00:00:00
all right everybody welcome back to the number one podcast in the world i'm your
00:00:05
host and executive producer for life isn't that right Dave Freeberg Jay Cal
00:00:10
Jason not at all what you are make sure you tune in startups and apply to
00:00:15
Founder University you're something very different with us again today the Sultan of Science David Freriededberg
00:00:24
can I just congratulate you on your fourth baby if you double that number you're going to be able to catch up to
00:00:29
Chimoth and his five plus three illegitimate how are you doing
00:00:35
[Music] let your winners ride
00:00:40
[Music] we open sourced it to the fans and
00:00:46
they've just gone crazy with it
00:00:52
how you feeling you're tired and grumpy aren't you you're a little transition for me i didn't have to do the work it's all Are you tired and grumpy and how's
00:00:59
Allison how's How's the Everyone's wonderful thank you for asking and a beautiful boy beautiful is crushing
00:01:05
nothing more amazing than seeing a child how's magnificent magnificent thank you
00:01:12
for asking thank you for asking that yeah okay let's move on thank you thank you for all the kind words and uh just
00:01:19
we sent over a gift basket Chimath and I longhorn Pana Stakes uh a 10-year uh
00:01:26
membership for Oh hey congrats to Olivia Landon by the way of Long Hill Wagyu she had twins
00:01:33
that means she's going to have more people to work on the ranch and slaughter cattle to send us our pana congratulations shout out congrats to
00:01:40
Olivia Landon it's so funny cuz we love this we love these steaks so much she
00:01:45
doubled we mentioned it on the pod and you idiots started like searching for it lunatics and they ordered out all the
00:01:50
Koolette steak so now Chant and I are screwed no crew no they ordered out everything everything was sold up
00:01:57
everything was sold up so now we have to gatekeep with us again your chairman dictator Chimath Polyhapatia he of two
00:02:04
votes in our fine organization how you doing Chimath i love voting control i'm
00:02:10
doing great he starts Thomas Lefont with a uh tie and then all of the
00:02:17
gamesmanship happens between the team of rivals me and Freeberg with us again Thomas Lefant
00:02:23
a gentleman a scholar no idea why he's here a true I don't know how he wound up
00:02:28
on this podcast but a true gentleman a true scholar and host of Easts meets
00:02:34
West an incredible conference that I attended this week with our besty David
00:02:40
Saxs who of course is at the White House and can't join us uh but Thomas what a great event thank you for including me
00:02:47
no box lunches by the way we we took your feedback from a couple of years ago so I hope that we met your standard you
00:02:54
did up highlights for you guys at your conference Thomas i mean I think for me obviously I think a lot of news in AI
00:03:01
this week so I think that was kind of the center piece of most of the panels
00:03:06
pretty much up and down the stack from SAS companies trying to transform into AI to obviously the big Zuck news on
00:03:14
scale and then potentially I saw in the information today the the Nat Friedman news so it feels like there's a lot
00:03:20
going on in the industry so should be fun to talk about yeah and we're going to talk about it all today we got a a really full docket rick Caruso the uh
00:03:29
mayor who would have saved Los Angeles from the fires he was there and you actually hosted at his incredible
00:03:35
facility what a we did we talked about the the state of LA which JCL is that is
00:03:41
that it looks like that's where you're at right yes i'm at my uh LA home which
00:03:46
uh aka the compound uh yeah it's uh it's available on Airbnb so I'm here in LA
00:03:52
but yeah Rick Caruso what a great speaker interestingly Jake Al today a friend just sent me a chart showing the
00:03:59
recovery of restaurants postco andif uh LA is 50% behind on the
00:04:06
recovery per store location versus the national average what do you attribute that to or what did they attribute it to
00:04:13
i think I think there's kind of a couple different things right and I think one the economy which you know unlike the
00:04:20
San Francisco economy being levered to to AI and on the upswing is more levered
00:04:25
to entertainment and I think you know secular decline I think you know someone mentioned at the conference
00:04:32
that filmmings in LA are down 50% from peak so I mean that's just a a massive
00:04:39
move down losing share to other geos both in the US I think Georgia right Jay
00:04:44
Cal was mentioned I mean Ted Cerrone has explained exactly how aggressive New York is being uh the UK is being Atlanta
00:04:52
I mean so many different hubs for movies giving much better deals than Los Angeles is yeah so I think it's a it's a
00:04:59
combination of I think you know being levered to one industry that's kind of
00:05:05
in secular decline i can tell you for Mr beast that for Beast Games we had a deal
00:05:10
in Las Vegas and in Toronto we got huge tax credits and in the second season
00:05:15
that we're doing for Amazon we did an enormous deal with the Kingdom of Saudi Arabia and so we're filming a
00:05:23
bunch of episodes there we're building the sets there we're actually going to keep them there after it's all said and done we would not film in Los Angeles
00:05:31
unless we absolutely had to we will stay as far away from California as possible and regulations are such a big part of
00:05:38
this it's on economic you can't make it work yeah 30% more expensive I think it's is the kind of the official number
00:05:44
on on well there's also speed right Thomas like how quickly can you stand something up how many how much paperwork
00:05:50
do you have to file James Beard Foundation I'm seeing here from the research has found that all
00:05:56
these independent restaurant owners said they just can't get staff here so in Los Angeles it's just hard for people to
00:06:01
live here and it's hard to get through the regulations and if you make it hard there are other options for people this
00:06:07
idea that California has a lock on uh anything other than incredible weather and beautiful people is Farsome there's
00:06:14
a lot of beautiful people in other places with decent weather and you can you can go do your projections there so
00:06:19
another topic that came up that a lot of people were talking about something that I know you've talked a lot about our our
00:06:26
debt issue and the debt to GDP ratio there was a lot of talk on the on the
00:06:32
flip side on the GDP side what if actually AI can increase productivity and regrow GDP faster than expectations
00:06:41
right and perhaps that's one of the reasons why you know interest rates might not be quite as high as you might
00:06:48
expect given some of the trends that you guys have talked about so I think a lot of a lot of discussions around
00:06:55
AI productivity and what we could look at over the next you know five to 10 years because of the the improvements
00:07:01
we're seeing this is particularly beneficial to the US right i mean if you
00:07:06
think about where AI is going to acrue economic surplus first it's likely going to be in the US not global GDP so the US
00:07:15
kind of does it compete dollars or it increases overall productivity or both ahead of the rest of the world if we do
00:07:22
see advances from AI to accelerate GDP growth is that because of all of the
00:07:27
onshoring of manufacturing and industry that we outsource today like do you
00:07:33
think that that goes handinhand with AI acceleration i think that's part of
00:07:38
it and I think the other part is just getting even out of the you know the knowledge worker workforce right just
00:07:46
getting significant productivity productivity improvements there one of the things that we showed in our keynote
00:07:53
is the adoption of these technologies and even taking doctors as an example right an area you know well you know
00:08:00
this new company um kind of coming in and and developing kind of a diagnosis kind of engine right
00:08:08
that's now used by a third of doctors so you know I I think that uh it's open
00:08:15
evidence by the way is the name of the company and already a third of US physicians are on the platform using it
00:08:22
you know 10 times a day to kind of help diagnosis so in particular in oncology as an example it's seen significant
00:08:28
traction so you know you multiply that by the legal profession coding i think
00:08:34
we're already seeing you know what if we just see kind of a an explosion of productivity gains across you know both
00:08:40
the physical and the digital economy yeah the doctor one's a good example if someone had the opportunity to go get
00:08:46
more regular preventative checkups um they would the problem is it's very
00:08:51
expensive it's hard to get an appointment or insurance won't cover it but if the cost to a doctor goes down
00:08:58
because they can leverage AI the throughput goes up by 10x they can see 10 times as many patients per day then
00:09:04
suddenly diagnostic care becomes more available they can charge for that they don't need to charge the same amount the
00:09:10
price will come down per checkup but you'll more people will be able to get a checkup per day so that grows GDP in
00:09:17
diagnostic care that grows the size of that piece of the economy it's a very good example give you by the way
00:09:23
anything where AI provides leverage to a service provider where their throughput now goes up um I'll give you another
00:09:29
example of that um Dave uh so there was an LA dentist that kind of hit got viral
00:09:35
this week i don't know if you guys saw this story but basically he um he
00:09:40
created an ad using V3 about a skydiving gorilla yeah I saw that who you know ultimately needs to
00:09:47
get his teeth fixed because he was drinking while he was jumping out of the plane and you know it's a very kind of funny viral ad he probably made it for a
00:09:55
couple you know hundred bucks and now his practice is totally full he's been flooded with requests right for the new
00:10:02
dental implants so you know to your point about increasing productivity boom there's how how V3 can help a local
00:10:09
dentist all right everybody welcome to the number one podcast in the world we got a full docket full docket but we're
00:10:17
going to rocket the docket because there's so much going on here zuck is tilted clearly uh this has been the big
00:10:23
discussion in Silicon Valley for the last 10 days or so according to reports
00:10:29
Zuck is super frustrated that Meta is falling behind in AI so he is swinging
00:10:34
for the fences sam Waltman said Meta has offered top Open AI employees a $100
00:10:39
million wait for it signing bonus that's not comp that's a signing bonus who knows if this is true or not but he's
00:10:46
also offering 100 million a year in annual comp he's clearly cut out tens of billions of dollars for this effort not
00:10:53
dissimilar to when he did his VR efforts that didn't work out so well here's a 30
00:10:58
secondond clip of Sam Alman talking about this on his brother Jack's podcast Uncapped they started making these like
00:11:04
giant offers to uh you know a lot of people on our team um you know like $100 million signing bonuses more than that
00:11:11
comp per year crazy and I'm actually It is crazy i'm really happy that at least
00:11:16
so far uh none of our best people have decided to take them up on that i think that people sort of look at the two
00:11:23
paths and say all right OpenAI's got a really good shot a much better shot at actually delivering on super
00:11:28
intelligence uh and also may eventually be the more valuable company meta just also vested over 14 billion I'm using
00:11:35
invested in quotes in scale AI for 49% stake and uh this probably is better
00:11:41
described as a shadow aqua hire to get around antitrust scrutiny you remember Microsoft did that with Inflection AI
00:11:46
back in the day google did it with Character AI and Amazon did it with Adept AI i'm not sure if this is necessary anymore uh since Lena Khan's
00:11:53
no longer in the position scale CEO Alexander Wang and others will be joining Meta to work on a new super
00:11:59
intelligence team they're saying that Scale is going to remain an independent company and get a new CEO not sure if that's going to happen
00:12:06
and if you don't know uh Scale does data labeling they get experts to help train
00:12:11
language models two of their biggest customers are OpenAI and Google and they both canled their contracts so Zuck is
00:12:18
taking that chess piece off the board so he can get all that data into his LLMs
00:12:23
he's also reportedly in talks to hire former GitHub CEO Nat Friedman and Daniel Gross to work on AI they have a
00:12:30
incubator investment fund for AI daniel Gross had a really cool startup
00:12:36
incubator called Pioneer Labs i had him on this week in startups a couple years ago really smart cat meta has 70 billion
00:12:42
in cash thomas Lefant when you see Zuck doing this what's your take not only on
00:12:48
what Zuck's doing but how big of an opportunity is this you know in terms of
00:12:54
the prize of having the best large language model what is he going for here and uh what's your take on these really
00:13:02
aggressive packages and 49% purchases i mean look I think one it it feels
00:13:08
highly rational right if you think about Meta's market cap is uh rough math 1.7
00:13:13
trillion if you're the CEO and you ultimately believe that maybe 50% of your market cap is at risk because of AI
00:13:21
850 billion why would you not spend maybe four or 5%
00:13:26
of that if you think it increases the odds even slightly that you're going to win the market so to me it it kind of
00:13:33
