00:00:00
all right freeberg is back welcome back
00:00:02
to the Allin podcast episode 160
00:00:04
something your favorite podcast in the
00:00:05
world yada yada yada with me again the
00:00:08
chairman dictator moth po hoaa the
00:00:10
Rainman yeah definitely David Sachs is
00:00:13
here and back from his time in the
00:00:17
metaverse we found him somewhere out in
00:00:19
space in the solar system in his Apple
00:00:22
goggles your favorite sultant of science
00:00:26
David fredberg is back from the
00:00:28
metaverse I miss you guys come home
00:00:31
thanks for having me what what did you
00:00:33
discover when you went to Uranus in
00:00:35
Google class sorry Apple you actually
00:00:38
use the Apple Vision Pro jcal I ordered
00:00:41
them I ordered them and I walked by the
00:00:42
Apple Store and I was going to go in and
00:00:44
try them and there were so many lunatics
00:00:46
in there I was like yeah I'm not doing
00:00:48
it but I ordered them you use you
00:00:49
actually used them what I ordered one
00:00:51
online to be delivered and it was like
00:00:53
delayed by a month so I went down to the
00:00:55
Apple Store and picked one up okay and
00:00:58
my kids cannot stop using it really I
00:01:01
went down to the Apple Store but got
00:01:02
cleaned out by the thief that stole
00:01:04
everything so the Oakland
00:01:08
one let your winners
00:01:10
[Music]
00:01:15
ride and instead we open source it to
00:01:18
the fans and they've just gone crazy
00:01:20
with
00:01:20
[Music]
00:01:23
it that was crazy that was crazy we'll
00:01:26
put the video in here to the idiots who
00:01:28
are robbing apples SCE all the devices
00:01:30
get bricked when you steal them and they
00:01:31
all have GPS in them have you tried it
00:01:34
your mouth no I was too busy working out
00:01:36
making love and winning oh okay got it
00:01:39
so you were you were making sweet love
00:01:42
you were watching your portfolio go up
00:01:44
and you were just generally winning got
00:01:46
it got it yeah yeah so freeberg the rest
00:01:48
of us were being men in the world
00:01:51
accomplishing stuff but but do tell us
00:01:53
about your time in the metaverse do
00:01:54
those goggles come with a lifetime
00:01:57
prescription of ssris you sound like one
00:02:00
of these like Tech journalists that are
00:02:02
actually anti-tech people you guys are
00:02:04
actually Tech journalists like it tech
00:02:06
journalist seem gen Computing platform I
00:02:09
remember when the iPad came out and
00:02:10
everyone poo pooed the iPad I thought it
00:02:12
was stupid I tried to use it I couldn't
00:02:13
get any value out of it and in 2010 or
00:02:16
2011 when did it come out 2010 2011 we
00:02:18
started using it with our sales team
00:02:20
selling to Farmers and we gave every
00:02:22
sales guy an iPad and they went out in
00:02:24
the field with 3G and they were able to
00:02:27
close sales in the field meeting with
00:02:29
Farmers which had never been done before
00:02:30
you usually had to get a farmer to come
00:02:31
into an office how many IP to sell the
00:02:34
product Oh so we we had like they were
00:02:37
selling cl.com software we had we had
00:02:39
dozens of these sales guys we gave them
00:02:40
out to our sales agents as well the
00:02:42
independent agents they started using
00:02:43
them and it was like a real GameChanger
00:02:45
in how sales was done in agriculture and
00:02:48
I had never even contemplated that when
00:02:49
I first used the iPad let's let's get to
00:02:51
Brass tax here what is the killer app
00:02:53
what do you think in the next five years
00:02:55
people are going to be doing with this
00:02:57
thing on a daily basis is there a daily
00:02:59
use case I'll say a couple things one is
00:03:01
like I feel the same way I did about the
00:03:02
iPad which is I don't know what it is
00:03:04
today but I can tell that there's
00:03:05
something there and I'll give you an
00:03:06
example of something I thought about
00:03:08
first of all the AR is gamechanging okay
00:03:10
if you've used like The Meta yeah the
00:03:12
Oculus Quest it like makes me super
00:03:15
dizzy makes my head hurt makes my eyes
00:03:17
hurt like you're super disoriented what
00:03:19
Apple solved is that you're like still
00:03:21
in reality but then you get to interact
00:03:24
with these three-dimensional kind of
00:03:25
objects in reality and it's like really
00:03:27
well done it's definitely V1 and there's
00:03:30
going to be incredible changes in the
00:03:31
next couple Generations but it gets rid
00:03:33
of all that dizziness disconnected kind
00:03:35
of stuff that happens with the the full
00:03:37
VR experience which I thought was really
00:03:39
incredible then last week and I'm sorry
00:03:42
I missed the show we have a facility
00:03:44
with my company in North Carolina we
00:03:46
have this giant Greenhouse facility and
00:03:47
I was doing meetings with farmers and
00:03:48
stuff I go to the greenhouse facility
00:03:51
and there's so much work that the
00:03:52
greenhouse text and lab teexs are doing
00:03:56
where they're using an iPhone and a
00:03:57
barcode scanner and a printer and
00:03:59
they're holding all these pieces of
00:04:01
equipment scanning the QR codes on
00:04:03
flowers taking the pollen out putting it
00:04:05
in the next flower training each other
00:04:07
how to do it and I was like I put this
00:04:09
apple Vision Pro on and I was like man
00:04:11
all the the apps and all the tools that
00:04:13
we had all these different pieces for
00:04:15
that was taking people tons of time
00:04:17
image collection data collection could
00:04:19
all just be done streamlined while
00:04:21
you're working you could have a report
00:04:24
yeah yeah you have a task on the right
00:04:26
cameras are taking images in the middle
00:04:28
QR codes are automatically scanned data
00:04:30
is being ingested the task list is kind
00:04:32
of you know giving folks next steps they
00:04:35
can listen to music while they're
00:04:36
working and I realized for that job and
00:04:38
I met with all the the team out there
00:04:39
and spent time with them and I actually
00:04:41
did the work that they do to get a
00:04:43
better sense for the workflow and I was
00:04:44
like man literally every aspect of this
00:04:46
job will be massively improved and
00:04:48
productivity will go up by 10x with
00:04:50
these goggles will it happen in the next
00:04:51
couple weeks or months I don't know but
00:04:53
my engineering team is looking into it
00:04:54
can we take it can we use some software
00:04:56
can we build some software and can we
00:04:58
put this on folks to give them a better
00:05:00
work experience to increase our
00:05:02
productivity to do automated data
00:05:03
capture so I don't know exactly where it
00:05:05
goes but I could start to see how this
00:05:07
can become a more ubiquitous part of a
00:05:09
Workforce setting and not just be a
00:05:10
video game and movie tool for consumers
00:05:13
so I'm I'm reasonably optimistic about
00:05:15
where this goes it's definitely V1 I
00:05:16
feel like it's the iPad days where no
00:05:18
one's really sure where the applications
00:05:20
are but yeah yeah Enterprise
00:05:21
applications unbelievable makes total
00:05:23
sense and also training training right
00:05:25
assembly line Workforce warehouse
00:05:28
workers where you're real time kind of
00:05:30
task updates data's being ingested all
00:05:32
in real time and and by the way the
00:05:34
other thing I'll say is training is
00:05:35
incredible there's spatial video
00:05:37
recording on it so it looks like you're
00:05:39
living through the experience that
00:05:40
someone else had so you can train
00:05:42
someone how to do a difficult task and
00:05:44
rather than have a human go spend hours
00:05:46
training a Workforce the workforce can
00:05:48
be trained by the goggles in a way that
00:05:50
you cannot do a two-dimensional video
00:05:51
today so I don't know I'm I'm I'm pretty
00:05:53
optimistic very strange days right I
00:05:55
don't know you're you're a fan of SciFi
00:05:57
but remember Strange Days tot trath
00:05:59
what's going to happen first here are
00:06:02
humans going to become more like robots
00:06:04
by putting these on and do this fact
00:06:05
reer work or is Elon with Optimus and
00:06:08
some of Humane I think is the other one
00:06:09
there's a couple of other people
00:06:10
building General use robots figure is
00:06:13
the other one figure yeah which one wins
00:06:15
the day is it going to be humans having
00:06:18
eyes and you know data collection like
00:06:21
robots or robots having appendages like
00:06:23
humans well let me let me put two ideas
00:06:26
together and see what you think of this
00:06:29
argument
00:06:30
if you think about the generation of
00:06:32
human beings that have as close
00:06:36
to any other generation before it lived
00:06:39
in a totally immersive world I would say
00:06:42
the best representation of that are
00:06:44
current teenagers and 20-year-old people
00:06:48
and maybe at the upper Edge the early
00:06:50
30s
00:06:51
people and why is that you know they've
00:06:54
lived inside of social media their
00:06:55
entire lives they've lived inside of
00:06:56
immersive video games their entire lives
00:06:58
but the question is is are they better
00:07:01
off and happier as far as we know from
00:07:03
an evolutionary perspective and I would
00:07:05
tell you that the answer is a is a huge
00:07:08
gaping no so if you believe that the
00:07:15
rise in depression the rise in suicide
00:07:19
the dependency on drugs the dependency
00:07:21
on ssris the sexual promiscuity the lack
00:07:24
of marriage the lack of kids if all of
00:07:27
those things are in some way
00:07:31
a correlated byproduct let's not say
00:07:33
it's causal right let's just say it's a
00:07:34
correlated byproduct of this entire
00:07:38
immersive almost exclusionary detached
00:07:41
world that these folks have grown up in
00:07:44
taking that to the
00:07:45
Limit I'm just going to put out there
00:07:47
may not be the solution to our problems
00:07:50
and so I guess the more directed answer
00:07:52
to your question is I would hope that
00:07:54
the ladder wins so that we take these
00:07:56
goggles off and actually learn how to
00:07:58
talk to each other and look each other
00:07:59
in the eyes get married and have
00:08:01
children because I think that's actually
00:08:03
better for the
00:08:05
world and I would probably say that it's
00:08:07
almost better for the world than a 10
00:08:09
Xing of
00:08:10
productivity interesting and then you
00:08:12
see the correlation to cancer and
00:08:14
disease that is disproportionately
00:08:16
higher amongst these young people so I
00:08:18
think it's at some point to ask
00:08:20
ourselves what is structurally happening
00:08:23
in the lives of these 16 you know 15 to
00:08:26
31 year olds that is just so in terms of
00:08:30
outcomes and if you look at some of the
00:08:33
environmental variables that they live
00:08:34
in and then take some of those and take
00:08:36
them to the Limit I think that there's a
00:08:38
reasonable argument that their lives get
00:08:40
worse before it gets better yeah I mean
00:08:42
the amount of time you spend on social
00:08:44
media is correlated uh with depression
00:08:46
not just social media I'm just saying
00:08:47
just this immersive like I'm going to
00:08:49
detach from the world and live through a
00:08:52
microphone and glasses taken to the
00:08:54
Limit I'm not sure is the solution to
00:08:57
these kids feeling detached lonely
00:09:00
isolated isolated yeah yeah I mean it it
00:09:03
correlates all of these things that
00:09:04
we're seeing in this younger generation
00:09:05
correlates with the introduction could
00:09:07
it be a good productivity device yes do
00:09:09
I hope it's a good productivity device
00:09:11
yes but if we try to make it the Panacea
00:09:13
for anything and everything I think
00:09:15
we're going to we're going to compound
00:09:18
the systemic issues that these young
00:09:20
people have and I suspect on the margin
00:09:23
if you were going to bet all of these
00:09:25
things that we see in these young people
00:09:26
today will get worse as a byproduct of
00:09:29
technology not necessarily get better so
00:09:31
if you can take a different path like
00:09:34
Optimus or the figure AI robots where
00:09:37
that work is done at least we have a
00:09:39
different problem probably maybe even
00:09:41
more existential abundance but a
00:09:44
different problem which is now how do
00:09:45
you find purpose but maybe you can find
00:09:46
purpose through connection and the types
00:09:48
of things that humans have been bred
00:09:50
over billions of years to actually
00:09:52
optimize for okay saak I remember when
00:09:57
you were starting craft you fir it up
00:10:00
like a group for VR and you got pretty
00:10:02
heavy into it you made a couple of small
00:10:04
bets I remember I don't think it any of
00:10:06
it worked out really you could tell me
00:10:08
if I'm wrong here but you got in a
00:10:09
little bit early there maybe you could
00:10:10
talk about the business case for this
00:10:12
and has that changed because you you
00:10:14
believed I've believed a lot of folks
00:10:16
thought hey maybe this is the time when
00:10:19
Zuck really start you know had bought
00:10:21
Oculus and and they started putting out
00:10:22
some good product seemed like it was a
00:10:24
false start is this the actual starting
00:10:26
