direct war with the Russians the Walts
01:31:22
into World War II is what it should be
01:31:25
called and the reason why this could
01:31:26
happen is because we don't have a fair
01:31:28
media that's fairly reported anything
01:31:30
about this War I mean Trump is on the
01:31:32
campaign Trail making I think very valid
01:31:35
points about this war that the Ukrainian
01:31:37
cause is doomed and that we should be
01:31:39
seeking a peace deal and a settlement
01:31:42
before this cons spiral into World War
01:31:44
II that is fundamentally correct but the
01:31:46
media portrays that as being pro-russian
01:31:49
and pro Putin and if you say that you
01:31:50
want peace you are basically on the take
01:31:53
from Putin in Russia that is what the
01:31:54
media has told the American public for 3
01:31:56
years the definition of Liber liberalism
01:31:59
has always been being completely against
01:32:02
war of any kind and being completely in
01:32:05
favor of free speech of all kinds that's
01:32:08
what being a liberal means we've lost
01:32:11
the
01:32:12
script and I think that people need to
01:32:14
understand that this is the one issue
01:32:17
where if we get it wrong literally
01:32:19
nothing else matters and we are
01:32:22
sleepwalking and tiptoeing into a
01:32:27
potential massive World War Jeffrey sck
01:32:30
said it perfectly you don't get a second
01:32:32
chance in the nuclear age all it takes
01:32:34
is one big mistake you do not get a
01:32:36
second chance and for
01:32:40
me I have become a single issue voter
01:32:43
this is the only issue to me that
01:32:45
matters we can sort everything else out
01:32:47
we can figure it all out we can we can
01:32:50
find common ground and reason should
01:32:53
taxes go up should taxes go down let's
01:32:56
figure it out should regulations go up
01:32:58
should regulations go down we can figure
01:33:00
it
01:33:01
out but we are fighting a potential
01:33:05
nuclear threat on three fronts how have
01:33:08
we allowed this to happen Russia Iran
01:33:12
China you
01:33:14
cannot
01:33:15
underestimate that when you add these
01:33:17
kinds of risks on top of each other
01:33:20
something can happen here and I don't
01:33:23
think people know they're too far away
01:33:24
from it they're too many generations
01:33:26
removed from it war is something you
01:33:28
heard maybe your grandparents talk about
01:33:30
now and you just thought okay
01:33:33
whatever I lived it it's not
01:33:36
good jth you're right I mean during the
01:33:38
Cuban Missile Crisis all of America was
01:33:40
huddled around their TV sets worried
01:33:43
about what would happen there is no
01:33:45
similar concern in this day and age
01:33:46
about the escalatory wars that are
01:33:50
happening there's a little bit of
01:33:51
concern I think about what's happening
01:33:52
in the Middle East there's virtually no
01:33:53
concern about what's happening in
01:33:55
Ukraine because people think it can't
01:33:57
affect them but it can and one of the
01:34:00
reasons it could affect them is because
01:34:01
we do not have a fair debate about that
01:34:04
issue in the US media the media has
01:34:06
simply caricatured any opposition to the
01:34:10
war as being protin so I would say that
01:34:13
when every pundit and
01:34:17
every person in a position to do
01:34:21
something about it says you have nothing
01:34:24
to worry
01:34:25
about you probably have something to
01:34:27
worry
01:34:28
about and so when everybody is trying to
01:34:31
tell you everybody that this is not a
01:34:34
risk it's probably a bigger risk than we
01:34:38
think yeah they're protesting too much
01:34:41
how can you say it's not a risk me think
01:34:43
thou do
01:34:44
protested all right love you boys
01:34:48
[Applause]
01:34:49
byebye let your winners ride
01:34:53
Man
01:34:56
David and instead we open source it to
01:34:59
the fans and they've just gone crazy
01:35:01
with it love queen
01:35:04
[Music]
01:35:09
of Besties
01:35:12
are that's my dog taking your
01:35:17
driveways oh man myit will meet me we
01:35:21
should all just get a room and just have
01:35:22
one big huge or because they're all this
01:35:24
usess it's like this like sexual tension
01:35:26
that they just need to release
01:35:33
somehow we need to get mer
01:35:39
[Music]
01:35:43
our going
01:35:45
[Music]