reminded me of a few few things number one the scale and size of the
00:13:38
opportunity right obviously people think AI is massive but frankly um Jake I'm
00:13:44
even wondering putting the regulatory scrutiny to the side if it was time he just didn't want to wait and obviously
00:13:49
doing it this way I think Alex literally the next day who's the co of scale can show up to work at Meta so I think it's
00:13:56
it's urgency of a large opportunity um I'm curious to get Chamas's take because
00:14:01
it reminded me a little bit of the pivot away from HTML 5 and also So a a much
00:14:07
smaller acquisition but one that we really felt which was of a company called Onavo
00:14:12
and for those that may not remember Onavo was a small data service provider but what it did is it had a panel of
00:14:18
phones and we as investors could see what people which apps people were using
00:14:24
and the data was incredibly valuable because it was the only service that gave you true engagement data and so
00:14:30
obviously as an investor you felt wow this is an incredible tool and eventually it sold to to Facebook and
00:14:37
Facebook used it internally and didn't allow anybody else to use it and we lost one of our key abilities right in the
00:14:44
mobile app revolution to tell who was winning and losing so um and you're saying the scale acquisition is you know
00:14:51
uh parallels that in a bit there's this great service a lot of people rely on it he buys it shuts it down for everybody
00:14:57
else gets the tool for himself gets the data for himself correct so I definitely see parallels and I think given this you
00:15:03
know their market cap and the size of this opportunity I think it makes a lot of sense shimath your thoughts on Zuck's
00:15:10
action obviously folks know you worked with him as you went from tens of millions of Facebook users to hundreds
00:15:16
of millions and you were there actually during the uh HTML rapper app disaster
00:15:22
uh that uh I think maybe that was a debate at our executive team at our M team and I was on the side of apps and
00:15:30
well without embarrassing him somebody else was on the side of HTML 5 i thought it was [ __ ] stupid why why why was
00:15:37
that but that decision one because you know all of my political capital at the time was also wrapped into native apps
00:15:45
our own phone an entire verticalized integrated stack and politically
00:15:52
I think I made the decision for them very hard because I was not a very
00:15:57
play nice in the sandbox with others kind of executive i was more of a scorched earth get it done kind of
00:16:02
person okay so no changes over the last 15 years that's good to know they made they made an enormous mistake but then
00:16:08
they admitted it about a year after I left they said "This was the single sucks at Explain in plain English why
00:16:13
HTML 5 rappers versus native apps." I can't cuz it's [ __ ] okay great uh I
00:16:19
can explain it so like native apps are was obvious in 2010
00:16:26
and the only the only reason to use HTML was as an endound for different carriers
00:16:33
and for different ecosystems that were trying to charge us a toll so in 2010 I
00:16:39
went to Mobile World Congress and I took a group of my most talented developers
00:16:45
and we built an entire replica of Facebook that we called Facebook zero which was only available via URL and we
00:16:51
launched it at Mobile World Congress and we did it and I announced it there because if you went to India as an
00:16:57
example all of the folks there would try to charge us a tax but if you could
00:17:04
navigate through the browser you wouldn't have to pay it right so that was a good example of what to do in a
00:17:09
developing market when people were toll taking but the real solution was to build an extremely integrated app from
00:17:15
the software all the way to the hardware and the only way to do that was as a native application and that has
00:17:21
tremendous applications to today but just to finish on yesterday my proposition was full phone full stack
00:17:28
full app all of this other HTML stuff should only be as a side thing that we do in markets where they try to make it
00:17:35
difficult for us instead it became politicized and it became a big bet on HTML 5 which I thought was absolutely
00:17:40
stupid and unjustifiable and that was also when I said "Okay well this phone's not going to happen so let
00:17:47
me leave." And a year later I think Mark to his credit said "This was really stupid." And ripped all the HTML 5 stuff
00:17:53
apart went native and the rest is history so let's fast forward to today
00:17:58
yeah there it is biggest mistake was betting too much it was It was an And that was again I'll just say it people
00:18:05
politicizing what should have been an obvious technical decision okay the other piece to that just to add to it
00:18:10
was it was also a religious decision then people liked the open standards of HTML 5 so there was certain developers
00:18:17
who felt like we have to support open stupid people thought that only stupid nontechnical people thought that it was
00:18:23
stupid it was obvious you'd have to be a [ __ ] [ __ ] and there were [ __ ] morons at the executive team that
00:18:29
advocated for this anyways we were right they were wrong and he was fine okay fast forward to the where are we today
00:18:36
it's the exact same story playing out now what do I mean you have to look very carefully at Microsoft's deal with
00:18:43
OpenAI why because what you see is the compounding of secrets there are secrets
00:18:50
in the training layer there are secrets in the model layer there are secrets in how these things are tightly coupled to
00:18:57
infrastructure and compute and what we have to remember is what Open AI got
00:19:02
from Microsoft was an extremely competent partner that built an enormous
00:19:08
Azure compute infrastructure to train everything from chat GPT all the way up
00:19:14
to the 03 model everything why is that important because you start to figure
00:19:20
out these tricks how do you really optimize these models to be extremely
00:19:25
performant and now if you look at all of the other models they've also had some
00:19:31
level of that advantage so if you look at Deep Seek what did they do well we don't know but what we have been told is
00:19:37
that there's very tight coupling to hardware if you look at what XAI is doing I think what you can bet is that
00:19:44
there's an extremely tight coupling to hardware and infrastructure and compute
00:19:49
if you look at what Facebook is doing they generically train on Nvidia and they launch it in the open source so I
00:19:56
think that what they need to do is more of the open AI more of the Google
00:20:02
playbook look at Google google's Gemini models are extremely tightly coupled to
00:20:07
TPU and it enables and unlocks an entire stack of secrets and capability that
00:20:14
then get manifested in model quality so I think the first thing that Mark has to do if I were him is start to chip away
00:20:22
at all of the sets of secrets so what secrets do you get from Alexander Wang
00:20:28
and scale it's what are the labeling techniques that allow these models to be
00:20:33
more and more performant what labeling techniques are used in the reasoning models what labeling techniques are used
00:20:38
in more traditional LLMs it is clear that Llama doesn't know this meta doesn't know this that well because
00:20:44
their model quality is meh so now what you get is that set of secrets so what do you get from Nat Freeman and Daniel
00:20:50
Gross you get what are the apps doing how are they approaching writing agents these agentic tips and tricks that make
00:20:57
usability and value more obvious but then what's missing i think the
00:21:02
thing that's missing is the infrastructure and compute set of secrets i think it's insufficient to buy
00:21:08
stuff off the shelf from Nvidia and expect these models to fundamentally compete so I think if I were a betting
00:21:14
man he's bought the training secrets he's bought the app secrets and now he has to buy some infrastructure and
00:21:20
compute hardware secrets you put it together and he's got a pretty good strategy here and also just to add to
00:21:25
that Shimothnat and Daniel have invested in a lot of AI companies and those companies are have
00:21:32
secrets of their own yeah and those are and actually I think they have some along the full stack freeberg your
00:21:37
thoughts on this strategy as described by Thomas and Shimoth and just the data
00:21:42
we're seeing on the playing field aggressive acquisition of talent and companies
00:21:49
i don't know if I have much to add here okay one additional point Shimoth by the way that you mentioned if we look at the
00:21:54
winners right in models of the past 12 months anthropic the same right they've been very um kind of deliberate and have
00:22:01
explained how TPUs right they've been a big user of them how it's helped define their training models so I think you're
00:22:07
100% right if we look at the models that have really performed it's ones that have that that quote secret as you
00:22:13
mentioned when I first started 8090 a year ago one of the key bets I made
00:22:18
which was a mistake and we unw wound the bet but the first bet that I made was can we build a transpiler which is to
00:22:25
say can you take a CUDA workload and then can you redirect it away from Nvidia to different hardware and basically what
00:22:32
I learned in that process are all of the attention mechanisms that are built into transformers
00:22:38
that really differentiate how good the models are need to
00:22:44
literally be handtuned for every single target of silicon that you have so when
00:22:49
Amazon just kind of wakes up and says here's this chip it means nothing unless you can incentivize somebody to build to
00:22:56
it but the opposite is also true if you have a model and you just run it generically you're not going to get the
00:23:02
gains and it's not going to be as special as if you have a dedicated infrastructure and compute architecture
00:23:07
and say we're going to tightly couple these it's been clear that OpenAI has
00:23:12
had that Anthropic has had that Google has had that Deep Seek has had that and
00:23:18
I think Meta needs to do that otherwise they're always going to be floundering on their back heel one quick misnomer I
00:23:23
think you know when people hear labeling they kind of assume a photo of a dog and someone says "This is a dog right?" I
00:23:29
mean that's definitely how it started but if you look at sales business it's completely more from that so you could
00:23:36
actually label a problem so for example in in simple terms 2 plus 2 equals 4 is
00:23:41
actually um a reasoning data set right so you got to think of labeling not just
00:23:46
in the simple terms of you know this image but of massive data sets of of outcomes and that's what's kind of
00:23:53
really used to tr uh train these reasoning models um but I think there's another Yeah there's another story here
00:24:00
guys in my opinion and it's the performance of the Mac 7
00:24:06
right and I I'm going to have to check with my data science team but I'm wondering if we're this is the year
00:24:12
where we've seen the greatest divergence amongst the Mac 7 right so if you look at the Mac 7 and if I just gave you
00:24:20
right this performance you can see okay so Meta's up 18 Google's down Nvidia's up 8 Tesla down 20 Apple down 21 Amazon
00:24:30
down three and Microsoft is plus 13 right so it's kind of interesting in a
00:24:35
market that you know historically over the past few years where we feel the Mac 7 have been truly correlated now the
00:24:42
market is saying wait hold on we might start to see diverging performance what I read from that in in in one element is
00:24:49
the market's starting to try and sort out who are going to be the winners and losers who's well positioned versus
00:24:55
maybe falling behind right so I think we're going to start to see some divergent performance from the Max 7 i
00:25:02
think it's going to reward not Can you put that back up there for a second i mean I think that's so interesting because if you look at the conditions on
00:25:08
the field today you know Google's down 8% but again I would tell you as a user
00:25:16
Gemini models are exceptional like absolutely just baron exceptional i
00:25:25
think Anthropic is incredible for Codegen incredible
00:25:30
what I see is every single company on this list that isn't Nvidia
00:25:36
baking and rolling their own silicon yet Nvidia is up and the rest are down i
00:25:43
told you that I spent time last week at Tesla i would not be sleeping on this business i think that it is yet again
00:25:50
back into the land of being misunderstood the only one that I understand why it's down this much is Apple because it's not
00:25:58
clear that they're even baking something in private there's nothing public there's nothing private it just seems like they're transitioning into being a
00:26:04
cash cow and getting into sort of that cash harvesting mode but it's almost weird that the price action is what it
00:26:10
is because I would have thought that Google would be up meta would maybe be a little flattish to down nvidia is up but
00:26:17
maybe it could be down tesla's down but it should probably be up amazon's basically break even and Apple is down
00:26:22
and I think that kind of makes sense that's sort of how I read this table yeah i mean what I love Chimoth by the way on that is that like now there's
00:26:29
debates right and and you can argue whether you know you agree with Chimath or whether you don't spending 20 billion
00:26:36
cuz he's cuz he's not afraid correct yeah no let's pull the chart up again here by the way I think this is an
00:26:42