pistol and is this the start of the VR
00:10:28
AR
00:10:30
adoption race I don't think we're quite
00:10:32
there yet okay we've been talking about
00:10:35
VR being a thing for over a decade yeah
00:10:39
no more like 30 remember the Nintendo VR
00:10:42
stuff it's like always on the verge of
00:10:44
happening I think that the big complaint
00:10:45
about the Apple device is has a lot of
00:10:47
capability but it's still a pretty huge
00:10:50
device to wear on your forehead it's
00:10:51
just not really going to be comfortable
00:10:53
enough to be something that people want
00:10:56
to use all the time h i mean there's
00:10:59
also a question of use cases but they're
00:11:01
getting there with the use
00:11:03
cases in any event I I do think that
00:11:06
Apple Vision Pro is it's like I said
00:11:09
last week it's a useful prototype or
00:11:12
proof of concept and it will get better
00:11:15
so I'm glad they did it because I think
00:11:17
you need to start somewhere and then
00:11:19
just keep iterating but eventually for
00:11:21
this to I think really take off you need
00:11:24
to shrink the form factor miniaturize
00:11:28
the technology just every version of it
00:11:30
make it simpler lighter easier to use
00:11:33
yeah I mean eventually it'll feel like
00:11:36
sunglasses and so that is I guess if
00:11:39
they become like regular glasses I think
00:11:40
we all agree it becomes the next I got
00:11:42
to tell you I feel like it's pretty damn
00:11:44
comfortable I don't know if you guys you
00:11:45
guys haven't really used it but that's
00:11:46
what I've heard that's the
00:11:48
surprising online saying it's unlike any
00:11:50
other headset I've ever worn they did an
00:11:52
incredible job designing like does it
00:11:55
feel like ski goggles it doesn't feel
00:11:58
heavy it doesn't feel pre pressure
00:12:00
compared to ski goggles how if you were
00:12:02
wearing ski goggles it's less
00:12:03
constricting than ski goggles it's more
00:12:05
comfortable it like floats on you a
00:12:07
little bit they did a great job with
00:12:08
this cushioning device they built and
00:12:10
the band you put on it feels very
00:12:12
natural it's apple design right it's
00:12:13
like a really well-designed product
00:12:15
that's unlike anything else you've ever
00:12:17
tried I've always felt like when Apple
00:12:19
comes into the race that's the starters
00:12:20
pistol and I think this is it because i'
00:12:22
I've heard the same thing from everybody
00:12:25
you you have to try it it feels like
00:12:27
different than Oculus and some of those
00:12:28
version that came out
00:12:30
previously and they have the app
00:12:33
ecosystem and I I would not discount
00:12:35
that when you know the ability to
00:12:36
monitize the app ecosystem and have all
00:12:38
the people who are already building the
00:12:40
com app the Uber app whatever notion you
00:12:43
know all the stuff that people use and
00:12:45
love Spotify YouTube and then poured it
00:12:48
over here fortnite whatever I think
00:12:50
that's going to be the magic and the
00:12:52
statistics are not lying here I mean
00:12:54
this is unbelievable they've sold
00:12:56
already 200,000 units which doesn't seem
00:12:59
like a lot but for a V1 that is a lot
00:13:01
and they're going to sell a half million
00:13:02
this year it's going to be close to like
00:13:05
that's not that many well it's a couple
00:13:06
of billion meta sells more they do yeah
00:13:09
but you know this is
00:13:11
$4,000 this isn't 500 so to sell that
00:13:14
many of a $4,000 device is incredible
00:13:16
it's a proof of concept it's not like a
00:13:19
regular Apple product that is a mass
00:13:22
Market device that tens or hundreds of
00:13:25
millions of people are going to buy but
00:13:26
it puts them on a path yeah to where
00:13:28
they can iterate and keep making it
00:13:29
better see I think and this is I guess
00:13:32
what I'd ask freeberg do you compare
00:13:34
this to buying a MacBook Pro buying an
00:13:37
iPhone or buying the Oculus you know
00:13:39
whatever the you know $500 unit because
00:13:42
everybody I see talking about online is
00:13:44
comparing it to the purchase of a laptop
00:13:47
because of the desktop and you can kind
00:13:49
of do your coding or surf the web and do
00:13:51
all that where do where do you put this
00:13:52
is it buying a TV is it buying a laptop
00:13:54
is it buying a smartphone what would you
00:13:55
have to have a keyboard to be really
00:13:57
productive on it uhhuh if you're going
00:13:59
to use it for writing purposes or coding
00:14:01
purposes so it doesn't really work with
00:14:03
just the the headset but you could do
00:14:05
that yeah it's definitely like buying a
00:14:07
new Computing device but people felt the
00:14:09
same way about the iPad get go back to
00:14:11
2010 when the iPad came out and everyone
00:14:14
was like who's it for it's a whole new
00:14:16
computer who's it for you already have a
00:14:18
phone you already have a computer why do
00:14:20
you need an iPad and then they sell tens
00:14:22
of millions a quarter now yeah so I
00:14:24
really I as I do the math on this I was
00:14:26
just kind of doing some back of the
00:14:27
envelope stuff I think they're going to
00:14:28
sell 100 bill
00:14:30
of Apple Vision Pros not this version
00:14:32
but this version plus the next version
00:14:34
probably over the
00:14:36
next I would guess for them to get to
00:14:38
100 billion in sales it'll take them
00:14:41
less than five years I think they're
00:14:42
going to run the table on everybody I
00:14:44
think they're going to own the entire
00:14:45
space I think everyone's underestimating
00:14:47
this as a new Computing platform and
00:14:48
once these applications particularly in
00:14:50
the Enterprise setting start to kick in
00:14:52
and I will say that the movie watching
00:14:54
experience is way better than watching
00:14:55
on a TV in your living room my kids
00:14:57
cannot stop asking me to use the goggles
00:14:59
to watch instead of an iPad or
00:15:01
TV because you see 3D like all Pixar
00:15:04
movies are natively 3D and so you got
00:15:07
the Disney Plus app on there you watch a
00:15:08
Pixar movie and you're watching in 3D
00:15:10
the kids are blown away so I think we're
00:15:13
all going to be surprised by how this Go
00:15:14
Disney's all in on it remember when our
00:15:16
parents told us not to sit too close to
00:15:18
the TV now we're just strapping the
00:15:20
thing to our face yeah I I had the most
00:15:23
Silicon Valley moment ever I go to buy a
00:15:27
cup of coffee I was going for my little
00:15:28
walk I see blue Bott I'm like oh you
00:15:29
know I get myself a mocha you know I
00:15:31
lost a little bit of weight I'm going to
00:15:32
treat myself $9 for a mocha number one
00:15:35
that in the C tilted me $9 for a it was
00:15:39
$8 and then I gave a dollar tip and then
00:15:42
I felt cheap giving a dollar tip you
00:15:43
know it's $88.99 for a carton of clover
00:15:45
milk all
00:15:47
organic I mean you can make infinite
00:15:49
lattes at home anyway where did you go
00:15:52
for your $9 mocha I was I'm in Palo Alto
00:15:55
right now because we lost like the blue
00:15:56
bottle yeah said this I'm like $9 what
00:16:00
am I doing you know I just I felt like
00:16:02
buying a chocolate bar the stain dirty
00:16:04
lips left on the cup oh my God look at
00:16:07
so you know what you're a little
00:16:08
obsessed with my lips take it easy there
00:16:11
so anyway then there's a kid in the
00:16:14
place wearing the goggles with the
00:16:16
keyboard he's pounding he's getting work
00:16:19
done this kid was doing work and I tell
00:16:22
you the trth put in the hours he was
00:16:24
putting in the hours no one looks at
00:16:25
your laptop no one looks at your screen
00:16:27
that's what work without anyone seeing
00:16:30
what you're doing this kid had four
00:16:31
desktops up this guy was probably on
00:16:33
PornHub Spotify writing code how many
00:16:37
words did this person say to another
00:16:38
human being while you were there no zero
00:16:40
and you know what when they're on a
00:16:41
laptop they're the same what's the
00:16:43
difference he's he's coding nobody B it
00:16:45
and I I think this is gonna they're
00:16:46
gonna run the table on this I think it's
00:16:48
100 billion 100 billion sales under five
00:16:51
years yeah I take the over I take the
00:16:53
over what do you you got the over or the
00:16:55
under cuz even if they keep it at three
00:16:57
grand they got to sell 30 million UNS to
00:16:59
get to 100 bil they're going to make up
00:17:01
a lot of money on this app store too by
00:17:02
way I you guys are right that it's going
00:17:03
to be successful in terms of Revenue
00:17:06
what I'm asking is a more societal
00:17:08
question is do you guys actually think
00:17:09
it's better no I don't want my kids in
00:17:11
this all day no and I can see this
00:17:12
becoming super predicting hey freeberg I
00:17:14
I can I buy three for your kids just
00:17:16
have them walk around with them I have a
00:17:17
no iPad in the house rule as well but
00:17:19
wait a minute hold on what about
00:17:20
productivity freeer my kids aren't
00:17:22
trying to be productive they're using it
00:17:24
to burn it's called childhood you don't
00:17:26
have a productive childhood it's
00:17:27
supposed to be not productive you guys
00:17:29
understand that at some point you guys
00:17:31
will be the only six kids whose parents
00:17:35
haven't given them this stupid thing to
00:17:36
put on their face no this is going to be
00:17:38
Tim restricted I have a no iPad no phone
00:17:41
no like I let them use the headset good
00:17:44
for them it's so good no no it burns
00:17:46
their Burns their brain away Burns their
00:17:47
brain away it's terrible man I totally
00:17:51
agree with you social interaction the
00:17:53
loss of our ability to communicate as
00:17:54
hum it's critical and it's a fail point
00:17:57
I do think that there where these things
00:17:59
create great unlocks I think this is an
00:18:00
Enterprise device can you imagine giving
00:18:02
a field sales team on the farms to go
00:18:04
there they can take off their swey
00:18:06
headset when the sun is shining and then
00:18:08
give it to the farmer to put on and then
00:18:10
he can put it on and feel the sweat and
00:18:12
the the headband will be wet that's not
00:18:14
the use case it doesn't by the way it's
00:18:16
it's a very personal device in order to
00:18:18
log in you know it does like a I scan um
00:18:21
or you have to have like a lock in like
00:18:23
login like you do with your phone but
00:18:24
then you got to reset the eye because it
00:18:25
automatically sets the eye position so
00:18:28
when put on someone else's headset you
00:18:29
got to reset the IP it's a whole thing
00:18:31
so it's not a transferable device it's a
00:18:33
very personal Computing you know kind of
00:18:35
thing so I don't think it's going to be
00:18:37
the same as like an iPad or a phone it's
00:18:38
a very different kind of thing I don't
00:18:39
know what it's going to look like yet I
00:18:40
don't know I I say next week we do the
00:18:42
show inside of these or at least me and
00:18:43
you freeberg will be will be does is is
00:18:45
there a zoom very funny there's um
00:18:49
there's an a there's an avatar thing and
00:18:51
so what it does it scans your face while
00:18:52
you're talking all four of us can see
00:18:54
each other as the Avatar yeah all right
00:18:56
let's do it it'll be hilarious I had a
00:18:58
moment this week in parenting I had a
00:19:00
moment this week where I told one of my
00:19:03
children that when I send a text message
00:19:06
I expect an immediate
00:19:08
response otherwise I am going to cancel
00:19:11
that child's phone and take it away and
00:19:13
then separately when they respond it has
00:19:15
to be in structured wellth thought out
00:19:18
perfectly formatted English and then
00:19:20
then third I said every single email I
00:19:23
see from you interacting with your
00:19:24
teachers or anybody else that's there to
00:19:27
help you needs to be incredibly well
00:19:29
written and formatted and if I see
00:19:32
garbage English I'm going to take your
00:19:33
phone away oh okay so you don't want
00:19:36
them on their phones but they have to
00:19:38
respond right away well they have very
00:19:39
strict rules on what they can use
00:19:40
they're there for literally all they can
00:19:43
do is communicate like they can use
00:19:45
iMessage but it is shocking to me that
00:19:48
despite the lack of games that they have
00:19:50
or whatever how poor they are in being
00:19:54
able to communicate and what little
00:19:57
access to devices
00:19:59
they have have already made them orders
00:20:02
of magnitude less able to communicate
00:20:04
than frankly I was able to when I was
00:20:05
their age and so yeah I can just imagine
00:20:08
what happens when you become even more
00:20:10
ins sconed in something that you can
00:20:12
cocon yourself with and not have to
00:20:13
interact with the rest I don't disagree
00:20:15
with you I don't disagree with you not
00:20:16
to say that it's not going to be a
00:20:18
revenue generator but I think that you
00:20:20
could just as easily
00:20:22
frankly instead of impacting Apple's
00:20:24
revenues you can probably go along the
00:20:26
makers of ssris
00:20:29
pot here comes the spread trade pot
00:20:32
Bumble and Tinder and you'll get to the
00:20:34
same place economically all right all