interesting way to only the only reason Microsoft is not on
00:26:48
this list is because of the limitation of the DOSs era interface of the Bloomberg terminal where it will only
00:26:54
allow you to compare six charts and not seven but we know that Microsoft is up 13 q perplexity yeah yeah
00:27:01
so you know when you also when you look at these there are some extenduating circumstances here like
00:27:07
Tesla's car sales are down all car sales are down and I think that's the piece that maybe isn't being accounted for
00:27:12
here and they're in a transitional period apple obviously Yeah there's a lot of regulatory
00:27:19
overhead so Tesla losing solar and EV tax credits yes apple Apple being told
00:27:24
to onshore and stop buying from China so their supply chains being disrupted because of tariffs those two companies
00:27:30
in particular are far more affected than the rest and even Amazon you know
00:27:35
there's been some conversation about tariff effect on Amazon but obviously that's offset with some of the benefits
00:27:42
they've been realizing and promoting as Jasse spoke in his letter this week uh from AI so I think that there's a
00:27:48
variation here that's probably a little bit more Thomas kind of tuned to
00:27:54
these conditions that aren't necessarily call it natural market forces but are kind of infl influenced market forces
00:28:00
associated with the the new administration and some of the policy choices that are being made if we were
00:28:05
looking at those number one and number two which one do you think gets to AGI first Thomas well wait hold on by the
00:28:12
way the other thing you should note Jason which I find really interesting is nobody talks about AGI anymore if you
00:28:17
listen to the language of all the companies it's all super intelligence which is a much more achievable goal because it's defined as being you know
00:28:24
multiples more intelligent than a human being but I think you're I think if you actually
00:28:29
did a search for the number of times AGI is being said today it's meaningfully less because I think people have
00:28:35
realized that that's not in the offing yeah by the way another lens chamat that I think about on these is who controls
00:28:41
their own destiny of these seven companies in AI right and I would argue
00:28:47
most I would argue Tesla does Nvidia and then it's kind of interesting right
00:28:52
neither Amazon doesn't have its own foundation model right they're kind of dependent on others right microsoft
00:28:59
49% does right because of this kind of relationship they have with open AI it's
00:29:05
both you know uh they they own a big share but they don't control it so there's kind of interesting and then
00:29:12
maybe 6 months ago we would have said well Meta absolutely does maybe Zuck's trying to question that a little bit and
00:29:18
you know it's it's fun in my opinion to kind of bring different lenses to this list right there's the regulatory one
00:29:24
that Friedberg was just talking about I kind of think about if I towards the co do I control my own destiny in this
00:29:30
market right and I expect these companies are not going to want to be dependent on others and are going to at least want to say no i'm going to
00:29:36
control my own destiny whether I win or lose who's your number one who's your number two if you had to could only bet
00:29:41
on two here to achieve super intelligence AGI let's just say win the AI re win the AI uh big prize the big
00:29:50
prize super intelligence AGI you know in the midterm five years five years from now we're sitting here thomas give me
00:29:56
your number one give me your number two look I I think to me number one I I still think Nvidia right i don't see the
00:30:02
GPU kind of getting displaced i see additional architectures kind of coming on board right and growing the market
00:30:08
but um at the end of the day all roads still lead to the GPU for all of these models so I would kind of still put um
00:30:15
kind of Nvidia on that my number two more of a dark horse but I I would pick
00:30:21
Tesla i do think it has the most potential for vertical integration right from all the
00:30:28
way the silicon to the model to actually the hardware right that might become super important not just in cars but in
00:30:34
Optimus so Nvidia 1 Tesla is my dark horse wow stunning chimoth who's your
00:30:41
number one and number two in the midterm 5 years from now we're sitting here on allin episode 700 Tesla's one and
00:30:47
Google's two and the reason is because they are the closest to having that
00:30:54
vertically integrated stack that I spoke about i think that Tesla has the best vision models now with XAI they'll have
00:31:02
one of the best LLMs and reasoning models and they'll be able to eventually
00:31:07
stick that on Dojo and then all of that will be in all of the physical AI that
00:31:14
you will interact with in your daily life whether it's a robot or whether it's a car or whether it's a robo taxi
00:31:20
so that's number one and then number two for many of the same reasons I think Google because you'll have the Gemini
00:31:27
family of models which just absolutely kickass like VO3 which we haven't really spoke about is going to destroy
00:31:34
Hollywood like in the next year like Hollywood is done I think but they're
00:31:40
landing model after model they have the TPU and the next generation TPU I think
00:31:45
is exceptional they're baking quantum and then they have an entire funnel of billions of people that they can direct
00:31:52
experiences to so Tesla one Google 2 chimath quick followup on that i'm curious on Google this is the because I
00:32:00
I oscillate a lot on this particular name can Google win if search declines
00:32:05
yes and I think that what probably has to happen is bear with me when I say this but if you had to boil down
00:32:13
Google's economic northstar metric right not the value northstar the economic
00:32:18
northstar metric would be price per click and I do think that Google is extremely
00:32:25
well positioned to pivot that to price per token and I think that they have
00:32:30
some emergent classes of physical AI but they have the largest pool of people where they can generate a price per
00:32:37
token value framework through YouTube through Gmail through workspace I think
00:32:42
through search but probably it's a different kind of model it just requires them to rip the band-aid off at some
00:32:49
point but yeah I think Google can do it i'm going to go with you Chimath i'm one
00:32:54
uh my one and two are either Google uh or Elon and I I'll just say Elon because
00:33:00
I uh like you I spent a day up at um XAI and I saw what a magnet for talent he is
00:33:06
i got to sit in some meetings and just he was interviewing people and he was working with that talent 8:00 at night
00:33:12
there's a lot of people there on a Saturday grinding it out it was nuts i first went to XAI
00:33:18
in the 15 minutes that I was in the parking lot finishing a call the kinds of people that were walking in and out of there you could tell they were big
00:33:24
brains i don't know how you know what I mean like from every walk of life they all just looked much smarter than the
00:33:31
rest of us yeah it some of them were like chain smoking cigarettes and just like stressed out it was crazy i hit a
00:33:37
couple of zins i'll be totally honest um but the reason I say Elon versus Google is I think Elon's in a unique position
00:33:43
and I don't have any insider information here and and I haven't talked about this or I'm not back channeling from Elon lest anybody aggregate this i think what
00:33:52
Colossus has done and what Tesla has done both of these things Tesla with their own stack of hardware to your
00:33:58
point Chamoth hardware plus software plus the user application of FSD and Optimus then you put that together with
00:34:04
the data the real-time data of X formerly known as Twitter plus um you
00:34:09
know what he's building with XAI and obviously those two companies merged i think Tesla board XAI board have to get
00:34:15
together put those two companies together one's worth a trillion one's worth 100 billion put them and just have
00:34:20
all that brain power going in one direction as opposed to Elon test switching between the two you do that I
00:34:26
think he wins number one you don't do that I think he either gets one or two and then I think Google um is going to
00:34:34
have a better search product thomas I think it's a really important point do they lose search share doesn't matter
00:34:39
what I think matters is are their ads more effective is their ad network more effective and I think based on what they know on you from your chat searches and
00:34:47
your discussions and what they analyze in your email just analyzing your Gmail and your surfing behavior and Chrome if
00:34:53
they get to keep it your Android phone if you use it your YouTube list and what you how when you drop off allin and when
00:35:00
you start listening to another podcast whatever it is all that data all that data is going to lead to an ad network
00:35:05
that performs so much better that even if they lose search hair their ad network is going to continue to grow and
00:35:10
I think it will increase in velocity so those are the my top two freeberg I'm curious from your position
00:35:16
which one you think is number one and number two i saved you for last because you know what we do here we save the best for last freeboard go ahead i think
00:35:24
there's a difference in how I would kind of lump them i I think that Tesla
00:35:29
probably has the it is the best place to invest if you want to have a shot at a
00:35:37
massive new industry so they've got a baseline business in in obviously the automobiles but I think this humanoid
00:35:45
robot opportunity is absolutely mind-blowingly ginormous and I don't
00:35:51
think that there's a better company on Earth positioned to execute against this humanoid robotics opportunity than Tesla
00:35:59
so you know it's sort of like I would call it a low probability high upside sort of call option embedded within that
00:36:06
business and obviously you're paying a premium for that because it is still a very healthy premium you pay for that business i think Nvidia to Thomas's
00:36:14
point I think the common thesis is it is the most protected the durability of the
00:36:19
business is there but I would argue that there's actually a low probability but very high severity risk to Nvidia in
00:36:25
China there was just a demonstration last month of a 1 nanometer semiconductor manufacturing process out
00:36:32
of China i think the more that we continue to try and isolate China from a
00:36:37
policy perspective the more we are emboldening investment in China meaning
00:36:43
from the government from private industry into China to create alternatives to the chip stack where the
00:36:49
United States companies particularly Nvidia have emote today so I do think that there's going to be an emergent
00:36:55
competitive threat coming out of China to Nvidia and just like we were knocked over by DeepSeek I think we will be
00:37:01
knocked over by some semiconductor manufacturing processes um coming out of
00:37:06
China in the near term but the overall kind of by the way Dave just on that point I think Sax's work on the
00:37:13
diffusion rule just generally I don't think has kind of gotten enough attention in the rescending of the
00:37:18
diffusion rule which essentially handicapped our ability to even arm our
00:37:23
allies right with our semi with our semiconductor technology um in my
00:37:29
opinion was kind of a milestone and very important moment um to to try and offset
00:37:35
exactly what you were just describing that's exactly right i mean there there there was a report a few months ago and I mentioned it on the show or maybe I
00:37:42
didn't or maybe I sent it to Sax and we talked about it offline i I can't remember but it was about a $40 billion
00:37:47
investment being made in developing competitive semiconductor manufacturing full stack solutions out of China so I I
00:37:54
do think that the lithography IP moat is being crossed in China i do think that
00:38:00
China is developing actually new technology for uh DUV and EUV systems i
00:38:05
I do think that there's a risk uh to Nvidia's core now look Nvidia is such a durable business there's great modes
00:38:12
great advantages but we're creating every incentive for an alternative to Nvidia to emerge from China and then my
00:38:18
my third kind of categorization would be what's the portfolio uh solution i think that's Google i think that there's a
00:38:24
diversification of high beta bets inside of Google of any one of which could have
00:38:31
call it a trillion dollar market cap outcome ranging from Whimo to quantum
00:38:36
computing to the biologics work that Demis is working on out of um isomorphic
00:38:43
uh there's a number of things that do not get a lot of attention at Google so yes there's a there's a core business
00:38:49
that that may be at risk Thomas but I think that there's a a portfolio of options you get at Google and you just
00:38:56
need any one of them to hit to kind of make up for the loss but I do think also Sundar in my interview with him which we
00:39:02
put out a couple of weeks ago is very thoughtful about where search evolves to and he is being I think reasonably
00:39:08
aggressive in in trying to evolve the search product architecture to meet the market to meet the consumer i do give
00:39:15
him credit for that so Google would be in a good place for me as an overall kind of pick in that set of options so
00:39:20
just to be clear Nvidia 1 Google 2 or Nvidia Tesla like I said I think in terms of like having the right sharp
00:39:26
ratio is how I would think about it the alpha and the beta adjusted returns I would put Google number one i would
00:39:32
probably put Tesla uh Tesla's valuation I think already has a premium associated
00:39:37
with those options so I don't know yeah so I don't know if I would really pay that premium i think um aside from the
00:39:43
valuations let's take valuations out of it just the the game here is who wins
00:39:48
the AI prize 5 years that's how I understood it as well yeah so valuation irrelevant valuation irrelevant who wins
00:39:55
the AI prize one you're saying Google two you're saying Tesla i think Google's in such a position i I mean look Demis
00:40:01
uh Demis I think has been fairly koi about where they are they obviously promote Gemini 2.5 but there's a lot
00:40:09
still coming and it's and and as Chimath pointed out it's not just LLMs there's a
00:40:15
pretty sizable family of models including a a lot of these um graph-based models that are being used
00:40:21
in really novel applications that no one else is even close to no one spending time on i mean some of the weather
00:40:28
forecasting it might seem small and trivial but it's a demonstration of Google's competency in in core model
00:40:34
development that shows an understanding and a depth of research and work that goes well beyond LLM so I'm pretty
00:40:41
bullish on the depth of talent the full stack yeah yeah and whatever they learn there could apply to Gmail could apply
00:40:46
to search could apply to ads could apply to YouTube algorithm right it's just goes up and down yeah yeah from a product perspective I do think you see
00:40:53
this kind of multi-model emergence that that we're now seeing that no one talks
00:40:58
about the single model that sits behind the application there are multiple models that work together and obviously
00:41:06
this agentic architecture unlocks another layer of not just kind of
00:41:11
solutions to complexity sure and so there's there's quite a lot I think that's emergent here um that Google will
00:41:17
start to kind of benefit from uh in the year ahead i mean for those of us you know who love tech right if we if we
00:41:23
step back for a minute I really feel like to use the analogy of this podcast like we are now at the WSL World Series
00:41:31
of Poker right we got seven companies around the table the stacks are trillion
00:41:36
in size right and all of us are going to get a front row seat to see what happens over the next 5 years i mean and on top
00:41:44
of that we're going to get to analyze bet ourselves on who we think's going to win we know there's some other companies
00:41:49
that are pushing to get at that table right with some sharp elbows i mean what a time to be doing what we're doing i
00:41:55
don't know if I love the analogy because I don't think first of all it's a zero sum game where there's this x number of chips and someone ends up with all the
00:42:01
chips i do think you could see as an example just talking about the scenarios we we just described Tesla developing an
00:42:08
extraordinary humanoid robot business that's worth a trillion dollars google building you know to Chim's point a
00:42:15
media empire based on generative AI in media and then you know Nvidia building an entirely new chip stack that
00:42:21
everyone's participating in so all of them in an ecosystem based way could could be major winners here yeah you're
00:42:26
right i I didn't mean it in the zero sum nature of it i meant it more in the in the stakes right and and and there's a
00:42:33
lot of hands to be I like the analogy because there's a lot of hands to be played and there is a price pool right
00:42:39
and and you could have three or four people at that table one thing I just want to point out here is just speaking
00:42:44
of regime change what is going on at Apple like they Siri was just the early idea of an AI agent it's just totally
00:42:51
disgrat it's disgusting it doesn't work it's embarrassing and then their biggest developer conference they're redoing the
00:42:58
UI like time for regime change at at at Apple no this has happened many many
00:43:05
many times in many industries before which is that companies that were
00:43:11
stalwart organizations transition themselves from being a growth business to being a cash cow and
00:43:18
these are well doumented transitions and it requires an extremely brutal reset if
00:43:26
you want to shake that up yes I think that the same thing that I think you have to respect Apple for which is
00:43:32
stability the duration of some of their best longest
00:43:38
serving executives are there for 20 and 30 years on the scale of innovation it's
00:43:43
a horrible thing and the reason is that we all just get old our skill sets
00:43:49
become rusty and we don't have the energy or the capacity to think about
00:43:54
what the future actually looks like because we are not living it and then what happens is you task those decisions
00:44:01
to people that you try to hire but you know you saw it in the clip with Sam
00:44:07
even in all of that crazy recruiting chaos that's happening right now for these brilliant machine learning and AI
00:44:13
people maybe that's a fight between OpenAI Meta and maybe Google but what you don't
00:44:20
hear is Apple so who's Apple getting i have to think that Apple is not getting any of those people so by the time you
00:44:26
end up at Apple it's just a different caliber of person that is true and
00:44:32
they're living inside of a cash cow organization that's going to optimize for don't make mistakes right but that's
00:44:40
h it's happened to HP it's happened to Lotus it's happened to Intel it's happened to General Electric it's
00:44:46
happened to companies it's just and it's happening to Apple so we should just not
00:44:52
sweat it and move on i don't know thomas what are your thoughts i mean it's kind of shocking with all that cash and they
00:44:59
don't acquire anything they had project Titan $10 billion to build their own car and they just shut it down imagine if
00:45:05
they kept going with that you think regime change time maybe Tim Cook retires and put somebody who's a product person in charge of it or maybe they
00:45:12
should merge with Tesla and put Elon in charge of it all there just seems to be no new products coming out of there like
00:45:18
it's absolutely uh confounding that they're optimizing for share buybacks and earnings per
00:45:24
share instead of having some amount of that money go towards innovation and acquiring companies biggest acquisition
00:45:31
is Beats give me a break i mean it's interesting right for me and I've studied Apple basically my whole career
00:45:38
and it's kind of interesting right because if you think about the their defining
00:45:44
competitive advantage right was the integration of hardware and software that led to the beautiful MacBook that we're all using it led to the iPhone and
00:45:51
right the fact that they were so coupled between hardware and software the user interface you know etc and I think it
00:45:57
directly led to them winning let's call the the mobile era right but I back to
00:46:02
Chamas's point and I think the analogy holds in AI they're the opposite right
00:46:07
they don't control I don't you know the silicon they don't control the underlying models um and so now they're
00:46:14
back to maybe you know using a historical analogy the PC makers who didn't control the OS that's right so I
00:46:21
I think the good news for them is look they still have a monopoly on users they have three trillion of market cap to
00:46:27
kind of play with so I think it's way too early to count them out but I think
00:46:33
you know the market let let's posate what's the most extreme thing that they could do right just for just for
00:46:39
intellectual sake right uh buy OpenAI for 500 billion i'm just going to put a crazy thing out there right so you think
00:46:47
okay that's the most extreme well is it even that extreme and what would Apple's stock do that day go up that's my view
00:46:54
too right i actually think it would go up not down even if they did something like that so I do think they need to be
00:47:01
kind of aggressive i do think to your point I think Freeberg it is important that you know all seven of these
00:47:06
companies could actually win and do well right that that is a absolute possibility but I I would love to see
00:47:13
them be a little bit more aggressive i mean you guys remember when Steve Jobs bought Finger Works right it was this
00:47:19
tiny acquisition they made this little trackpad that you could use your fingers on no one figured out why they did this and then in turn into multi-touch and
00:47:26
scrolling right so I think it's it's going to be fascinating to see what they do thomas that was a great question I
00:47:33
was about to ask if Apple could do one thing they could do one internal project or buy one external company maybe we
00:47:39
could do both around the horn what would we advise them to do my number one is build a humanoid robot like how does
00:47:45
Apple not have a humanoid robot that seems like that's obviously the next giant consumer market is having Optimus
00:47:51
or figure in your house freeberg I'm going to go to you first since I went to you last last time is there a product
00:47:58
that they could do that they could build that they would be uniquely suited to that would turn this all around if you
00:48:04
could pick it on their road map what would it be i do think there is i do think they're doing it and I do think they have a shot at winning which is
00:48:10
this kind of ambient AI assistant i don't know about you guys I must own 30 friaking Apple devices uh I have many
00:48:17
Apple computers I use in different offices i have phones i have many AirPods i got everything watches
00:48:23
everything i'm ubiquitous on the Apple platform so I'm an easy transition into this if it works so as everyone races to
00:48:31
build kind of the Agentic AI assistant that uh is sort of in my ear all the
00:48:38
time or available where I don't have to stare at my freaking phone like this um it is a great unlock for humanity it's a
00:48:44
great unlock as a consumer it's feasible technically and I'm sure Apple of
00:48:49
everyone that we've referenced today is best suited to both access the consumer
00:48:55
design and engineer this solution in a way that can be truly transformative i think it references a little bit what
00:49:02
Johnny IV and Sam Alman have been talking uh about doing but I do think that this is exactly the direction Apple
00:49:08
is headed and I do think that they've got a very great shot at at winning at it don't think they need to own the full
00:49:14
stack to be successful here got it okay so we got Optimus we got the device you're talking about this ambient
00:49:20
assistant is part Siri and part maybe a pendant that records your behavior in
00:49:25
the world and gives you feedback to it and that's what they're calling a puck perhaps that Johnny IV has made or these
00:49:31
pendants that record everything thomas what's your thought on the one product they could create to that point um it
00:49:38
it's interesting to think that the AirPod business at Apple is 3x Open's revenue base today that's right and
00:49:45
that's just the AirPod business and by the way let me let me just say one thing about this we all think about devices in
00:49:50
the context of a single device being an assistant i think if there are more
00:49:56
devices integrated into our lives and the assistant is ethereal and ubiquitous
00:50:02
amongst the devices it's almost like uh the Star Trek Next Generation you walk in you say "Hey computer." And there's
00:50:08
always a device available that's doing things there's always a device observing there's always a device able to take care of things for you whether it's in
00:50:14
your ear whether it's your phone whether it's your watch but basically these devices all instead of acting
00:50:19
independently they all know what you've been asking or talking about with the other devices and so you could get in
00:50:24
your car and you could pick up you know the conversation you were having you know while you were sitting in your
00:50:30
office in front of your computer to do work and so the agent effectively is almost like this ethereal ambient
00:50:36
assistant so everywhere you go the agent is there they could even be in a candle lit bath with you Freedberg they could
00:50:42
be in there they could Well I mean by the way think about also you know it it it know having identity so it knows who
00:50:47
you are but I could be in your in your home Jal not that I would ever get invited to your home but let's say I was
00:50:53
there uh you know I could walk into the the living room and there's your puck and it starts talking to me because it
00:50:58
knows who I am and yeah it's like it knows me yeah or you and I have a bath for two you and I could be a candidate
00:51:04
for two and it would know the when each of us are fighting over what music we want to play the assistant will you know
00:51:10
hear out the debate playlist do you have a uh a device before we go on to IPOs here do you have a device or an angle
00:51:17
for Apple to go after if they were truly ambitious or maybe they are and it's just in stealth what do you think you
00:51:22
think it's the goggles the glasses you think it's a pendant you think it's optimist what do you think i don't think they have any chance of anything great
00:51:30
love it i would take the exact opposite of what Freebrook says look at this chart and I'll tell you why okay here we
00:51:35
go this chart is not This chart is not a strategy so this is a chart of Apple's revenue and what you see is iPhone has
00:51:42
completely stalled out and so to Thomas's point where do you make money you make money in other hardware this is
00:51:49