00:20:37
right here we go we got a lot on the
00:20:39
what a Great Leap Forward for Humanity I
00:20:41
can't wait s i I just see this as the
00:20:44
laptop replacement okay I wanted to talk
00:20:47
a little bit about what apparently is
00:20:49
going to be the the spread trade of the
00:20:50
last year meta is continued their
00:20:54
unbelievable run and snap dropped like
00:20:58
30% here's a chart for yall of snap
00:21:00
versus meta you can take a quick look at
00:21:02
it here and just for context both
00:21:04
companies did great during Co and Zer
00:21:07
hit all-time highs in 2021 but they both
00:21:09
got crushed due to the ad spend pullback
00:21:11
obviously but then meta started to get
00:21:14
less focused on their headsets and more
00:21:16
focused on AI started doing their
00:21:18
reduction in headcount
00:21:20
22% year-over-year from 86,000 to 67,000
00:21:24
the last quarter for meta and they're
00:21:27
quarterly profits have increased to an
00:21:29
all-time high of $14 billion that's
00:21:32
profits folks in Q4 for meta all-time
00:21:35
high for the stock price $470 a share
00:21:38
$1.2 trillion market cap sna down 60%
00:21:43
from its closing price on its IPO Day in
00:21:45
2017 let me just jump to chth before I
00:21:48
get into more charts and everything you
00:21:49
pointed out chth and maybe you could
00:21:51
explain to the audience just how
00:21:54
ridiculous the voting rights were and
00:21:57
the mass massive dependence that the
00:22:01
snap team and the executives had on
00:22:03
stock based comp two issues for
00:22:06
youth well I mean I think I said it
00:22:08
before I think that case studies have
00:22:11
been written about
00:22:13
how tilted the governances in snap I
00:22:16
think the point is that they basically
00:22:17
have infinite to zero voting power over
00:22:20
common shareholders so there's no real
00:22:23
feedback loop and I think that that has
00:22:25
probably adversely affected
00:22:28
the types of people that traffic in
00:22:31
their stock now look
00:22:35
activists and short sellers sometimes
00:22:38
have a very bad
00:22:40
reputation but if you steal man their
00:22:42
side of it what they are there to do is
00:22:45
to shine a light on inefficiency and in
00:22:48
the short seller case sometimes
00:22:49
impropriety but it should all lead to
00:22:52
companies being better run right I think
00:22:55
meta had this example where they had a
00:22:58
really big Hiccup and everybody
00:23:00
including us sort of pointed out the
00:23:03
levels of
00:23:04
spend that they were making really
00:23:07
didn't make any sense I think we had a
00:23:08
chart that compared the level of spend
00:23:10
of meta second only
00:23:12
to like the spaceship program right just
00:23:15
like Bonkers an enormous amount of money
00:23:17
and look Mark got the message he heard
00:23:21
it loud and clear I think he got fed up
00:23:23
with whatever was going on there and he
00:23:25
fixed it and it's in the number
00:23:28
now I don't know snap because to be
00:23:30
honest with you I've never taken more
00:23:32
than one second to look at that company
00:23:35
and the reason is there's just zero
00:23:37
ability for me to have any useful say so
00:23:40
I've never honestly looked at its
00:23:41
performance I've never studied a single
00:23:44
characteristic I've never trended it and
00:23:47
I think the point is that I am probably
00:23:50
where a lot of other reasonably smart
00:23:52
folks who could give a reasoned opinion
00:23:54
on how to make it better land and part
00:23:58
of the reason is because there is no
00:24:00
feedback loop that matters yeah and when
00:24:02
you know that why would you waste your
00:24:04
time at least in other options right
00:24:06
there are other options and and meta was
00:24:08
another one you know you can write a
00:24:10
letter it gets picked up on CNBC and
00:24:13
Bloomberg and whatever and all of a
00:24:15
sudden they kind of pay attention and I
00:24:17
think and you look at Disney Nelson
00:24:19
pelts goes and gets Ike pearlmutter
00:24:21
shares buys some more takes a Lars yeah
00:24:24
we'll see whether that fixes itself the
00:24:26
point is that and of these other cases
00:24:29
people are investing the time because
00:24:32
they think that there's even a small
00:24:33
shred of a chance that the company
00:24:35
listens but if you literally have no say
00:24:39
you couldn't even do a proxy you
00:24:40
couldn't vote the shares why would you
00:24:42
bother and I think that that's more of
00:24:44
an example where maybe there is a I so I
00:24:47
don't even know why snaap it poorly and
00:24:50
again I'm not going to really take the
00:24:51
time because it's like why bother taking
00:24:53
the time sack should should they unwind
00:24:55
this like no voting common shares super
00:24:58
voting shares nonsense and and should
00:25:00
this go away as a concept in the stock
00:25:03
market well I mean Facebook or meta has
00:25:06
a pretty similar concept I mean I guess
00:25:09
Zuckerberg has 60% voting control
00:25:12
whereas Evans spel has 99% so snap is
00:25:15
more
00:25:16
egregious the difference is that
00:25:19
Zuckerberg is listening and Spiegel is
00:25:21
not the the reason why snap is doing
00:25:25
poorly is not because its Revenue has
00:25:27
erated so I looked up or let's put it
00:25:30
this way I asked chat GPT for their key
00:25:33
metrics so assuming GPT is not
00:25:35
hallucinating if you compare 2021 to
00:25:39
2023 their total revenue went up from
00:25:41
4.1 to 4.5 billion and gross profit went
00:25:47
from call it 2.4 to 2.5 billion so not a
00:25:51
huge increase but revenue and gross
00:25:53
profit were slightly up but if you look
00:25:55
at operating expenses they went from 3
00:25:58
billion to 4 billion a year and that is
00:26:01
why their operating income or operating
00:26:04
loss went from a $700 million loss to$
00:26:07
1.4 billion loss in two years so that's
00:26:11
the source of the problem is that they
00:26:14
increase their operating expense by a
00:26:16
billion dollars a year from 2021 to 2023
00:26:21
it's pretty simp they seem like they're
00:26:22
the last ones to get the memo yeah they
00:26:24
they were the last ones to get the memo
00:26:26
and just just finish the point so you
00:26:27
saw that a few days ahead of this
00:26:30
quarterly announcement where their stock
00:26:32
got crushed they put out a press Le
00:26:34
saying they're going to cut their head
00:26:35
count 10% it's too little it's too
00:26:38
little too late yeah they knew right
00:26:40
they knew they had a problem so they
00:26:42
released the the Press Le saying oh
00:26:44
we're going to cut well you should have
00:26:46
done what Zuckerberg did you know
00:26:48
Zuckerberg did a 20% cut last year he
00:26:51
got serious he got lean and fit and
00:26:55
instead these guys held out did nothing
00:26:58
then when they know that the Market's
00:26:59
going to crush them they put out this
00:27:01
lame announcement 10% no not 10% really
00:27:04
if you just want to get back to where
00:27:07
you were two years ago in terms of
00:27:09
operating expense you need a 25%
00:27:12
reduction yeah yeah but it's more than
00:27:14
that if you look at the numbers let's
00:27:16
use operating cash flow was 165 million
00:27:18
for SNAP for the quarter so their
00:27:20
operations generated 165 million of
00:27:22
profit but for the entire year because
00:27:25
they lost money in the quarters prior
00:27:27
they generated free cash flow of only
00:27:29
$35 million so the business net in
00:27:33
produced $35 million of incremental cash
00:27:36
you know how stock-based comp accounting
00:27:37
works the charge happens when it vests
00:27:40
so this is what employees are vesting
00:27:42
during the year of 2023 employees vested
00:27:45
$1.3 billion of stock-based comp so that
00:27:48
means new shares or options were issued
00:27:50
that on an accounting basis the options
00:27:52
are valued using black schs and the
00:27:54
shares are valued based on the share
00:27:55
price so they issued 1.3 billion of
00:27:57
stock based comp so they generated 35
00:27:59
million of free cash and they used $1.3
00:28:02
billion to compensate employees beyond
00:28:04
their Opex so that means that they paid
00:28:07
employees 40 times the free cash flow
00:28:10
that was generated for shareholders
00:28:12
during the year which is also equivalent
00:28:14
to 10% of the Enterprise market value of
00:28:17
this company so the Enterprise value of
00:28:19
the company is $15 billion 10% of that
00:28:23
was issued to employees to compensate
00:28:24
them now let me give you the the the
00:28:26
story of another city meta and by the
00:28:28
way snap's share count because they
00:28:30
issued all the stock the number of
00:28:32
shares outstanding increased by 4%
00:28:34
during the year during the year meta's
00:28:37
number of shares outstanding decreased
00:28:39
by half a percent because they used cash
00:28:41
to go and buy back stocks so they were
00:28:43
able to reduce the shares outstanding
00:28:45
now as you guys talked about meta cut
00:28:47
employee count by 22% and snap cut
00:28:49
employee head counts by 3% during the
00:28:52
year but here's the crazy difference in
00:28:54
performance the stock-based comp expense
00:28:57
for meta during that year was about $14
00:29:00
billion invested that year that company
00:29:03
generated 71 billion of operating cash
00:29:06
flow so while while snap gave employees
00:29:08
40 times the free cash flow meta gave
00:29:12
employees you know about a 20% of the uh
00:29:15
of the free cash flow and then and then
00:29:17
meow went around and they used some of
00:29:18
that extra cash to buy back $20 billion
00:29:20
of stock so they bought back more shares
00:29:23
than what the employees were issued that
00:29:24
that year were so it shows such a differ
00:29:28
in looking out for shareholders so if
00:29:30
I'm an investor and by the way meta is
00:29:32
trading at like 25 times free cash flow
00:29:34
which is not a crazy multiple given all
00:29:36
the new businesses that they have in llu
00:29:37
and the progression to cloud and other
00:29:39
things that they might do if I'm looking
00:29:41
at those two businesses as a shareholder
00:29:43
you got this guy that controls the whole
00:29:45
stock he's giving employees a billion
00:29:47
three of shares a year when he's only
00:29:49
making $30 million of free cash flow a
00:29:51
year and then the other guy is issuing
00:29:54
$4 billion of shares buying them all
00:29:56
back and he's making 70 billion of free
00:29:58
cash flow a year I don't know it's very
00:29:59
hard to decide which one to go after
00:30:01
Spiel brought it up in an interview I
00:30:03
saw and a lot of the layoffs were
00:30:06
topheavy so he got rid of a lot of the
00:30:08
top people who had these huge comp
00:30:10
packages and then what I'm hearing from
00:30:13
a lot of Executives is cutting these
00:30:15
highly stock comped Executives who are
00:30:19
you know also have big cash comp cutting
00:30:21
them putting lieutenants in charge and
00:30:23
then moving more jobs to other locations
00:30:26
where people don't expect stop
00:30:28
stock-based comp you know if you're in
00:30:29
India or you're in South America
00:30:32
whatever you know stock-based comp is
00:30:34
not like the obsession it is here so as
00:30:36
everybody optimizes these businesses I
00:30:39
mean Facebook even why do they need
00:30:40
5,000 employees so they announced
00:30:43
roughly 500 job cuts out of what 5,500
00:30:47
employees that's crazy I mean should
00:30:51
that company be operating with 2,000
00:30:53
employees it's good question cut the
00:30:55
number of Twitter employees from 8,000
00:30:57
to 1500 when you look at the number of
00:31:00
apps that they're running and the number
00:31:01
of products that they're running
00:31:02
compared to meta right meta has far more
00:31:04
apps far more infrastructure meta is
00:31:06
serving 3.2 billion daily active users
00:31:10
snap is about 400 million So Meta is 8x
00:31:14
the users with many more applications
00:31:16
and much more
00:31:18
infrastructure so I think it's a it's
00:31:20
another great kind of ratio to look at
00:31:22
the performance of these two I think
00:31:24
you're exactly right yeah the other
00:31:26
advantage that it has is because they're
00:31:29
so profitable they have the resources to
00:31:31
go big in AI big time which is very
00:31:34
expensive so yeah so they are the leader
00:31:36
you get all this option value at meta
00:31:37
which you don't get at snap there's all
00:31:39
this infrastructure that they can
00:31:41
leverage much like Amazon did with AWS
00:31:43
into things like Cloud AI tools for
00:31:46
thirdparty developers third party
00:31:48
applications and then obviously the you
00:31:51
know meta is the biggest advertising
00:31:53
platform next to Google in the world now
00:31:56
and there's much more that they can
00:31:57
start to do to to extend further into
00:32:00
the the they get they did get an awesome
00:32:02
save remember Apple screwed them and was
00:32:04
like you can track devices now and like
00:32:07
that just took a massive hit in the ad
00:32:09
Network and it was all those headwinds
00:32:11
they were like okay we're just going to
00:32:12
use AI to optimize ads and supposedly
00:32:15
the AI optimization of ads I was talking
00:32:16
to somebody on the inside they said like
00:32:19
yeah we got it all back we gained it