not a strategy of success this is a strategy of inefficiency i lost my AirPods i need to buy a new
00:51:55
pair oh the cables changed i need to buy a bunch of those this and that and a
00:52:00
this and that strategy is not a strategy it's a tactical play for revenue optimization in the short term a company
00:52:06
that focuses on this kind of revenue growth is not capable of creating
00:52:11
something that's exceptionally unexpected that will come from a new company who
00:52:16
has no ties to the past has nostalgia on the fact that we're going to swap out
00:52:22
the connector type and you know book another billion dollars the what Thomas said is an indictment actually about
00:52:28
their ability to do it when your AirPods business is two or three times bigger than Open AI what there is internally
00:52:36
when you try to have a strategy meeting about what to do is derision about Open AI because you're like that's small and
00:52:43
even our AirPods business is three times big that's what some smartass MBA will say in that meeting and it'll shut the
00:52:50
meeting down so how do you expect that culture to then all of a sudden get their act
00:52:55
together i think it's exceptionally hard and here's the clip on Q play the clip
00:53:01
Nick it's a great point here's the clip i'm Apple nostalgic me too bring Steve Jobs back watch this lunacy you probably
00:53:08
saw that Johnny IV is linked up with Open AI to create some sort of future AI device
00:53:14
yeah I don't know what that is i don't either yeah is this a space that Apple's looking at is this a space that goes
00:53:19
beyond what you have in the current lineup of devices something that is more personal maybe you wear it glasses
00:53:27
i I think I mean I think we have some extremely personal wearable devices if you want something that's uh aware of
00:53:34
your environment with with audio I think you're you're wearing one right now on on your wrist um if you want something
00:53:41
that you can capture the environment with and see and also receive visual content you might just have one in your
00:53:47
pocket right now um are there other form factors that can make sense to AI uh
00:53:53
sure but uh pretty hard to beat something that's uh with you all the
00:53:58
time and glancable or you know provides a nice screen that you can interact with um so uh yeah I I don't know what
00:54:05
they're working on what do you think Jimoth again I think I want to be very clear about what I'm saying that is a
00:54:11
very competent Craig Federi very very competent executive
00:54:17
and whoever the person beside him is that guy's I'm going to assume competent as well they're competent at making
00:54:24
money the way that they've made money for the last 17 years with no meaningful
00:54:32
disturbance and I think it's just something to appreciate that after 17 years of
00:54:38
unmitigated linear success it's very difficult to retool yourself it's like
00:54:45
asking Michael Jordan to go and all of a sudden become an all-star baseball it doesn't work
00:54:50
and so I think I I think it's okay though this is my point it's okay guys to have creative destruction of
00:54:56
companies like there was probably a version of us blathering on about HP and
00:55:01
being nostalgic about the transistor radio that they made and the you know HP12B calculator that they made and oh
00:55:08
my god why can't they figure their [ __ ] out and where are we today hp doesn't even exist it's okay i mean just Thomas
00:55:17
the fact that they launched Siri they bought that company and Siri can't do
00:55:22
anything other than like an alarm can barely play a song it barely can do
00:55:27
directions i I I mean literally we're in year like 27 of Siri and it can't do
00:55:33
anything and then I have the the Google and Gro voice and when I turn that on it
00:55:38
does whatever I want it will load on my Pixel it loads other applications fires it off does specific tasks in it it's
00:55:44
absolutely descriat on your Pixel i have a Pixel when I when I flip open my Pixel i have
00:55:52
to I have the Pixel 9 Chimath it's the Anaconda of smartphones pixel 9 foldable
00:55:58
got it it's the greatest assistant ever it's what Siri it's what Steve Jobs showed Siri i had you at nine he had me
00:56:04
at Anaconda yeah I had you at 9 in and we can all aspire maybe get Roman extra
00:56:09
get that extra inch thomas go ahead chamath I would argue to you that I think this management team has done it once and it's in the transition of their
00:56:17
gross profit base which doesn't show in the chart that you just highlighted but was something that I kind of lived as an
00:56:23
analyst covering the stock for a long time where if you remember over a decade ago 90 plus% of their gross profit was a
00:56:30
onetime hardware sale on the iPhone and no one thought that they would ever be
00:56:36
able to get away from the drug of selling that one iPhone unit right and cut to you know over a decade later it's
00:56:43
40% right and I don't think they get enough credit for actually transitioning
00:56:48
from hardware to a recurring gross profit base but look you might argue that that was an easier pivot and
00:56:54
challenge than what they're going to face and so let's see whether they can do it
00:56:59
the other thing guys I wonder about um let's I know we want to talk about IPOs but I do wonder whether Zuck buying
00:57:07
scale for 15 billion gives air cover for other companies to really start being
00:57:13
aggressive right and and to me as we think about circle and coreweave two
00:57:18
companies that have gone IPO recently it's it's kind of amazing kind of
00:57:23
numerically that the charts are almost identical even you know on a dollar basis on a share price right because to
00:57:31
me what it says we were talking about the dispersion of the Mac 7 before right which are going to do well which are not
00:57:36
i expect we're going to have a lot of opinions on this over the next few years and frankly they may change we you know we may think Apple one way today it may
00:57:43
change in a month right but I do think the market is starting to realize that there is dispersion that AI might create
00:57:50
some all winners or some winners and then some losers right and is starting
00:57:56
to think about okay how do I want to be positioned for the next five years what are big open-ended growth opportunities
00:58:02
and here comes two companies one lever to crypto right and the other lever to AI so I don't think it's a surprise to
00:58:09
me these things are intertwined you're 100% on because here's the thing the average profit margin of the S&P 493 is
00:58:18
drum roll please 12% the average growth of the S&P 493 is drum roll please
00:58:25
single digits so to your point why would you belong any of these 493 companies
00:58:31
that may turn around and one day just get decapitated by something you don't even know that's getting cooked up by a
00:58:37
couple kids in a garage using OpenAI or Grock or what have you it just makes a
00:58:42
lot more sense when you find investable companies in the big themes of the future to at a minimum hedge right be
00:58:50
less long the past and frankly make some bets about the future and I think that
00:58:57
that's where you're seeing these IPOs just absolutely rip what is a better comparison in my opinion are the
00:59:03
companies that are truly levered to the future themes of AI and crypto versus any of these IPOs that have happened of
00:59:11
companies that are not and I think what you see is there's a dispersion there as well and they are being treated almost
00:59:17
as similarly Jason as the S&P 493 it's like yeah it's good yeah
00:59:24
it's fine they get some reasonable gains but if you're lever to any of those two two trends you're off to the races
00:59:31
because it's just so disruptive people don't want to be bag holding these old legacy companies we're already into our
00:59:38
next topic which is IPOs and M&A lena Khan is no longer in the building and M&A is back on the menu as are IPOs as
00:59:47
Tom has pointed out three IPOs March 28th June 5th and June 12th coreweave Circle and Chime obviously Coreweave up
00:59:53
4x after going public $81 billion market cap absolutely stunning circle 25x
00:59:58
oversubscribed 6x from its opening price $ 48 billion market cap chime that's a
01:00:04
NEO bank like New Bank which is already public that was up 40% uh in its IPO price but then it went down 20% $12
01:00:10
billion market cap on the other side of the ledge we have a ton of M&A this year
01:00:16
so when you look at what's happening under the Trump administration look at what's actually happening the game on the field is three major IPOs uh and
01:00:24
then massive amounts of billion dollar acquisitions obviously we talked about Google acquiring Whizford 32 billion uh
01:00:30
SoftBank bought Emperor I don't know what they do 6.5 billion openai bought two companies one for three billion one
01:00:37
point for 6.5 billion developer co-pilot Windsurf 3 billion johnny Ives IO making
01:00:42
some sort of a puck or hardware device data Bricks brought Neon for a billion salesforce uh did an $8 billion
01:00:48
acquisition and then interesting Door Dash bought two companies uber made two smaller acquisitions there is a ton of
01:00:55
activity here what does it say about the market David Friedberg that we're seeing
01:01:01
so much M&A and these amazing IPOs coming out all within the last 3 4
01:01:07
months okay so let me just follow up to a comment Chimath made and ask Thomas
01:01:13
his view i have a a theory and I haven't looked empirically to see if it makes
01:01:20
sense for most of the S&P 500 the fundamental profit growth is pretty
01:01:25
anemic with the exception obviously of a couple of the big tech outliers the MAG7
01:01:30
and a few others but for for the majority of the S&P this is a pretty kind of anemic environment relative to
01:01:37
the transitions that are underway in the world fundamentally with with AI and ancillary technology so are the
01:01:44
institutional fund managers hungry for access to some of these new you know
01:01:50
high growth offerings and they have been held off because and just to kind of go back I think it was around 2008 or so
01:01:57
public institutional fund managers started to do crossover investing into private equities and that scaled up and
01:02:04
scaled up and it it entered obviously a stage where it was a heavy flurry a lot of activity and a lot of crossover late
01:02:10
stage investing um you right until 2021 when things started to pop 2022 and
01:02:17
because they were overexposed with their private equity portfolios relative to their public equities they
01:02:24
came out of 2122 with the market declining and they now had a higher concentration of private equities than
01:02:30
they were supposed to have and so they have been kept out of the market for the last 3 or so years of the private market
01:02:36
and now is there kind of this pentup hunger or pent-up demand for new issuances for high growth tech issuance
01:02:42
Is is that what we're seeing is there kind of this pent up demand because they've had to stay out of the the private market for 3 years and if there
01:02:49
is obviously it bodess well for latestage growth startups that are looking to go public because the demand
01:02:54
will be there and I think the reports were that the Chime IPO was like 18x overs subscribed i think you're right
01:03:00
and and something that you know I've talked about with you guys and uh was a was a big conversation at our at the
01:03:07
all-in summit last year was the health of the uh private ecosystem right and we
01:03:13
talked about the concept of look if you put a dollar in you need to get a dollar out right and so I do think that we're
01:03:20
starting to see a healthier market where we know a lot of dollars have gone in but now we're starting to see some
01:03:25
dollars coming out so I think that's both in M&A by the way and it's also in IPOs So I think that's one element but I
01:03:31
also think the second element which is we're the tailwind of the mobile and SAS era right and even if you look at the
01:03:38
SAS companies we kind of put this together in our deck when we were preparing it for our conference um this
01:03:44
week chamath I think you'll find this interesting right if you look at SAS in 2021
01:03:50
the median growth rate for SAS companies was 17% and a quarter of those were growing over 25% mhm okay if you look at
01:03:58
SAS today the growth rate has been cut in half 17% to 9% and only 5% of that
01:04:06
cohort is now growing above 25% so I think Dave what's clearly happening
01:04:11
right is other sectors which were predominantly seen to be growth are now slowing down right so that's kind of one
01:04:18
piece so the market can no longer just rely on saying "Oh I'm just going to own the Bessemer SAS index right for the
01:04:25
next decade and I'll do great." Because those companies have really slowed down and I think it's starting to look
01:04:30
forward and think okay now over the next 5 to 10 years what are the companies
01:04:36
that can compound at maybe 25% per year over that time frame and I think companies like Cororeweave and Circle
01:04:43
and Chime by the way and others are going to kind of fill that gap i um I really like this chart
01:04:49
if I had to guess about what has changed from 2021 to 2025
01:04:56
is that most companies have realized that buying yet another
01:05:02
vertical software solution is not going to help their business that it typically
01:05:08
adds bloat it adds cost and it adds people and I think starting in 2023
01:05:16
what people started to guess is at some point in the near future you're going to
01:05:21
have some AI way of rewriting all of this vertical software and I think
01:05:27
that's why it stopped growing i don't think this SAS market ever had the
01:05:33
return on equity that it was supposed to and I think so many companies have woken
01:05:39
up from this hangover saying there's got to be a better way it can't always be yet another tool yet another program yet
01:05:47
another multi-year delay yet another price escalator and I think
01:05:52
that that the jig is totally up for software you're referring to the
01:05:58
Salesforce and the SAS category Chimoth and what you're doing at 8090 specifically yeah well it's it's not
01:06:04
just us but like if you look at anybody that's rebuilding software it is so much easier to rebuild software
01:06:12
from scratch today like my team of 30 people can transact hundreds of millions
01:06:18
of dollars of work not because we are so prolifically amazing but frankly because
01:06:23
well I think the team is good but honestly because the underlying tool chain gives you a level of leverage and
01:06:29
so if you rebuild the software development life cycle using these tools
01:06:36
you can't help it but become much more efficient and you can't help it but deliver custom solutions that are
01:06:42
meaningfully cheaper and I think Jason if you look at the entirety of the software that runs the world we're going
01:06:48
to rebuild it soup to nuts all of that and the tool you're referring to just for the audience is the AI co-pilots
01:06:55
that are making that are contributing 30 40% to code bases at Microsoft and less
01:07:01
less specifically that because those are those are good for individual people but the software development life cycle is
01:07:06
more the horizontal end to end of making things got it so what we do internally at 8090 is we have an entire process
01:07:13
that starts from the PRD all the way out to the functioning code and we use different techniques at each step but
01:07:19
what you get is a 50 60 70% increase at each step which then compounds and so
01:07:26
you have the ability of a team that would otherwise be able to service tens of millions of dollars be a team that
01:07:33
can service hundreds of millions and then a team that would otherwise service hundreds can service billions let me ask you guys your response to this theory if
01:07:42
there is going to be this kind of accelerated call it custom software rebuild of
01:07:48
business models and you take the S&P 493 do you think that we enter an era where
01:07:53
there is a similar dispersion as we're talking about seeing in the MAG 7 with the S&P 493 where there are going to be
01:08:02
probably the biggest money-making opportunities for investors that we've seen in decades
01:08:08
between those that do adopt and do rebuild using AI and those that don't or
01:08:15
100% i had a call yesterday with one of the largest private equity funds in the
01:08:21
world hundreds of billions of dollars under management and we're doing something with them at 8090 with one of
01:08:26
their most important assets and when you're an owner of a business
01:08:33
and you can direct specific change and you can rip out
01:08:40
hundreds of millions of dollars of software licenses and replace it with
01:08:46
tens of millions of dollars of highly customized software
01:08:52
it's an enormous lift to opex and business model quality so why doesn't it happen more the reason it doesn't happen
01:09:00
right now for this S&P 493 is that the IT organizations inside all companies
01:09:07
essentially speak a different language than the CEO the CFO and the board so if the CEO CFO and the board of directors
01:09:14
of the S&P 493 speak English the IT organization speaks Mandarin Chinese and
01:09:19
you get away with saying all kinds of [ __ ] i'll give you an example i went to a CIO conference one person that I
01:09:27
met an $18 billion a year IT budget
01:09:33
what the [ __ ] does that actually even mean to spend $18 billion a year on it
01:09:39
i'm not saying that this is a mag seven company guys and when you take that example and you multiply it by 50 and
01:09:46
100 and 493 examples of people spending money there's an entire cartel of
01:09:52
influence that's been built in software that's going to get undone because you're not going to be able to justify
01:09:58
it free absolutely correct and the response from the SAS industry is changing from the per seat model as the
01:10:05
number of employees at these companies continues to get lowered obviously Microsoft a lot of layoffs andy Jasse
01:10:11
talking about layoffs they're moving from the per seat model they're not taking this uh laying down uh they know
01:10:18
that people are going to make custom software so what they're doing is they're moving to a consumption model so you're seeing people charge per call per
01:10:24
customer support call etc and well it's I'll tell you why it doesn't work working in combination hold
01:10:30
on hold on let me finish the other thing they're doing is they're dramatically lowering the number of people and the developers they have on their team and
01:10:37
then a lot of what's happening in the background is the third piece they're doing is they're starting to uh do rollups and people are starting to talk
01:10:44
about how can we take you know 20 of these SAS companies lower them just like you're doing to compete your thought
01:10:52
playbook well I just wanted to comment on this like consumption based pricing it doesn't work and what I mean is you
01:10:58
can have some adoption in the short term the best example is Snowflake but in the long term it destroys your business and
01:11:05
the reason is because you don't know which data is valuable and you're not going to put up with a variable business
01:11:10
model that increases more and more cost because you need to trap everything and so what happens is all of these other
01:11:17
companies develop around you people go back to Postpress people go to Superbase they find all of these ways of saying
01:11:24
snowflake makes no sense and the reason is because in this world nobody's going
01:11:29
to pay consumption because you're like how do you expect me to you know hold and store and pay for terabytes and
01:11:35
terabytes potentially a day of data it's not sustainable we'll see if intercom
01:11:41
Salesforce HubSpot and we see if all of those people start Slack start losing their customer base or if they lower
01:11:47
their pricing to make it just too easy to keep those systems in thomas your thoughts yeah so two quick thoughts uh
01:11:52
number one Chimath to put a kind of a mathematical frame on this right we know that Anthropic is kind of the level zero
01:12:00
of code generation they're they're doing incredibly well powering companies like Cursor right i think and this is order
01:12:06
of magnitude correct that Enthropic in Q1 added 70% of the net new ARR in the
01:12:12
SAS industry right defined by public SAS companies right so let's just think that
01:12:18
the company in AI that is most powering the disruption of SAS added 3/4 of the
01:12:23
net new of the entire industry right so that's kind of point number one I think Freedberg point number two I think what
01:12:30
we're seeing in the Max 7 right where we're starting to have debates about who's well positioned and who isn't who's going to win and who isn't right
01:12:36
is actually as it was in the past 5 years going to be a broader lens into the S&P 493 i think inside of boardrooms
01:12:45
inside of every investment committee you're going to see the exact same conversations that we've been having
01:12:50
about the MAX 7 right who who's well positioned who can win what are the management teams maybe like Zuck that
01:12:57
are being aggressive and bold and capturing the opportunity and which are the ones that are not so for me as a
01:13:02
stock picker right I think over the next 5 years I couldn't think of a more
01:13:07
interesting time where we're actually going to see dispersion between winners and losers and do you think that these
01:13:13
rollup models make sense so you've probably heard uh and I don't know if you guys have considered this but
01:13:19
obviously some fund managers are putting together pools of capital to go out and buy businesses that they can then apply
01:13:26
their knowhow they're bring in smart people in AI to then create a category killer and go after that market and are
01:13:34
you guys participating in that and how do you kind of view that opportunity are all the public companies basically too
01:13:41
mature or are some of them going to kind of go after this type this model as well it goes back to whether you can attract
01:13:46
the talent to go and do these things my advice to this large private equity firm is you can probably try to stand up your
01:13:54
own AI org but I suspect you're going to get the person that didn't get an OpenAI
01:13:59
offer didn't get a Meta offer didn't get a Google offer didn't get an 8090 offer then didn't get an Apple offer and then
01:14:05
that's the person you'll hire how good that person will be who the hell knows i think the problem is that even if you
01:14:11
take some of these kind of meh industries and roll them all up you
01:14:18
ultimately have to find a buyer who wants to own that business after you so the question is like if you were to buy
01:14:24
a bunch of accounting firms or law firms or IT services firms and
01:14:32
you do an incredible job who wants to buy that in seven years
01:14:37
meaning if you talk to like if you went to the OpenAI demo day there was this really interesting chart where Andre
01:14:44
Karpathy talked about integrating Google login into one of his apps i think it was the his
01:14:50
menu gen app and the comment he made which profoundly hit me is like why am I
01:14:56
doing any of this why isn't this just one click behind the scenes and you could take that generalization and apply
01:15:03
it to all of IT services why does any of that exist why isn't it all one click
01:15:08
and eventually if these agents become smart enough the fear that I have is that there is no terminal buyer for many
01:15:14
of these companies mhm but they could still be public chimat i mean they could they could
01:15:21
trade at some multiple of cash flow and you're basically arbiting the cash flow but I'm not talking about the private equity trade i'm actually talking about
01:15:27
the public equity trade if you look at the 493 companies those are better positioned i think like instead of an IT
01:15:34
rollup I think what you could do is probably sort like here's what I would do i would take the 493 and the filter
01:15:41
that I would apply is what offline assets do they have what online assets
01:15:47
do they have what percentage of those assets are defensible and unique and exist in a postAI world and what
01:15:53
percentage of those assets disappear in a post AI world and I think where I would end up is I'd like own a specialty
01:15:59
chemicals company or something you know like you're still gonna need lubricants and stuff and you can find some way to
01:16:05
make it but if you're like a You need lubricants sorry go ahead you know I love the lubricants
01:16:12
but no dy no Diddy but uh baby oil making you know like five by the crate
01:16:18
timoth do you want to talk about your spack uh tweet uhoh you know the market's back can we see this much can
01:16:24
you play the siren can you play the siren well as with all my tweets like a combo
01:16:30
like a combo beach party as with all my tweets it starts when
01:16:36
Look here here's what X is an incredible platform i use it for Pull up the tweet thing pull up the tweet i use it for a
01:16:41
lot of things but your villain phase right now man you full super villain it's so great the retweet is more
01:16:48
important yeah I love that quote retweet here we go here's the tweet chimamoth says "Incredible that almost 58,000
01:16:54
people voted in his tweet if he should launch a new spa." So uh give the people what they want Chimamoth or what well I
01:17:02
first I first started this because I when I use X sometimes to to just to like sound off because it d-stresses me
01:17:09
during the day okay i like I'll troll people or whatever and then I just did this and I was so impressed that 58,000
01:17:16
people voted but really what happened was I had a lot of very smart money people on Wall Street and some crypto
01:17:22
folks call me that I respect and and basically what they said is like it would be really good if you did it so I
01:17:28
don't know if I'm going to do it but I'm heavily leaning towards doing it well the argument to do it is you learned a lot since last time there's a lot of
01:17:34
inventory there you've got a lot of access to pre-market companies i think what people need to understand is when
01:17:39
you're doing spaxs and correct me if I'm wrong here here's what here's what I'll say Jason this poll and this community
01:17:46
note will be in every single document I do nobody that is listening to this
01:17:51
should participate in this this is going to be for me and a handful of you know
01:17:56
advanced large pools of money you should stay as far away as possible whatever I
01:18:03
do next don't participate in back that's that's the rule here stay on the
01:18:08
sideline do something else don't come in the arena cuz we're trying things timoth don't you have enough going on like why
01:18:14
would you sp Why would you do this when you have fate loves irony fate loves loves irony bro fate loves Absolutely
01:18:19
this will be hilarious it would be the greatest IPO of all time if the poll was
01:18:25
yes I'd be like "Oh [ __ ] this is the last thing I need." All in spa let's go
01:18:30
thomas commentary thomas are you going to buy the all-in spack what's going on the spa market coming back i'm open to
01:18:36
all great companies coming to the public market love it love it i mean but So
01:18:41