00:32:21
back we've got massive AI advertising
00:32:24
optimization going on so totally yeah
00:32:25
that's great that Tim Cook you know
00:32:27
kicked us in the nuts but we don't care
00:32:30
by the way that's a great point jcal it
00:32:31
really says a lot about how meta was
00:32:34
able to respond to that change which a
00:32:37
lot of people speculated would destroy
00:32:38
the advertising business and the fact
00:32:40
that they were able to engineer
00:32:41
solutions to drive advertising Revenue
00:32:43
up to $40 billion it's just mind-blowing
00:32:47
it's a really kind of impressive outcome
00:32:49
for the team and I think it speaks a lot
00:32:50
to the quality of the engineers there
00:32:52
yeah I think it's a great Point yeah
00:32:53
saaks you tweeted that you're seeing a
00:32:55
little SAS bounce back all of a sudden
00:32:58
that's interesting I am seeing something
00:33:00
similar last year last two years you had
00:33:02
a ton of people cutting their SAS Bend
00:33:05
maybe removing the number of SAS vendors
00:33:07
they had consolidating vendors uh you
00:33:09
tweeted many public and private software
00:33:11
companies are experiencing accelerating
00:33:12
growth after six to 7 quars of
00:33:15
deceleration SAS recession appears to be
00:33:18
over according to the SAS Master David
00:33:21
saxs you want to unpack this for us what
00:33:23
do you seeing well it's still pretty
00:33:24
early because not everyone's reported
00:33:26
but if if you looked at the big Tech
00:33:29
Cloud performance in Q4 you could see
00:33:32
that there's a bounce back in here this
00:33:34
is net new ARR added for AWS Azure and
00:33:39
Google Cloud so you see here in Q4 that
00:33:43
there's a huge increase in net and U AR
00:33:46
for the the big cloud computing
00:33:48
platforms and then I think another Bell
00:33:50
weather is at lassan so we're still
00:33:52
waiting to hear from HubSpot Salesforce
00:33:55
Zoom Adobe companies like that they
00:33:56
haven't reported yet but if you look at
00:33:59
makes jira amongst other products
00:34:01
they're based in Australia yeah the
00:34:03
major yeah exactly collection of SAS
00:34:05
companies right it's a collection of SAS
00:34:07
products yeah so net new ARR would be
00:34:10
the amount of growth in that quarter and
00:34:13
this is on a year-over-year basis so you
00:34:15
can kind of see Q4 of 21 was the
00:34:19
absolute Peak and then it plummeted and
00:34:23
then it actually went negative for about
00:34:25
a year that's that's tough to be in a
00:34:27
company with new AR going negative yeah
00:34:31
yeah that doesn't mean by the way the
00:34:32
company's shrinking it just means that
00:34:34
the amount of net new ARR which is the
00:34:37
amount of growth is actually smaller
00:34:40
than that same quarter a year before
00:34:43
yeah and then in Q4 you could see
00:34:46
there's some acceleration here that
00:34:47
they're starting to add more they added
00:34:49
more net new ARR I guess 33% more in Q4
00:34:53
than they did over the previous year and
00:34:55
part of that sacks is because the comps
00:34:57
are lower and they kind of bottomed out
00:34:59
yeah they bottomed out now they're re
00:35:01
accelerating so you know we're starting
00:35:03
to see this in some of my board meetings
00:35:05
as well where in 2022 everybody was
00:35:09
missing their numbers and reforecasting
00:35:11
down and then they would miss the
00:35:12
reforecast yeah so by
00:35:15
2023 the forecasts were very very
00:35:18
conservative and I would say and now I'm
00:35:20
seeing companies beat the the sort of
00:35:23
the lower forecast in Q4 this wasn't
00:35:27
happening earlier in the year but
00:35:28
finally I think people are starting to
00:35:29
beat their sort of their lower forecast
00:35:32
for Q4 that's the question that I was
00:35:34
curious about what do you what do you
00:35:36
actually think is happening is that
00:35:38
we've Reb baselined these businesses so
00:35:40
now what would have looked like just a
00:35:42
massive Miss over the last two years now
00:35:45
looks like a beat because we've just
00:35:47
completely reset expectations is it that
00:35:50
or is it that the economy is actually
00:35:52
expanding and we can count on some
00:35:56
reasonable growth rates is it a combo of
00:35:58
the two what do you think it actually is
00:36:01
yeah I mean it's definitely a new
00:36:04
Baseline in the sense that and if you go
00:36:06
back to 2020 or 2021 we considered good
00:36:10
growth to be you know 2 to 3x year-over
00:36:13
year and now if it's going from 60 to
00:36:16
80% growth year-over year you're happy
00:36:18
so there's definitely been a lowering of
00:36:20
expectations that being said you still
00:36:23
see in these numbers there has been a
00:36:24
bottoming out and we're starting to now
00:36:26
grow from this new
00:36:29
Baseline so for example I think with
00:36:33
atlassian here we are seeing an increase
00:36:37
in spend basically in in growth right so
00:36:39
the way a recession is typically defined
00:36:40
is uh two quarters of negative growth
00:36:43
right we had six to seven quarters of
00:36:46
decelerating or negative growth in SAS
00:36:49
in Tech in SAS which is why called it
00:36:51
the session or be the yeah it was
00:36:54
actually kind of a depression you're
00:36:55
right but now we're seeing quarter over
00:36:58
quarter growth so growth is
00:37:00
reaccelerating growth is higher than it
00:37:02
was so is it going to get to where it
00:37:04
was that probably will take some time
00:37:06
but it feels like the problems in the
00:37:08
ecosystem work themselves out and now
00:37:10
we're back to growth again that yeah I
00:37:12
can add psychologically because I'm on a
00:37:14
couple of SAS boards as well and
00:37:16
psychologically it felt like you tell me
00:37:18
if I'm right SAS saak if you saw the
00:37:19
same thing there were two years of
00:37:21
calling up customers and they were like
00:37:23
we're we're consolidating vendors and by
00:37:25
the way we did a riff and so we need 20%
00:37:27
less seats so we're going to have 20%
00:37:31
less SAS companies that we're buying
00:37:32
from and we're going to have 20% less
00:37:35
seeds so you start putting that all
00:37:36
together man everybody was just in
00:37:39
psychological triage mode we cannot
00:37:41
spend money I don't want to lose my job
00:37:43
so you're if you're a procurement person
00:37:45
you're the CTO you don't want to lose
00:37:46
your job you don't want to have more
00:37:47
Cuts so you're like well I can cut some
00:37:49
software costs do I get points for that
00:37:51
and the points you would score for the
00:37:53
last two years was cutting costs with
00:37:55
the market RI in and you you know now
00:37:58
got a really you know efficient company
00:38:00
you're like hey can we spend a little
00:38:01
bit on SAS to make the remaining
00:38:03
employees even more you know productive
00:38:07
okay maybe that's a reasonable
00:38:08
discussion and then people are playing
00:38:09
ball in terms of negotiating prices so
00:38:12
that's the other thing I see is like
00:38:14
people are like we we'll take your
00:38:15
software but here's what we want to pay
00:38:16
and then they're coming to the board and
00:38:18
saying can we do this deal would have
00:38:20
been a million dollar deal but it's a
00:38:21
$200,000 like yeah take the money take
00:38:23
the money let's let's be hug that
00:38:25
customer the market is generally an
00:38:27
escalator on the way up an elevator on
00:38:29
the way down so the recovery is going to
00:38:31
take a long time but at least we've
00:38:33
bottomed out and we're in recovery as
00:38:35
opposed to continuing declines yeah by
00:38:38
the same token if you're a startup and
00:38:40
you're not seeing Improvement in your Q4
00:38:42
sales then you no longer have a macro
00:38:45
excuse for why you're not doing well
00:38:48
interesting and then freeberg you added
00:38:50
you know you're like I'll I'll make my
00:38:52
own software you said uh you know some
00:38:54
softare software is too expensive I'll
00:38:56
put a developer on it and so how's that
00:38:59
working out for you are you still in
00:39:00
that mindset of like yeah maybe we just
00:39:02
build our own software yeah I mean I
00:39:04
it's not just us I think we're seeing a
00:39:06
lot of companies pursuing this path a
00:39:09
couple Engineers can rebuild the
00:39:11
functionality of core applications
00:39:14
particularly because I think if you
00:39:16
think about the business model that
00:39:17
makes SAS so great is they could value
00:39:20
share rather than charge the cost of an
00:39:22
engineer plus some margin the business
00:39:26
model the equity value that comes in
00:39:28
software if you can build something once
00:39:31
that creates $100 of value you could
00:39:33
probably charge your customer $30 $40
00:39:35
for that product because it's saving
00:39:37
them 60 bucks 70 bucks and they'll make
00:39:39
that switch to software so you know the
00:39:42
ROI driven value share model in SAS has
00:39:46
made it incredibly valuable the problem
00:39:48
now is that an engineer can be hired to
00:39:52
build the replacement and so it creates
00:39:54
price compression so the SAS company can
00:39:57
no longer capture that much value
00:39:59
because the savings is actually less
00:40:00
than that because the Enterprise might
00:40:03
say hey I'm going to hire someone and
00:40:04
instead of spending 60 Grand a year on
00:40:07
your software I'm going to allocate a
00:40:08
quarter of an engineer's time to build
00:40:10
that software and it's going to replace
00:40:11
that that cost so I think that that's
00:40:13
still the case so while there might be
00:40:15
bookings there's still which are driven
00:40:18
largely by a search for efficiency gains
00:40:20
a search for more profitability for more
00:40:22
productivity within an Enterprise there
00:40:24
are other options for that Enterprise to
00:40:26
realize that productivity gain today and
00:40:29
that's what's going to cause perhaps
00:40:30
price compression and more competition
00:40:33
than has been the case but I don't think
00:40:35
that the adoption of software is going
00:40:37
to slow down it certainly seems to be re
00:40:40
accelerating which is great competitive
00:40:41
right we're moving into a hyper
00:40:43
competitive market right especially with
00:40:45
AI it's a mix of internal software it's
00:40:47
a mix of internal software as you guys
00:40:49
know there are very few traditional
00:40:51
non-tech Enterprises now that don't have
00:40:53
a software team that can write code so
00:40:55
now that so many companies have software
00:40:58
teams that write code they're all going
00:40:59
to be asking the question should we be
00:41:01
buying the software or should we be
00:41:02
building something internal yep it's a
00:41:04
classic buyer build situation all right
00:41:06
let's talk a little bit about VCS and
00:41:08
how they're investing in AI there there
00:41:09
seems to be three camps shaping up here
00:41:11
chamath you know one group is like and
00:41:14
the incumbents are going to win you know
00:41:15
Microsoft Google Amazon everybody
00:41:17
they're going to win the day so they're
00:41:19
going to wait and see then there's
00:41:22
another group who's sitting it out
00:41:23
because they're like hey open sour is
00:41:25
going to win meta committed to open
00:41:28
source and collaborative platforms I've
00:41:30
been playing with hugging face with suep
00:41:34
as well as you Chim moth and it's pretty
00:41:36
amazing what's happening over there and
00:41:37
then a bunch are obviously placing bets
00:41:39
right now the valuations are absurd
00:41:42
Founders fund and inre and Hartz two
00:41:43
notable firms are approaching it
00:41:45
differently Founders fund bought into
00:41:47
open AI at a $29 billion valuation but
00:41:51
aside from that investment they're
00:41:53
generally avoiding the AI deals on the
00:41:55
other hand andreon is is betting heavily
00:41:58
character AI repet 11 Labs mistel you're
00:42:01
also in repet Sachs so what do you think
00:42:05
is open source going to win the day
00:42:06
you've been picks and shovels the whole
00:42:08
way you've been talking about
00:42:09
compression maybe this isn't actually a
00:42:10
good
00:42:11
Market what's you're thinking as a
00:42:13
capital allocator J I think foundational
00:42:16
models will have no economic value I
00:42:19
think that they will be an incredibly
00:42:20
powerful part of the
00:42:22
substrate and they will be broadly
00:42:25
available and and entirely free wow so
00:42:28
if you think about that any closed model
00:42:31
especially a closed model that operates
00:42:33
on open on the open internet is not very
00:42:37
valuable and any open source model that
00:42:40
operates on the that trains on the open
00:42:43
internet will make that so so in that
00:42:46
world things like mistol and
00:42:50
llama will essentially Decay the market
00:42:52
to zero so if you if you're looking at
00:42:55
any economic value that has been
00:42:57
captured up until today if it has been
00:43:00
captured by having a proprietary closed
00:43:02
model trained on open
00:43:05
data that economic value will go away
00:43:09
and I think Google and Microsoft and
00:43:11
Facebook and Amazon and all these
00:43:15
startups have a deep economic incentive
00:43:17
actually to make that so so now you can
00:43:20
evaluate what that means so if you get