Thomas can I ask you a question like tell us about the state of liquidity and actually about IPOs and spaxs in general
01:18:48
like where's your where's your temperature on it just give us a read on what you think i mean look I I think we're getting real world data Chimath
01:18:54
right like in real time um not just from kind of higher visibility companies like Circle and Coree but um Chime also did
01:19:02
really well um Caris uh company you know more in Dave's uh wheelhouse right um
01:19:09
also just coming out so and then wait till we see um the flurry of S1s that
01:19:15
have already been filed right figma is a is a generational potential company right that's going to be coming so I
01:19:22
think we're going to see fantastic assets coming out and I think the market is saying we're open for business the
01:19:28
the MAX 7 is controversial to Dave's point the the S&P 493 there's going to
01:19:33
be lots of winners and losers it's maybe not as obvious there's going to be some dispersion so bring on the new cohort i
01:19:41
think it's the first time you could probably argue that you could go short the S&P Yeah and pick a couple of
01:19:46
winners it's It might be the first time that I would feel in the last 20 years cuz I I'm pretty negative on people
01:19:52
being able to kind of pick stocks but I do think that this is such a transformative moment that if you really
01:19:57
have a sense for what's possible you could start to see category killers emerge out of the S&P and it's an
01:20:03
opportunity to short the S&P and pick a couple winners totally do you Thomas but do you do you care about how these
01:20:09
companies go public like do you care about spack versus direct listing versus IPO like I don't I only care about the
01:20:16
quality of the underlying asset and what I think it can be worth 5 years from now now obviously I do care about the
01:20:22
liquidity that I'm getting in the IPO Chimoth so you know am I getting a
01:20:27
million uh or 100 million or a billion as the float right that's number one and obviously I also do care about the
01:20:35
percentage that is floating and I do care about the lockup right so those those three elements are really
01:20:40
important in terms of a company going public and how we think about participating give the listeners the guidance there so for the first thing
01:20:47
bigger is better than smaller correct so it's number one can I even buy it right if if the IPO is so small um and you
01:20:56
know we can't get a large enough position it doesn't really make sense for us right so that would be kind of
01:21:01
point number one right point number two is how much of the company is publicly floating right
01:21:09
better there as well correct we you kind of get a truer price right when a higher percentage of the company floats um it's
01:21:16
also most likely going to be less volatile and less susceptible chimat to um you know pricing uh predatory pricing
01:21:25
and and manipulation and things like that what's the percentage float that I think 20% is in my opinion kind of a
01:21:33
minimum some have gone out you know I think I remember correct you may know this I think LinkedIn went out at like
01:21:39
10% or something i I remember it being really small and a lot of us thinking like wow that is a that is a small float
01:21:46
yeah yeah which ended up by the way being very volatile so so number two the float and then
01:21:52
number three the lockup right first is there one um in a direct listing there may not be one right so you may you may
01:21:58
get in that scenario to a truer price faster um and Thomas why do you think
01:22:04
there's been no direct listings like why has that totally fallen away after I
01:22:10
mean Spotify did one we did one at Slack and then where where are they like why
01:22:15
why don't people pursue those so here's a statistic i actually had to double
01:22:20
check this because I couldn't believe it right if you look at the cohort of companies that went IPO in 2021 right
01:22:26
and uh and I'm actually not including spaxs in this particular analysis right
01:22:32
if you look at that cohort t plus one year the cohort was down about 40% on
01:22:39
average right okay fine maybe they went up too high 2021 was a peak they didn't do well in one year t plus 5 years it's
01:22:45
down 50% right which which really kind of shocked me right so I think there's kind of scar
01:22:53
tissue on both sides of the table on the buy side about wait hold on what am I really buying and how do I make sure
01:22:59
that um it's kind of a sustainable kind of company but frankly probably also
01:23:05
from boards right who are taking their best assets public and may just want to
01:23:10
um pursue a more conventional approach in the beginning stages right I can tell
01:23:16
you for us direct listing versus IPO makes makes no functional difference you
01:23:21
know I think each has a benefit and I think in some depending on how how concentrated your ownership base is how
01:23:28
understandable your business model is and things like that but we just want these companies to come there's a market
01:23:34
behavior by the way in direct listings and I I've mentioned this once but I've been in two transactions with direct
01:23:40
listings the first was Slack and in the execution of it we misexecuted we
01:23:46
meaning me because I had a huge ownership of Slack but I didn't know
01:23:52
what to do with it and I ended up distributing portions along the way and
01:23:57
it then went through all kinds of turbulence and then it got acquired slightly above the IPO price and what I
01:24:04
learned in retrospect was the best trade is actually the first day trade on a
01:24:09
direct listing so then when it came back around and I got a distribution the day before of Coinbase and and I mentioned
01:24:15
this to Brian this was not a judgment on the company i said "If this direct listing process is going to map to what
01:24:21
I've experienced at Slack the right thing to do is to sell." And I sold that on day one at 335 bucks a share
01:24:30
and ju it's just it's I think Jason it's still not at the IPO price i think it
01:24:35
might be getting close but no it's not back so these Yeah so these direct listings are not what they're expected
01:24:40
to be either yeah if we look back on spaxs I think SoFi is above the price and that might have been one of yours
01:24:46
from Joby getting close these were venture investments these were latestage venture investments in your mind Thomas
01:24:53
and then retail tried to become venture capitalists and they didn't have the 5
01:24:58
10 year horizon that we as venture capitalists have is that your assessment of it and are there any great ones that
01:25:03
came out of the spa movement well I mean the the direct listing era as an example let's talk about Spotify right which
01:25:11
basically has 7xed right over that period so again I it's hard to tell
01:25:18
right causation versus correlation that's why like I think ultimately for me as an ultimate kind of long-term
01:25:24
owner of these businesses I really just care about the quality of the business and whether you chose to go spack or direct listing or IPO is a mechanical
01:25:32
decision um to me the output is quality of business and you know that's
01:25:39
ultimately what wins out okay I want to end on this uh you just shared a chart of app loving and the massive
01:25:46
revenue per employee this is just astounding Thomas apploven as we can see here had 3.6 million revenue per
01:25:53
employee in 21 now up to 7.6 million they peaked at a,000 employees now down to 750ish it looks like in related news
01:26:01
obviously Microsoft we talked about the other week let go of 3% they're planning on massive cuts again for sales these
01:26:07
are organizations that are at record cash record revenue in an industry where we had a tradition of not firing the
01:26:14
gray beards and people had been at the company for more than 10 years andy Jasse didn't come up as like one of the
01:26:19
companies we think is going to win at AI but it might be the company most impacted by deploying AI inside their
01:26:25
enterprise he launched Amissive here it is i suggest everybody read it when you
01:26:30
send a missive like this to your employees you're trying to communicate something to them and to the public markets so he published it on his
01:26:37
website he talks about dozens of AI projects AI tools for advertisers obviously Geni for sellers you know
01:26:44
their product detail page he's talking about Alexa coming back with a brand new version shopping assistance everything
01:26:52
but then he started talking about the work force size he says in this manifesto in the next few years we
01:26:58
expect this will reduce our total corporate workforce as we get efficiency gains from using AI extensively across
01:27:04
the company so my question to you Thomas is when you hear public CEOs talking
01:27:10
about lowering the number of employees while they're growing 10 20% per year
01:27:15
this is obviously awesome for earnings the share price but there's going to be massive job displacement any thoughts on
01:27:22
the job displacement job replacement and society navigating that and just as well
01:27:29
Andy Jasse specifically and what you think of Amazon as a business and them
01:27:34
being a player in AI and AI being a player in their business you know I think it's a it's an important question
01:27:40
and I'll defer to what Jensen answered on this topic because in my view it's still the most credible and cohesive answer I've kind of heard right and
01:27:48
Jensen is known uh the CEO of Nvidia an an incredibly long-term thinker and in
01:27:53
his view is he looks at a population that's getting older and he wonders who are going to be all the young people
01:27:58
that are going to take care of all the old people whether it's nurses or doctors or other things like that and in his view we better get a lot more
01:28:05
productive right to deal with our inverted demographic table so I ultimately think this is going to enable
01:28:12
more young people to take care of more old people right and it's just going to create I think knowledge workers are
01:28:18
incredibly flexible they can take their tools from you know one particular skill set to another so I think this is going
01:28:24
to unleash incredible opportunities for the economy i think it is going to make
01:28:30
us more productive and wealthier so I'm definitely on the more optimistic side of the scenario
01:28:36
shimoth any thoughts on Amazon they didn't come up but obviously AWS crushing it and they're a major player
01:28:43
and they have their own silicon they're making you mentioned that being an important part of the stack and then you
01:28:48
have Optimus and robots figure that are going to be in their factories that's a lot of jobs delivery robots they're
01:28:55
doing drones like Zipline they have their own version of it obviously and they're doing zuks so if you just look
01:29:00
at their behavior and you look at their investments they're massively massively investing in robotics self-driving and
01:29:07
chips so they're pretty hardware focused yeah for physical AI they're a kingmaker in parts because they're a a sync for
01:29:15
demand so they'll just generate so much demand for robots so if Figure lands the BMW or the UPS robot successfully Amazon
01:29:23
will buy a gajillion of them if Optimus lands a successful robot that they tune inside the Tesla factory and then are
01:29:29
ready to sell Amazon will buy a gajillion of them if there are drones that are delivering things Amazon will
01:29:36
buy a gajillion of them so on the one side there's a lot of typical opex lift that Amazon will get i think the problem
01:29:43
is more with AWS which is that their success is actually their biggest bottleneck the success is that they're
01:29:49
not necessarily kingmaking they're about being a purveyor of many many many
01:29:55
different things that you can find inside of AWS marketplace and so you
01:30:00
know the the thing that they'll have to embrace is well do I differentiate my
01:30:08
own hardware from Nvidia's at some point do I actually make a real bet on models
01:30:13
and try to frankly buy anthropic which is probably their only solution and tightly couple it in and say that you
01:30:20
know if you want to have next generation codegen experiences they need to run inside of AWS
01:30:25
these are the difficult decisions that I think that Andy will have to face and he's going to have to spend hundreds of billions of dollars but yeah the the
01:30:33
Amazon retail side is going to be a kingmaker for all of these physical AI things
01:30:39
freeberg any thoughts on Amazon just as a company broadly chamat saying "Hey they're a kingmaker." That seems like a
01:30:46
really interesting insight you have any insights there on Amazon and they're playing a part here in the future of AI
01:30:52
i don't thomas any closing thoughts here on you know the sort of old old guard
01:30:59
Microsoft Amazon and their employee count and the cuts we're seeing there uh
01:31:05
and what these companies will look like in the future in terms of revenue per employee they're not hiring young people
01:31:11
they're getting rid of the old folks they're just advancing it seems at a at a they're adopting AI pretty uh severely
01:31:18
at these companies what are your thoughts there i'm going to play I'm going to play the role of JCAL and I'm going to ask a question to all three of
01:31:24
you guys oh here we go so Microsoft's employee count peaked at about 250,000
01:31:30
you know call it about a year ago who here believes that in 5 years Microsoft
01:31:36
will have more employees than it does today more
01:31:42
i'm going to say the same i think they'll have just about 250 plus or minus 10% i don't think if I if I could