00:43:23
an opened model from hugging face that's
00:43:25
just kick out
00:43:27
where do you spend money well you're
00:43:29
going to have to spend money to actually
00:43:31
train it to finetune it maybe to have
00:43:35
some pretty Zippy
00:43:37
inference and all of that means that
00:43:39
there's a new kind of substrate that has
00:43:41
to be built which is all around the way
00:43:44
that the tokens per second are
00:43:46
provisioned to the apps that sit on top
00:43:48
of the model what that means is you need
00:43:50
to go back to 2006 and7 and say okay
00:43:53
when we first created the
00:43:55
cloud who made money and fast forward 18
00:43:59
years later it's the same people that
00:44:01
are still making money so the people
00:44:03
that made money in 2006 and 7 were
00:44:06
Amazon principally because of ec2 and
00:44:10
S3 The Perfect Analogy of ec2 and S3 in
00:44:14
2024 is the token per second provider
00:44:17
now there you have to double click and
00:44:19
say okay well what does a tokens per
00:44:21
second provider need to do to make a lot
00:44:23
of money and I think the ultimate answer
00:44:25
is you need your own proprietary
00:44:27
Hardware so who is in a position to do
00:44:29
that Amazon has announced that they have
00:44:32
an inference and training solution for
00:44:35
training cerebrus has announced a pretty
00:44:37
compelling solution Google obviously has
00:44:39
TPU then there's a handful of startups
00:44:41
including one that I helped get off the
00:44:43
ground in 2016 that I funded called Gro
00:44:47
all of those companies are in a position
00:44:48
to build a tokens per second service
00:44:51
then you have companies like together AI
00:44:53
which basically just go and take Venture
00:44:55
money and
00:44:57
wrap Nvidia
00:45:00
gpus and you can debate what the
00:45:02
advantage will be there one could say
00:45:04
well it's not really a huge advantage
00:45:07
over time so my refined thoughts today
00:45:11
are sort of what my initial guess was
00:45:14
when we started talking about AI a year
00:45:15
ago which is the picks and shovels
00:45:18
providers can make a ton of money and
00:45:21
the people that own proprietary data can
00:45:23
make a ton of money but I think open
00:45:25
Source models will basically crush the
00:45:28
value of models to zero economically
00:45:30
even though the utility will go to
00:45:31
Infinity the economic value will go to
00:45:33
zero did any of you guys see chim's
00:45:35
interview with Jonathan
00:45:37
Ross NOP not yet you put it out right
00:45:39
Cham you made it public you know I did
00:45:41
it just for my subscribers but Jonathan
00:45:43
is the founder and CEO of grock the
00:45:46
company that I just mentioned and the
00:45:48
quick version of that story is I was I
00:45:50
would I would pour over the Google
00:45:52
earnings results in the mid teens of
00:45:55
2000 cuz I was pretty actively investing
00:45:57
in a bunch of different public equities
00:46:00
and Sundar said in a press release he
00:46:02
mentioned that they had rolled their own
00:46:03
silicon for machine learning called TPU
00:46:07
and I was like what is going on that
00:46:09
Google thinks that they can actually
00:46:12
roll their own silicon what must they
00:46:14
know that the rest of us don't know and
00:46:17
so it took me about six or nine months
00:46:18
but through sunny I got introduced to
00:46:21
Jonathan and then we were able to get
00:46:24
Jonathan to leave Google and he started
00:46:26
and he Jonathan was a founder of TPU at
00:46:28
Google and then he started grock which I
00:46:30
was able to lead that funding round in
00:46:34
in 2016 so eight years
00:46:36
ago anyways I did a a spaces with
00:46:39
Jonathan talking about the entire AI
00:46:41
landscape and AI acceleration to my
00:46:43
subscribers but it was so good I got to
00:46:46
say he is he was so
00:46:49
impressive that we kind of like figured
00:46:52
out a way to just play the space and
00:46:56
tape it and then we published it to
00:46:57
everybody so it's it's on it's on my
00:46:59
Twitter for anybody that wants to listen
00:47:00
to it it is amazing he is really
00:47:04
impressive I was sitting on the 17 going
00:47:07
to Santa Cruz not moving for hour and a
00:47:11
half and I listened to it so I kept me
00:47:13
alive but I thought it was really great
00:47:15
what you think he is great no he's great
00:47:18
some great insights and I think he's
00:47:20
very
00:47:21
compelling in arguing why some of the
00:47:24
big cloud providers today that are
00:47:27
offering infrastructure for AI model
00:47:31
training and
00:47:32
inference are going to be challenged if
00:47:35
someone can build full stack and be and
00:47:38
do it successfully so it was a really
00:47:40
good interview I actually think it's
00:47:41
really worth listening
00:47:43
to but I enjoyed it yeah thanks for
00:47:45
putting it out there I was like
00:47:46
literally just sitting I was sitting in
00:47:47
the car browsing Twitter and I saw your
00:47:50
thing and I clicked on it and then I
00:47:51
just ended up listening it was a little
00:47:53
it's a little hard actually when you do
00:47:54
a space for your Subs you can't actually
00:47:57
just flip a switch and then release it
00:47:59
to all of your followers so we actually
00:48:03
had to like literally play it and then
00:48:06
just capture the audio out and then
00:48:07
republish it but anyways despite that
00:48:10
inconvenience if anybody's interested in
00:48:12
learning about AI Hardware he is very
00:48:14
compelling and he's very educational so
00:48:16
saak your thoughts on just how you're
00:48:18
approaching investing in AI if you're
00:48:20
specifically investing in the
00:48:22
underpinnings of AI pix and shovels yada
00:48:24
yada or if you're just looking on the
00:48:26
application Level and it's you know you
00:48:28
know that kind of approach well we we
00:48:30
divided the space into three categories
00:48:33
uh one is the the models themselves the
00:48:35
foundation models which can be either
00:48:37
open source or closed
00:48:39
Source there's infrastructure so like
00:48:42
jam saying it could be like model
00:48:44
training it could be Vector databases
00:48:47
tools that developers use to create the
00:48:51
AI stack typically inside their
00:48:52
Enterprise and then the third would be
00:48:54
applications which can be things like
00:48:56
co-pilots or it could be a pre AI app
00:48:59
that's using AI to kind of turbocharge
00:49:03
its capabilities yeah most SAS would be
00:49:06
in the application bucket and so that's
00:49:09
principally where we're focused although
00:49:10
we do look at infrastructure plays and
00:49:13
models however I do think there is an
00:49:15
argument for I mean really with the
00:49:18
question of commoditization well like
00:49:20
all the model companies just get totally
00:49:23
commoditized we really we're talking
00:49:24
about open AI right because they're the
00:49:26
leader so the question is can they
00:49:27
maintain their lead I do think there is
00:49:30
an argument that open AI
00:49:34
will stay in the lead and actually do
00:49:36
quite well and I think there's a few
00:49:39
points there one is that if you're a
00:49:42
consumer you just want to use the best
00:49:44
GPT you want to use Google Ex it's just
00:49:47
like search right if Google is a little
00:49:49
better or the perception is it's a
00:49:51
little better than Bing or the other
00:49:53
search engines you don't win a plurality
00:49:56
of search traffic you actually end up
00:49:58
winning it all because consumers just
00:49:59
want the very best one so most of the
00:50:02
tests show that open AI is still ahead
00:50:05
of the open source models and I think
00:50:06
even people in the open source movement
00:50:08
will tell you that open AI is call it
00:50:10
six months ahead they have no doubt that
00:50:13
open source will get to where open AI is
00:50:15
now in six months nonetheless if open AI
00:50:19
just maintains a little bit of a lead
00:50:21
over open source then it could compound
00:50:26
yeah it basically win the vast vast
00:50:28
majority of the call it consumer search
00:50:30
or consumer GPT market so that's Point
00:50:33
number one point number two is now that
00:50:37
open aai has these hundreds of millions
00:50:40
of consumers using it that's a pretty
00:50:42
attractive audience for developers to
00:50:44
want to reach and open AI has done a
00:50:48
really good job creating a platform for
00:50:50
developers to create you know what are
00:50:52
called custom gpts so most developers
00:50:56
don't want to go through the hassle of
00:50:59
training a Model fine-tuning A model
00:51:01
doing all of that work that you would
00:51:02
have to do in the open source ecosystem
00:51:04
they just want to point chat GPT at a
00:51:10
repository of data or documents
00:51:13
information have it learn what it needs
00:51:15
to learn fine-tune it in that way maybe
00:51:17
add some lightweight functionality using
00:51:19
open ai's platform to create a custom
00:51:22
GPT that's what I think most developers
00:51:24
want is they just want a simple stack to
00:51:27
work with and they're going to prize
00:51:30
again Simplicity and the power of the
00:51:32
developer tools over the theoretical
00:51:35
control they get by rolling their own
00:51:37
models training and funing their own
00:51:39
models in open source and so I think
00:51:41
what you're seeing now is I mean how
00:51:43
many custom gpts have already been
00:51:44
created on the platform it might be tens
00:51:47
of thousands I mean there's so Millions
00:51:49
yeah so easy to create them yeah so you
00:51:51
have a classic developer Network effect
00:51:53
where you've got open AI aggregating
00:51:55
hundreds of millions of consumers
00:51:57
because they perceive that chat GPT is
00:51:59
the best then you've got developers
00:52:01
wanting to reach that audience so they
00:52:02
build custom gpts on the open AI
00:52:05
platform that actually gives chat GPT
00:52:08
more capability yeah and that's
00:52:10
something that open source can't easily
00:52:11
catch up with actually actually just
00:52:15
finish the point so yeah so it is a
00:52:16
flywheel where you know classic
00:52:19
operating
00:52:20
system developer Network effect where
00:52:22
you want to use the operating system
00:52:24
that is the most programs written for it
00:52:27
yeah and interestingly hugging face has
00:52:30
realized this and hugging face released
00:52:32
this week their own version of gpts
00:52:34
which is really interesting and you can
00:52:36
pick Sachs which open source project you
00:52:39
want to use to make it so unlike gpts on
00:52:41
chat GPT we have to pick theirs on the
00:52:44
hugging face one you could pick you know
00:52:46
llama or whichever one you want there's
00:52:49
an account called artificial analysis
00:52:51
that you can follow the thing to keep in
00:52:53
mind saaks is that for any of this to be
00:52:56
true these apis need to be usable right
00:52:58
I mean I don't know if you remember but
00:53:00
when we were building apps even as back
00:53:02
as the late 2000s and early 2010s one of
00:53:07
the things was there was a pretty
00:53:08
important paper that was published by
00:53:09
Google about attention span and it would
00:53:12
look at page load times in a cold cache
00:53:15
environment right and it basically said
00:53:17
you have to be at like 150 milliseconds
00:53:20
right that's like Best in Class
00:53:22
performance or faster and I remember
00:53:24
when we read that at Facebook we went
00:53:25
crazy so much so that at one point a
00:53:28
small team and I kind of actually
00:53:30
launched a strip down version of
00:53:32
Facebook to compete with Facebook if
00:53:34
there's a Nick you can probably find
00:53:35
this article on Tech runch and we did it
00:53:37
without telling everybody was called
00:53:38
like Facebook zero anyways the point is
00:53:40
speed matters because in the absence of
00:53:43
having very Snappy response you could
00:53:45
have the best model in the world but if
00:53:47
it takes 10 20 30 seconds to basically
00:53:50
initiate and get back data from a fetch
00:53:52
request it's an impossible thing to to
00:53:55
do so I think one of the things that you
00:53:57
have to keep in mind is that there are
00:53:59
these two things that need to move at
00:54:00
the same time one is the quality of how
00:54:02
the model is but two is the speed and
00:54:04
its responsiveness which is a function
00:54:06
of again hardware and your ability to
00:54:08
basically tokenize tokens per second
00:54:11
very very quickly so that developers are
00:54:13
incentivized to not just play around in
00:54:15
a sandbox but to actually build
00:54:17
production code and I don't think we've
00:54:19
seen that second thing happen because
00:54:21
nobody is delivering it and that's the
00:54:22
big thing that nobody talks about for
00:54:24
example like AWS if you look inside of
00:54:27
how expensive it is to build an app
00:54:28
there I've tried even when they give you
00:54:31
credits the credits they give you aren't
00:54:32
sufficient enough to even pay for half
00:54:34
the
00:54:35
power and then the way that they
00:54:37
schedule and the way that they try to
00:54:39
orchestrate you to use Hardware makes
00:54:41
building production apps unless you are
00:54:43
willing to spend millions and millions
00:54:45
of dollars for a very slow app
00:54:48
unfeasible and so if you go back to a
00:54:50
startup economy raising money here the
00:54:53
Venture investor should start asking the
00:54:56
question well what is the speed and
00:54:58
usability of these services that I'm
00:55:00
funding and the reason is because you
00:55:03
could build the best experience in the
00:55:04
world that runs on Local Host but if all
00:55:07
of a sudden you actually try to launch
00:55:08
it as an app and the thing takes 35 and
00:55:10
40 seconds to generate something it's
00:55:13
DOA and I don't think enough people ask
00:55:16
those questions or understand that
00:55:17
that's true so this is why I think you
00:55:19
have to sort of be looking at both of
00:55:21
these two things at the same time but
00:55:23
this account is interesting because it
00:55:25
kind of just strips things down to the
00:55:28
bare facts and they start to allow you
00:55:32
as a third party to understand what you
00:55:35
can do yeah speed is just such a a
00:55:38
critical component of this and what
00:55:39
Google found was as you know free
00:55:41
Brokers you were there every time they
00:55:43
lowered a certain number of milliseconds
00:55:45
the usage went up right people did more
00:55:47
searches which makes sense if you get
00:55:48
your results back faster yeah it was a
00:55:50
key metric from day one at Google
00:55:53
Marissa Mayer ran all the consumer
00:55:55
facing products at Google during this
00:55:57
you know earlier era she was like beat
00:56:01
it into the team I mean if you guys
00:56:02
remember one of the first the the the
00:56:05
first kind of early feature of the
00:56:06
Google results page was the amount of
00:56:08
time it took to load the results they'd
00:56:10
show you how many milliseconds yeah they
00:56:11
show you that yeah they they literally
00:56:12
put your Northstar metric exposed to the
00:56:15
consumer which totally that must have
00:56:16
lit a fire under the asses of all the
00:56:18
developers and server people yeah well I
00:56:20
mean they were kind of showing off the
00:56:21
quality of the infrastructure and the
00:56:22
way they did indexing and everything but
00:56:24
the result
00:56:26
really played out in usage the the
00:56:28
faster the results the more frequently
00:56:29
you would use the search engine and the
00:56:31
more likely you were to come back and
00:56:33
it's amazing how much consumer Behavior
00:56:35
drifts based on milliseconds like you
00:56:38
have a few milliseconds of delay learned
00:56:41
this right I mean if you look at the if
00:56:42
you ever see the movie the founder where
00:56:43
they explain the McDonald's process they
00:56:45
learned it too guys look at this this is
00:56:47
really interesting on this analysis I
00:56:49
mean tth are you saying that you don't
00:56:51
think open AI can achieve the necessary
00:56:53
levels of performance no I'm saying two
00:56:55
things open AI is three different
00:56:57
businesses open AI has a closed model
00:56:59
that's trained on the open internet I
00:57:01
think economically it's going to be very
00:57:03
hard to sustain that unless they start
00:57:05
buying all number of apps so that they
00:57:08
can get some fine tunes that they
00:57:10
control that are proprietary to them so
00:57:11
for example if open AI were to buy all
00:57:14
of Reddit that would be a really
00:57:15
interesting development that would
00:57:17
improve the quality of open AI in a
00:57:20
unique and differentiated way relative
00:57:23
to where things like llama Mist will get
00:57:25
to at the same time as well as X's Gro I
00:57:28
think they're all going to converge to
00:57:30
the same quality in the next probably 12
00:57:34
to 18 months that's Point number one
00:57:35
your belief there is there's enough data
00:57:37
in those pools that everybody reaches
00:57:40
parody no did you guys okay Nick did you
00:57:42
so I I published this primer on a primer
00:57:45
yeah yeah there is a slide in there Nick
00:57:47
that you can pull out but it just shows
00:57:49
you that there is a
00:57:51
converging in the quality of the results
00:57:55
as the number of the parameters of the
00:57:57
model gets higher and higher and what it
00:57:58
effectively shows you is that we are
00:58:00
already in the land of diminishing
00:58:03
returns when models are trained on the
00:58:05
same underlying data so if you are using
00:58:08
the open internet llama mistl open aai
00:58:11
they're all getting to the same quality
00:58:14
code point and they will be there within
00:58:15
the next 6 to n months so that's
00:58:18
business number one on open AI business
00:58:20
number two is a consumer facing app
00:58:22
called chat GPT that has a lot of legs
00:58:25
because I think people are you know
00:58:26
develop habits it'll be very sticky and
00:58:29
I think it'll get better and better and
00:58:31
then the third business that they're in
00:58:33
is selling Enterprise services to large
00:58:36
Fortune 500s in fact if you look at
00:58:38
their open AI day what they talk about
00:58:40
is they sell they've sold already to
00:58:42
like 94% of the Fortune 500 what does
00:58:45
that mean I think what that actually
00:58:46
means is they've sold a lot of test
00:58:48
environments and sandboxing but again in
00:58:50
order to translate that into functional
00:58:53
production code that's used by Bank of
00:58:55
America right or Boeing in
00:58:58
production you have to have Zippy Zippy
00:59:02
fast SAS and a level of performance that
00:59:05
no cloud provider yet has delivered none
00:59:09
nobody so Nick if you just go to that
00:59:12
please the thing I just wanted to show
00:59:13
you this because it's a really
00:59:14
interesting chart this is not mine this
00:59:15
is theirs if you look at quality versus
00:59:17
price saxs it starts to starts to show
00:59:20
you like where do you want to be you
00:59:22
want to be in the upper left quadr in
00:59:25
their analysis
00:59:27
right and so the point is what you can
00:59:30
see is that a ton of different models
00:59:32
are getting to the same place and so
00:59:35
obviously you'd want to use the model
00:59:36
that's the cheapest or most
00:59:39
convenient well who's going to pay for
00:59:41
that if you if you and your LPS want to
00:59:43
pay for
00:59:44
that the person that figures out the way
00:59:47
that it's the cheapest to give you the
00:59:48
same answer will actually end up winning
00:59:50
because you will run out of money and
00:59:52
they will not I don't know I mean I
00:59:54
think that there's a lot of business
00:59:55
problems inside companies where people
00:59:58
just want to very quickly set up their
01:00:01
own again custom GPT without having to
01:00:04
go
01:00:05
through the time the cost the hassle of
01:00:08
trying to do model training or
01:00:10
fine-tuning so let's just back up here's
01:00:12
the path that open AI is on so step one
01:00:16
get hundreds of millions of consumers
01:00:18
using it and getting them to view open
01:00:21
AI or chat GPT as the Google in this
01:00:24
area right strong presumption this is
01:00:26
just the one you go to when you have a a
01:00:30
question step
01:00:31
two these same people these same
01:00:34
consumers now want to use chat GPT at
01:00:36
work because there's some research they
01:00:39
want to do so openai has just rolled out
01:00:42
um both Enterprise licenses and team
01:00:44
work spaces so you can work
01:00:46
collaboratively on the same queries in a
01:00:48
team work space step three is the
01:00:51
rolling out a very easy to use Dev
01:00:52
platform that allows devel Vel opers to
01:00:55
again create custom gpts by just
01:00:58
pointing open AI at
01:01:00
repositories okay and so let's say that
01:01:03
you're the customer support team and you
01:01:07
want to create a a GPT to help customer
01:01:10
support answer cases you could basically
01:01:14
then
01:01:16
train chat GPT on let's say every
01:01:20
customer support ticket and
01:01:25
email that the company has ever produced
01:01:27
right now you could wait for the
01:01:30
company's it department to get us act
01:01:32
together and figure out how to train an
01:01:35
open source model on the same thing but
01:01:38
do you really want to wait for that or
01:01:39
do you just want to get going you know
01:01:41
and now open AI has given you the
01:01:43
Enterprise license that you need
01:01:46
to pacify the concerns about security
01:01:49
and privacy and all that kind of things
01:01:51
to some degree there's always going to
01:01:52
be those super paranoid Fortune 500
01:01:54
companies that will insist on doing on
01:01:58
owning everything and and doing it doing
01:02:00
it open source let me build on your
01:02:01
example so I run a small software
01:02:04
company during the day called hustle and
01:02:07
we saw a lot of
01:02:10
tickets related to this specific
01:02:13
legislation that exists whenever you're
01:02:15
texting or you're doing Auto dialing
01:02:18
stuff called 10 DLC and so we wanted to
01:02:22
eliminate those those tickets right so I
01:02:25
actually went and I built a
01:02:27
GPT which was called the privacy policy
01:02:29
generator because a lot of these trouble
01:02:31
tickets were because the Privacy
01:02:32
policies were bad and we trained them
01:02:36
using a handful of ones that were good
01:02:38
and a handful of ones that are bad with
01:02:40
a bunch of rules and I train the model
01:02:42
and it's wonderful except I can't run it
01:02:45
in production because it's not the kind
01:02:47
of thing that is usable in that way
01:02:51
right now it's still very difficult and
01:02:53
so all I'm saying is I'm happy to keep
01:02:55
spending a few hundred dollars a month a
01:02:57
few thousand bucks a month whatever it
01:02:59
is that I'm spending I don't quite
01:03:00
exactly know and I agree with you it was
01:03:03
very easy I think open a does an
01:03:05
excellent job of getting off the ground
01:03:08
but what I'm also saying is that when
01:03:11
you actually translate that into a
01:03:13
Mainline use case right where I want to
01:03:17
now give it to my support team and say
01:03:19
this is now a tool you can rely on it's
01:03:21
integrated into your workflow into your
01:03:23
other tools it's integrated
01:03:24
to how you pipe out data into Salesforce
01:03:27
or what have you it's just very hard and
01:03:30
I'm not saying it's not going to get
01:03:31
fixed I'm saying we're just not there
01:03:33
yet and one of the Rays in which it's
01:03:36
not there is that there is no place I
01:03:38
can go including open AI that actually
01:03:41
makes it fast enough to be usable in
01:03:43
production you wrote this on open AI
01:03:46
stack you wrote a custom GPT yeah built
01:03:48
myself yeah and you can do them on
01:03:51
hugging face now it's going to be a lot
01:03:52
of options in terms of integrating into
01:03:54
your workflows I think it's a really
01:03:55
interesting point because I saw a demo
01:03:58
somewhere where now actually I think
01:04:01
openai announced this that you can at
01:04:03
mention a custom GPT yeah yeah Sunny
01:04:06
showed me that this week on the Pod yeah
01:04:09
in in chat GPT you can now mention a
01:04:11
custom GPT to kind of invoke it yeah so
01:04:14
how it works is you would say hey I'm
01:04:16
heading to New York what
01:04:18
flights can I get at Expedia at kayak
01:04:21
whatever and then it gives you you know
01:04:23
the results here and you you you're kind
01:04:25
of pulling that up just to the point
01:04:27
about about where data advantages lie
01:04:29
and that's ultimately going to drive
01:04:32
value I cannot I've tried to think a lot
01:04:35
about this I cannot think about a better
01:04:38
data
01:04:39
advantage that is orders of magnitude
01:04:42
better than anything else YouTube
01:04:45
YouTube
01:04:46
Say it
01:04:47
is so here's here's the numbers I I
01:04:50
pulled this up you guys know like gpt3
01:04:53
and three and a half were trained with a
01:04:55
heavy waiting on common crawl which is
01:04:56
this open sourc we talked about this
01:04:58
before Gil elbaz runs it open source
01:05:02
crawling of the web the total amount of
01:05:04
data in common crawl which I think it
01:05:06
counted and I could be off on this
01:05:07
something like 40 to 60% of the waiting
01:05:10
in GPT 3 or 35 I'm off on this probably
01:05:13
so the total amount of data in that
01:05:14
common crawl data set is about 10
01:05:16
pedabytes
01:05:18
okay based on YouTube's public statement
01:05:22
recently they're seeing about 500 hours
01:05:25
a minute of video uploaded or 720,000
01:05:28
hours a day and if you assume somewhere
01:05:31
between you know just under 1080p on
01:05:33
that video we're talking about probably
01:05:36
one to two pedabytes of data being
01:05:39
uploaded to YouTube per day so if you
01:05:43
assume like over time the definition of
01:05:45
the videoos gone gotten better and the
01:05:47
amount of uploads gotten up you could
01:05:49
probably assume that there's roughly I'm
01:05:51
guessing there's probably somewhere
01:05:53
between 2,000 and 3,000 pedabytes of
01:05:57
data in YouTube growing by 1 to two
01:06:00
pedabytes per day which makes YouTube's
01:06:03
data repository 300 times larger than
01:06:06
common crawl which makes it bigger than
01:06:08
anything else that anyone else has and
01:06:10
here's the amazing thing about it it has
01:06:13
video it has image it has audio it has
01:06:16
text it has everything multi and it is
01:06:19
growing so if you were to take a a bet
01:06:22
or build a thesis around this point that
01:06:24
the data Advantage is going to drive
01:06:26
value creation if Google gets its act
01:06:28
together and leverages the data
01:06:30
repository at YouTube it is an
01:06:32
insurmountable moat that will only
01:06:34
continue to extend because the quality
01:06:36
of the YouTube experience and the
01:06:38
network effects continue to accumulate
01:06:39
for them so I think it's the most
01:06:41
valuable asset in the world today based
01:06:44
on the thesis that AI value is going to
01:06:46
ACR to the data owner I think you're
01:06:48
making such an important point this is
01:06:50
why the
01:06:51
counterfactual is is true and it's
01:06:53
actually showing up in the data and Nick
01:06:55
will show you this slide again from from
01:06:57
the AI primer but that is why we're
01:06:59
seeing these diminishing returns
01:07:00
freeberg in all of these third party
01:07:02
benchmarks of these models using the
01:07:03
same data set it's all using the same
01:07:05
data set so what we are proving is not
01:07:07
that the underlying Hardware can't scale
01:07:10
nor that Transformers are only efficient
01:07:12
to a point that's not what all of this
01:07:14
convergence is showing it's that in the
01:07:16
absence of proprietary data you're just
01:07:17
going to get to the same model quality
01:07:19
and we're seeing a bunch of different
01:07:21
models get to a very early Finish Line
01:07:24
which again if people like Facebook are
01:07:27
doing for free that's much easier to
01:07:29
underwrite because you you don't have to
01:07:31
underwrite it being a
01:07:34
differentiator in five years but if you
01:07:36
have a if you have a startup with Equity
01:07:38
value tied to a model I think it's
01:07:41
very it's much more of a tenuous place
01:07:43
to be in the absence of proprietary data
01:07:46
and everyone in the world has a camera
01:07:48
and a microphone in their pocket and
01:07:51
high-speed internet now from the phone
01:07:53
in their pocket
01:07:54
and more and more people are uploading
01:07:56
that content that that data that's being
01:07:58
generated YouTube's got this free data
01:08:00
vacuum and they just out in the world
01:08:02
and most of it's getting upload you well
01:08:04
it is public facing though so it's not
01:08:07
just true for text it's also true for
01:08:10
you know all of the image generation so
01:08:12
like if you look they can train more
01:08:13
than just an llm on it right they can
01:08:15
build all sorts of yeah go ahead no no
01:08:18
no I was just going to say like the
01:08:19
version of common crawl for training
01:08:21
these image models also exists and so to
01:08:23
your point it's like we are all
01:08:25
operating from the same brittle very
01:08:28
fixed small Quantum of training
01:08:31
information and so that is why I think
01:08:34
like
01:08:35
Facebook and Google are doing a really
01:08:38
important job by deciding that these
01:08:40
models should be free right and then
01:08:44
being able to so then the question that
01:08:46
just accentuates their data Advantage it
01:08:49
does and and I think that it allows them
01:08:51
to decide how much to leak out so for
01:08:54
example whenever like if you were using
01:08:56
a lot of Google services like GFS big
01:08:59
table big query you know
01:09:02
tensorflow the versions that you had
01:09:04
access to Via gcp was always one or two
01:09:08
generations behind what the Google
01:09:10
employees got to use right but it was
01:09:13
still so much better than anything else
01:09:15
that we could get anywhere else that you
01:09:16
would still build to those endpoints and
01:09:18
I think there's a similar version of
01:09:20
this where Facebook and Google probably
01:09:22
realize like look we'll have version
01:09:25
five running internally to optimize ads
01:09:27
and all of this other stuff that makes
01:09:29
our business that much better and we'll
01:09:31
expose version three to the public but
01:09:33
version three is still trained on so
01:09:35
much proprietary data that it's so much
01:09:36
better than version 10 of anything else
01:09:38
that's just operating on the open
01:09:40
internet right and and you know to your
01:09:43
point freeberg that's the outward facing
01:09:45
stuff YouTube is a collection of things
01:09:47
people want to share what Google also
01:09:50
has is Google Docs and Gmail things that
01:09:53
people say say privately so they have
01:09:55
another data resource there that they
01:09:57
can tap you know and and there'll be
01:09:59
regulations of privacy around that but
01:10:01
maybe there's a difference there but I I
01:10:02
honestly can't think of the quantum
01:10:04
coming close to YouTube not even close
01:10:07
well the the thing to Jason's point
01:10:08
which is really interesting is like you
01:10:10
know there's a modality in AI called rag
01:10:13
where you can actually just augment with
01:10:16
very specific training on a very
01:10:17
specific substate of of documents to
01:10:19
improve it's like a it's like a hacked
01:10:21
version of a fine tune but the Beautiful
01:10:23
about that is like if you have a Google
01:10:25
workspace my entire company runs on on
01:10:27
Google workspace in fact most of my
01:10:29
companies do at this point to click a
01:10:32
button where all of a sudden now all of
01:10:35
that stuff in all of my G drives all of
01:10:37
a sudden is trainable so that the N plus
01:10:40
first employee comes in and has an agent
01:10:43
that's tuned on every deck every model
01:10:46
spreadsheet every document that's a huge
01:10:49
Edge totally huge Edge by the way and as
01:10:52
a CEO if you gave me that choice I don't
01:10:55
think anybody underneath that reports to
01:10:57
me has any right to make that decision
01:10:59
but as a CEO I would click that button
01:11:01
instantly and I had that right as a CEO
01:11:03
and so like that's the CEO pitch it's
01:11:05
like look I can just give you these
01:11:06
agents that are that are like the next
01:11:09
version of a knowledge base that we've
01:11:11
always wanted inside of a company right
01:11:14
notion has this you know they' basically
01:11:16
you can start asking your entire notion
01:11:18
instance questions about notion which is
01:11:21
incredible and uh yeah you can just and
01:11:24
as a CEO you can see across everything
01:11:26
chth because as you know with Google
01:11:28
Docs if you're in a compliance-based
01:11:30
industry like Finance you can see
01:11:32
everything every message every email
01:11:35
every document and you can search the
01:11:37
security model and the data model
01:11:39
becomes very complicated in all of that
01:11:41
stuff like for example like how do you
01:11:43
know that this spreadsheet is
01:11:45
actually you should learn on it but who
01:11:48
gets to actually then have that added to
01:11:51
the subset of of answers right all all
01:11:53
of a sudden like salaries
01:11:56
the information gets put into the
01:11:58
training model very dangerous or subset
01:12:01
a of a companies working on a
01:12:02
proprietary chip design that they
01:12:03
actually like the way that Apple runs
01:12:06
highly highly segregated teams where
01:12:08
nobody else can know so there's all
01:12:10
kinds of complicated security and and
01:12:12
data model and usage questions there but
01:12:14
yeah BR new world so there's been a lot
01:12:16
of discussion real estate you you shared
01:12:18
a video with us why don't you kick it
01:12:19
off for us here preber what's going on
01:12:21
in commercial real estate and saxs
01:12:22
you've got Holdings and a lot of as well
01:12:24
so let's kick up the commercial real
01:12:26
estate challenges of the moment well I
01:12:29
mean I think we're teeing off of Barry's
01:12:31
comments at this event last week he and
01:12:34
I met backstage because I spoke right
01:12:36
before him and then he gave this talk
01:12:39
which is available on YouTube where he
01:12:41
talked about the state of the commercial
01:12:42
real estate market and particularly he
01:12:44
talked about the office Market just to
01:12:47
take a step back to talk about the scale
01:12:50
of commercial real estate as an asset
01:12:52
class in the US Nick if if you'll pull
01:12:53
up this chart the total estimated market
01:12:56
value of commercial real estate in the
01:12:58
US across different categories is about
01:13:01
$2 trillion with about $3 trillion being
01:13:04
in the office Market which is
01:13:06
specifically what he was talking about
01:13:07
he was saying that in the US we're
01:13:09
seeing people not coming back to work
01:13:12
and all these offices are empty and
01:13:13
we've talked a lot about these offices
01:13:14
being written down so how significant of
01:13:16
a problem is this so $ 20 trillion asset
01:13:19
class obviously the multif Family Market
01:13:21
is probably not as bad as office and
01:13:23
retail which are the most heavily
01:13:25
affected Each of which are about $3
01:13:27
trillion a piece the rest of these
01:13:29
categories seem relatively
01:13:32
unscathed in comparison industrial
01:13:34
Hospitality healthare you know those
01:13:37
those real estate sectors are probably
01:13:39
pretty strong data Cent is obviously
01:13:40
growing like crazy Self Storage is a
01:13:42
great Market if you pull up the next
01:13:44
image so it turns out that of the 20
01:13:47
trillion do of market value there's
01:13:49
about $6 trillion of debt so you can
01:13:52
kind of think about that 20 trillion
01:13:53
being 6 trillion owned by the debt
01:13:57
holders and 14 trillion by the equity
01:13:59
holders and the debt is owned roughly
01:14:03
50% by Banks and thrifts and this was
01:14:06
this concern that we've been talking
01:14:08
about with higher rates is the debt on
01:14:09
office actually going to be able to pay
01:14:11
the debt on retail going to be able to
01:14:12
pay when half of that debt is held by
01:14:14
Banks and thrifts that as we' talked
01:14:16
about have such a close ratio to
01:14:20
deposits that you can actually see many
01:14:22
banks become technically insolvent if
01:14:24
the debt starts to default Barry's point
01:14:28
that he made was if you look at the
01:14:29
office Market which you know is marked
01:14:32
on everyone's books as $3 trillion of
01:14:35
market value he thinks it's probably
01:14:37
worth closer to 1.8 trillion so there's
01:14:40
$1.2 trillion of
01:14:42
loss in the office category and if you
01:14:46
assume 40% of that 3 trillion is held as
01:14:49
debt you're talking about $1.2 trillion
01:14:51
of office debt a reduction from 3
01:14:55
trillion to 1.8 trillion means that the
01:14:58
equity
01:14:59
value has gone down from 1.8 trillion to
01:15:02
600 billion so they've lost Equity
01:15:05
holders in office real estate have
01:15:07
probably lost 2third of their value
01:15:10
two-thirds of their investment and who
01:15:12
owns all of that most of that 60 plus
01:15:16
perent call it 2third of that is likely
01:15:19
owned by private Equity Funds and other
01:15:21
institutions where the end benefici is
01:15:23
actually Pension funds and retirement
01:15:25
funds and so if 2third of the value has
01:15:28
to be written off in these books and it
01:15:29
hasn't happened yet what's going to
01:15:31
happen to all these retirement funds and
01:15:32
this is where going back to my
01:15:34
speculation a couple months ago kind of
01:15:36
gets Revisited if you're actually
01:15:37
talking about a two-third right down on
01:15:39
the value in these funds most of that
01:15:41
being Pension funds you're not going to
01:15:43
see governments let that happen you're
01:15:45
going to see the federal government yeah
01:15:47
it's there's going to be some action at
01:15:49
some point and it's unlikely the office
01:15:52
Market is going to m rebound overnight
01:15:54
if this stays the way it is who's going
01:15:57
to fill that hole for retirees and
01:15:59
pensioners because we're not going to
01:16:00
let that all get written down someone is
01:16:02
going to step in and say we've got to do
01:16:03
something about this and there's going
01:16:05
to need to be some sort of structured
01:16:07
solution to support retirees and
01:16:09
pensioners because that's ultimately who
01:16:11
ends up holding the bag in this massive
01:16:13
write down he didn't go all the way
01:16:14
there in his statements he was talking
01:16:15
more about his estimate of 3 trillion to
01:16:17
1.8 trillion and then I tried to connect
01:16:19
the dots and what that actually means
01:16:21
and ultimately there's going to be some
01:16:23
pain felt by retirement funds that's
01:16:25
going to need to be dealt with somehow
01:16:27
so SX I don't know if that if that sits
01:16:29
right with you I mean I think the big
01:16:30
picture is right I think you're applying
01:16:32
a lot of averages right I think in the
01:16:35
office Market in particular the typical
01:16:37
office deal is more like onethird equity
01:16:39
and twoth thirds debt there's just a lot
01:16:41
more leverage right so that' be Point
01:16:44
number one which makes the situation
01:16:45
worse even worse yeah so I would say
01:16:48
that there's a huge amount of equity
01:16:51
that's been written off but in addition
01:16:52
to that there's a lot of debt
01:16:54
holders who are in trouble too and that
01:16:58
debt is is held by Regional Banks so
01:17:01
these commercial loan portfolios are
01:17:04
significantly impaired that's what we
01:17:05
saw with Community Bank of New York is
01:17:07
that their stock cratered when they
01:17:10
reported higher than expected losses in
01:17:12
their commercial real estate
01:17:14
portfolio so freeberg I think the point
01:17:17
is just the the pain from this is not
01:17:20
just going to be on the Equity holders
01:17:23
but also on these Banks which can't
01:17:26
afford to lose it's not evenly
01:17:27
distributed yeah right yeah right and we
01:17:31
saw this in San Francisco where some of
01:17:32
these buildings have 70% debt to equity
01:17:34
ratios and you know the the value puts
01:17:37
them in the hole and equities wiped out
01:17:39
completely and the debt holders have to
01:17:40
take a hit and normally you know that
01:17:42
debt is not really written off very
01:17:44
often it's well this is why that the
01:17:46
debt holders the debt holders don't want
01:17:48
to foreclose they don't want to get
01:17:49
these buildings back because when they
01:17:51
do they're going to have to write down
01:17:52
the low loan as long as the loan is
01:17:55
still outstanding and they haven't
01:17:56
foreclosed they can pretend that the
01:17:58
value of the building is not impaired
01:18:00
Kick the Can down the road is the best
01:18:02
strategy for them so it's it's called Uh
01:18:04
pretend and extend so what they'll do is
01:18:06
they'll work out a deal with the the
01:18:09
landlord the equity holder that the
01:18:11
equity holder would say listen I can't
01:18:13
pay the interest so they'll just tack on
01:18:14
the interest basically as principle at
01:18:17
the end of the loan and they'll extend
01:18:19
out the term of the loan which would
01:18:21
wipe out the equity at a certain point
01:18:22
yeah
01:18:24
well what it does it allows the equity
01:18:26
holder to stay in control own the
01:18:28
building right because yeah the equity
01:18:30
holder can't pay make their debt
01:18:32
payments today but they're going to
01:18:34
postpone those debt payments till the
01:18:36
end of the of the loan and again in the
01:18:39
meantime just kind of hope that the
01:18:41
market could that debt at some point
01:18:42
since they have so little equity in
01:18:43
these buildings typically just exceed
01:18:46
the value of the of the property and
01:18:48
it's like I'm just working for the bank
01:18:50
now and yeah why am I even putting this
01:18:52
working because everyone kind of hopes
01:18:54
that the market will recover the value
01:18:56
of their Equity will go up and they'll
01:18:58
be able to make their debt payments
01:19:00
again yeah so if you're the equity if
01:19:01
you're the equity holder you'd rather
01:19:03
hold on and have a chance of your Equity
01:19:05
being worth something in recovery then
01:19:07
definitely lose the building and if
01:19:09
you're a Regional Bank you'd rather
01:19:12
blend an extend or pretend an extend as
01:19:14
opposed to having to realize the loss
01:19:17
right now yep and showing the market
01:19:20
that your solvency may not be as good as
01:19:22
you thought the same thing happened with
01:19:25
government bonds remember that with svb
01:19:27
and these other Banks they had these
01:19:28
huge held to maturity port bond
01:19:31
portfolios y these are main mostly just
01:19:34
uh t- bills that were worth I don't know
01:19:37
60 cents on the dollar when interest
01:19:38
rates spiked from 0 to 5% but they
01:19:42
didn't have to recognize that loss as
01:19:44
long as they weren't planning to sell
01:19:47
them right and then when they had the
01:19:49
bank run they had to sell well yeah
01:19:51
that's right so when depositors left
01:19:52
left because they needed their money or
01:19:55
because there was a run or because they
01:19:57
could get higher rates in a money market
01:19:58
fund all of a sudden these Banks had to
01:20:01
sell their heal to maturity portfolios
01:20:03
they had to recognize that loss and
01:20:05
that's when everyone realized oh wait a
01:20:07
second they're not actually solving okay
01:20:08
so jamat Supply demand matters in real
01:20:11
estate we have a Tail of Two Cities here
01:20:13
on one side in real estate for
01:20:15
commercial real estate no demand for
01:20:17
office space which uh is and way too
01:20:20
much Supply paradoxically on the other
01:20:23
side we have this incredible market for
01:20:25
developers which is gosh there's not
01:20:28
enough homes I think we need 7 million
01:20:29
more homes and the demand is off the
01:20:32
charts for homes yeah yeah I mean I
01:20:34
think I think you're basically right
01:20:35
it's not I keep trying to explain
01:20:36
residential is not a great Market either
01:20:38
because interest rates have spiked up so
01:20:40
there's not a vacancy problem multif
01:20:42
family developers are still able to
01:20:44
lease the units they're still a able to
01:20:47
rent the problem is their financing
01:20:49
costs have shot through the roof so
01:20:52
again let's say you were a developer who
01:20:53
built multif family in the last few
01:20:55
years you took out a construction loan
01:20:58
that construction loan might have been
01:20:59
at 3 4% yeah now you want to put
01:21:02
long-term financing on it but if you can
01:21:05
even find debt right now because there's
01:21:07
a credit crunch going on you may have to
01:21:08
pay 8 nine 10% yeah but at least you can
01:21:11
find a renter you can find a renter
01:21:14
that's true but only at a certain price
01:21:16
and let's say you underwrote that
01:21:17
property to I don't know like a five cap
01:21:20
like a certain yield yeah but now your
01:21:22
fin costs are much higher than you
01:21:23
thought you might be underwater yeah
01:21:26
meing that situation isn't as bad as
01:21:29
what's happening in why I think it's I
01:21:32
think it's worse in some ways if you're
01:21:34
fully if you're fully
01:21:36
rented and your building is underwater
01:21:38
because now your debt payments are much
01:21:40
higher than you expected then there's no
01:21:42
business model yeah but are we seeing
01:21:44
that are we seeing tons of multi family
01:21:45
go under can I make two points one I
01:21:48
think I think David is is Right which is
01:21:50
that I don't know this Market very well
01:21:52
but just just as a as a bystander here's
01:21:55
what I observe it seems that the
01:21:56
residential Market has a
01:21:59
feature and I don't know whether it's
01:22:01
good or bad but that feature is that you
01:22:04
repic to market demand every year so to
01:22:08
the extent that Supply demand is
01:22:10
changing and default rates are up or
01:22:12
whatever that's reflected in rents and
01:22:15
you see that because rents change very
01:22:17
quickly and most human beings are
01:22:19
signing six-month to oneyear leases so
01:22:21
that reset happens very quickly so it
01:22:23
can more dynamically adapt so to the
01:22:25
extent that a market segment is impaired
01:22:28
you see the impairment quickly on the on
01:22:31
the office side what I see is that
01:22:33
there's been a structural Behavior
01:22:35
change in covid that has reset in every
01:22:39
other part of the world except for the
01:22:41
United States where there are these
01:22:43
frankly typically younger typically more
01:22:46
Junior employees that have held many of
01:22:49
these companies hostage in a bid to
01:22:51
return back to office office space and
01:22:53
so we know that there is this vacancy
01:22:55
Cliff that's going to hit commercial
01:22:57
real estate we just don't know when
01:22:59
because there are they're in long-term
01:23:01
leases they're canceling these leases
01:23:03
over long periods of time so the reset
01:23:05
cycle is longer that's just my
01:23:06
observation as an outsider I don't know
01:23:08
what that me for for prices or anything
01:23:10
else but it just seems that at least the
01:23:12
residential Market can find a bottoming
01:23:14
sooner because you can reset prices
01:23:16
every year but commercial just seems
01:23:18
like a melting ice Direction correct to
01:23:21
you SX that assessment commercial has
01:23:25
both a demand problem and a financing
01:23:27
problem multif family just has a
01:23:28
financing problem but it's important
01:23:30
office we're talking about office
01:23:32
there's retail and then there's office
01:23:34
and then there's other industrial you
01:23:36
see in China China has 50 million
01:23:39
homes ahead of schedule 50 million
01:23:42
additional Supply that can house 150
01:23:44
million people so as acute as our issues
01:23:46
are the China issue might be much much
01:23:50
yeah seismic can let me just give you an
01:23:51
example on the multif family side okay
01:23:53
let's say that you buy a building okay
01:23:55
let's say you bought a building in 2021
01:23:57
the absolute peak of the market and you
01:24:00
could get debt at say 4% okay and you
01:24:03
penciled out let's call it a 6% yield
01:24:07
that with the debt you getting so let's
01:24:09
say you did 2/3 debt at 4% you could now
01:24:13
lever up that 6% yield to 10% okay
01:24:17
that's like sort of the math right now
01:24:20
all of a sudden and and to get there
01:24:21
you'd have to do some added work on the
01:24:23
property you have to Spruce it up okay
01:24:25
now it's a few years later and your
01:24:28
short-term financing is running out and
01:24:31
you need to refy and you've done your
01:24:32
value added work but here's the problem
01:24:35
the overall valuations in the market
01:24:37
have come way down so before the bank
01:24:41
was willing to give you 2third loan to
01:24:43
value now the values come way down you
01:24:45
may not even be able to get two-thirds
01:24:47
loan value so you're going to have to do
01:24:48
what's called an equity in refinancing
01:24:51
you're going to have to produce more
01:24:53
Equity you're going have to Pony up more
01:24:54
money so instead of taking Equity out
01:24:56
like when the deal goes well you're
01:24:57
going have to put equity in you may not
01:24:59
have that Equity if you're the developer
01:25:01
the other thing is that your financing
01:25:03
cost now might be 10% so now you've got
01:25:06
negative leverage you're generating a 6%
01:25:09
yield but you're borrowing at 10% to
01:25:11
generate that 6% yield so the debt no
01:25:14
longer makes sense you're again you're
01:25:16
not positively leveraged you're
01:25:17
negatively leveraged so you're not going
01:25:19
to want to take out that debt and if you
01:25:21
do take out that debt
01:25:23
the the building's going to be
01:25:24
underwater it's not going to be
01:25:25
generating net operating income it's
01:25:27
going to be generating losses so that's
01:25:30
why even categories like multif
01:25:34
family where you don't have a vacancy
01:25:36
problem there's strong demand yeah those
01:25:39
properties still don't make sense if you
01:25:41
had long-term debt on your multif family
01:25:43
if you were able to lock in that 4% loan
01:25:46
for 10 years you're fine but for all the
01:25:49
people who are refinancing now who are
01:25:51
coming up this year last year next year
01:25:54
they're in deep trouble and that's why
01:25:56
there's a rolling crisis in real estate
01:25:58
is because the debt rolls over time it's
01:26:01
not like everybody hits the wall and has
01:26:03
to refinance at the same time well that
01:26:05
God right I mean this would be
01:26:06
cataclysmic if if it was if everybody
01:26:09
can you imagine if Silicon Valley and
01:26:11
San Francisco had to say here's actually
01:26:12
the reality anybody want to actually pay
01:26:14
for this office all in the same year
01:26:17
right that would be insane but the
01:26:20
crisis is growing is as the leases roll
01:26:23
and those old rents that were higher the
01:26:25
market roll off and now you have to take
01:26:27
on new leases if you can even get them
01:26:30
it's going to be Bal at a much lower
01:26:32
rate and as the old loans roll that were
01:26:34
at a much lower interest rate you have
01:26:36
to get financing even if you get it at a
01:26:38
much higher interest rate that's when
01:26:40
all of the sudden these buildings go
01:26:41
from being basically solvent to
01:26:44
insolvent yeah I mean Janet yellen's
01:26:47
just going to bail these folks out I
01:26:48
mean you won't bail out the banks
01:26:49
themselves but you'll bail out the
01:26:50
creditors obviously the people holding
01:26:53
the bag they'll get bailed yeah that's
01:26:54
everybody agrees Janet yelling yelling
01:26:58
our treasury secretary I don't know if
01:27:00
she's going to be the one to do it I
01:27:02
there's going to be congressional action
01:27:03
on this stuff yeah I mean they tend to
01:27:07
lead it so all right for the Sultan of
01:27:10
science David freeberg and David saaks
01:27:13
and chamar poaa the chairman dictator I
01:27:16
am the world's greatest moderator we'll
01:27:17
see you next time on the Allin pod
01:27:19
bye-bye
01:27:21
bye-bye will let your winners
01:27:29
ride we open source it to the fans and
01:27:31
they've just gone crazy with it love
01:27:34
queen
01:27:37
[Music]
01:27:41
of
01:27:43
[Music]
01:27:44
Besties my dog taking your
01:27:47
driveways man oh man my habiter will
01:27:51
meet
01:27:52
we should all just get a room and just
01:27:54
have one big huge orgy cuz they're all
01:27:55
this useless it's like this like sexual
01:27:57
tension but they just need to release
01:27:59
[Music]
01:28:05
somehow we need to get merch are
01:28:10
[Music]
01:28:14
all I'm
01:28:17
[Music]
01:28:19
going