01:31:48
pick push as the answer I would pick push which is they're going to get 10% better every year with AI 20% more
01:31:54
efficient therefore they don't need to add people but I also don't think they atrophy much more so maybe they have 225
01:32:01
250 why Why' you say more so quickly i'm curious oh so this chart which I think
01:32:06
is like a very dangerous vanity metric
01:32:12
is why so what Microsoft touts is what percentage of code is generated by AI
01:32:18
without answering the more important question which is is that code useful and good and if you ask that second
01:32:25
layer Nick I sent you this tweet from Yan Lun and I'll tell you that this is my lived experience as well is most code
01:32:33
generated by AI is crap and most of the tools that we use you know the reason we
01:32:40
call these tools app crappers is because most of the code that it generates is crap so it's great in a single player
01:32:48
mode but transitioning from single player mode to a complex enterprise environment is not possible today so I
01:32:56
think that Microsoft puts these metrics out because they want to seem that they're on the front line of it but I
01:33:02
suspect that this is just like you know how you used to hire Mackenzie consultants to fire people because it was good air cover it's probably just
01:33:09
air cover to fire a bunch of folks that they probably wanted to get rid of anyways but it's not related to that
01:33:14
chart and the reason is that Yan Lun's tweet is true when you allow these
01:33:19
models to run over complicated tasks over long periods of time the error rates compound to such a degree that the
01:33:25
that the resulting output is not worthwhile and so until that problem is fixed which I'm sure it will be and I
01:33:32
and I and I'm going to bet that it will be the idea that all of a sudden it's because of coding agents that people are
01:33:38
getting laid off I think is a fallacy so I suspect that Microsoft business on the margin grows back to Dave's point some
01:33:44
of the 493 shrink and go away it'll be cheaper for Microsoft to bundle together a bunch of other products that are point
01:33:51
features today and so they'll have more people they'll need more the people will be different they'll have different skill sets but I suspect Microsoft's
01:33:57
employee base grows freeberg what say you i think shrink wow so by the way
01:34:04
pretty interesting to think about we have one decisively more one medium about the same a and a and a less i only
01:34:13
say that because I do think that there's a real probability of revenue decline in the next 5 years so if you look at the
01:34:19
enterprise install base I think that cloud gets competed away so I do think like on the on the application software
01:34:26
layer they're going to have a really hard time in this new world because the old school customers that buy Microsoft
01:34:32
are going to die they're more likely to die in their marketplace compared to the folks that are going to build native
01:34:38
software native workflows and I'm not really where Chimoth is i think you may be right about where AI written code is
01:34:44
today i I don't think that that's true 3 years from now four years from now given the pace of improvement and so in a
01:34:51
world where you have software written workflows built for you through agentic tools I think that Microsoft's core
01:34:59
business for the is going to decline the the losers are their biggest customers and the winners are not going to use
01:35:04
them so I I you know that that would be Where are you at Thomas maybe you're the tiebreaker
01:35:10
i I'm in Chamas camp where I actually think the Microsoft business will be bigger if anything on on kind of alone
01:35:18
and that at the end of the day uh we'll just need more people to support it i just think they'll be they'll be more
01:35:24
relevant they'll have more productive employees but they'll still be more of them i'm
01:35:29
predicting incredible growth and the same number of employees so you guys are predicting incredible growth and
01:35:35
employee growth i think that that's interesting so So sorry less revenue less employees interesting so the thesis
01:35:41
as as your grows um is basically where the the application dollars go
01:35:49
effectively is one way to think about this right so application dollars go there and that more than makes up for
01:35:54
the decline in in that business over time right and there's multiple clouds by the way I went to the Google Next
01:36:01
event last year and so I I ended up going to these like special dinners or whatever a couple cocktail dinner thing
01:36:08
because I spoke there and I saw they put me with a bunch of these people and I CIOS of you know whatever Fortune50
01:36:13
companies and all of them said that they're multicloud like they're not no one's going to standardize on one cloud
01:36:19
so everyone has to be on Microsoft and Google and I had never really recognized this or thought about this as being a a
01:36:26
fact that it's not necessarily the best or the lowest price at the end of the day these guys are going to distribute
01:36:31
their exposure and so I think that maybe supports your case i'm very easily con
01:36:36
I'm very easily able to see other arguments today i'm very convinced here's the revenue what a spectacular revenue run uh just I think all four of
01:36:45
us would agree that if we could synthetically own AWS Azure and GCP if I
01:36:51
could somehow automatically create an index of all three of those businesses right over the next five years yeah yeah
01:36:58
you wouldn't need to own anything else you wouldn't need own anything else i wish Elon would take that so why don't
01:37:03
you put up with the shitty part of the rest of their businesses and just own all three and that's it call it a day cuz you've got to assume that if one of
01:37:10
them wins over the other two or accelerates ahead of the other two it's going to more than make up for the losses that the other two might
01:37:15
experience in their other businesses the multiples aren't crazy on those three companies by the way correct quite reasonable yeah i think if Elon took
01:37:22
what he did with Colossus and he had an AWS competitor he would be a serious competitor in the space but this is like
01:37:28
this the velocity at which he can build out data centers is extraordinary this is where Elon does better because he can
01:37:34
actually get a better like um fundraising uh in the private market
01:37:39
with XAI than what he has to deal with yeah he's really he's really struggling with that that's what I'm saying yeah no
01:37:46
no I'm saying it's better for him right hey guys look who's here couldn't stay away david Sachs look at here you can't
01:37:53
get away from it 11 o'clock happens on a Thursday and you start jonesing for your besties welcome to the ZAR David S good
01:38:01
to be back Jacob where are you in LA mhm i'm in LA this is You're at someone's guest house yeah actually this is one of
01:38:07
your guest houses you You just lost track i still have I still have the key code it's a J Cal Kalen jay Calen is at
01:38:15
your guest house jalen jalen here i'm here come down the hill he'll still get that
01:38:21
reference it's getting kind of dated now oh god ko Kalan is ride or die i mean he
01:38:26
would jump on a a vente or a grande for you for sure let's talk a little bit here since I got you Saxs would you be
01:38:31
willing to talk a little bit about the Genius Act we just passed the Senate i think you have your fingerprints on this
01:38:37
is that true yeah tell us everything well it's definitely something we supported and this is I think a huge
01:38:44
milestone i mean just uh you know what basically happened is we had this genius act which is the stable coin bill passed
01:38:50
the Senate with 68 votes got 18 Democrats they came on board we had to
01:38:56
hit that key threshold of 60 votes in the Senate that's the threshold you need in the Senate unless you know it's it's
01:39:03
um a narrow exception for reconciliation so it's very very hard to pass any bill
01:39:08
out of the Senate and you need a significant amount of bipartisan support and we got that now when you consider
01:39:15
where we were a year ago you know you realize what huge progress this is for
01:39:20
the crypto industry a year ago you had crypto companies being prosecuted you
01:39:25
had this whole regulation through prosecution approach where Gary Gendler who was the chair of
01:39:31
the SEC then he wouldn't tell startups what the rules were but they would just announce prosecutions and this was
01:39:37
driving all the crypto innovation offshore and I think we were basically poised to lose the crypto industry in
01:39:43
the United States what happened then is President Trump adopted this cause he announced that he wanted to make uh the
01:39:49
United States the crypto capital of the planet he really campaigned on this and as part of his administration he in the
01:39:56
very first week signed a new executive order making it clear that his administration supported crypto we've
01:40:01
been rooting out all the Biden war on crypto rules and regulations at the agency level and now we have this first
01:40:08
major legislative win and I would expect the House will act in the next few weeks on this and then the president will have
01:40:14
a bill he can sign this is uh great work and it's really important because to your point Gary Gensler's concept was
01:40:22
hey there's an existing playbook there's existing rules just follow those but none of these things actually match the
01:40:28
existing rules perfectly so you need some new rules they need to evolve it was much worse than that because he
01:40:34
would say things like "Well just come into the SEC and talk to us." You know so in other words you got to come in and talk to us and get our approval but then
01:40:40
when startups would go in there and talk to them there'd be enforcement people there writing down everything they said
01:40:45
and the next day they get a Wells notice and they would get investigated honey yeah they were honeypotted basically
01:40:52
and so the the response the industry was "Okay we're just going to leave the United States." And that that was what
01:40:58
was in the process of happening until President Trump won the election and then changed the tone in Washington i
01:41:03
think there was one other really significant thing that that happened because you know obviously President
01:41:09
Trump has gotten Republicans on board with this cause but the question is why are Democrats on board with it during
01:41:15
the Biden administration Elizabeth Warren really called the shots on crypto
01:41:20
and it was well reported that Gendler was was sort of her ally and her pick i've kind of joked that Warren
01:41:27
controlled the Biden autopen on on crypto because she really did exert that kind of influence
01:41:32
so the question is well what changed and I think one of the big things is that in this last election Sherid Brown who was
01:41:39
the chair of the banking committee for the Democrats in the Senate lost his seat in a close election against Bernie
01:41:45
Moreno and I think there were many reasons for him to lose that seat he was far to the left of voters in Ohio
01:41:52
nonetheless he had been a successful politician there for a long time and one of the reasons why he lost is because
01:41:58
the crypto industry really got behind Bernie Mareno because Sher Brown was just a a total blocker to any crypto
01:42:06
legislation in the mold of Elizabeth Warren and I think that a lot of smart Democrats looked at that and said "Why
01:42:12
are we dying on this hill again?" You know and I think she Well it's also extraordinarily popular sachs with
01:42:18
consumers and businesses so there is a demand here and Korea we've got you've
01:42:24
got something like 50 million wallet holders in the US and their their their voters so that's one out of five
01:42:29
Americans adult so I think a lot of Democrats said "Well wait a second why are we just blindly following Elizabeth
01:42:34
Warren on this what exactly is so harmful about this?" Particularly when what we're talking about here is
01:42:40
creating a regulatory regime you know it shouldn't be hard to sell Democrats on
01:42:45
new regulations uh but in this case the reason why there's broad bipartisan support is because the crypto industry
01:42:52
itself is calling for those regulations because having regulatory certainty is
01:42:57
better for them than the possibility of the return of a Gary Gendzer-L like figure who just prosecutes them without
01:43:04
telling them what the rules are so this is why I think you're getting some significant bipartisan support and and
01:43:10
as you said bringing this on shore is such a great portion of it there are tons of actors who some people might
01:43:16
describe as bad or gray or dark tether comes to mind with a lot of regulation
01:43:22
against it and now those folks who are running away with the industry Thomas
01:43:27
now they have to compete with people like Jeremy Circle which are totally buttoned up here in the United States
01:43:33
and it levels the playing field so it's an example of actually good regulation bringing
01:43:38
this opportunity back on shore and taking it out of the gray area just on the whole offshore versus onshore so it
01:43:45
is true that the number one stable coin issuer on the planet right now is an offshore company and that is partly
01:43:52
because there has not been a regulatory framework in the US and there's been hostility towards the crypto space and
01:44:00
so the logical reaction to that is to either not get involved in the crypto
01:44:06
space which is what the banks have done until now or you go offshore neither one is good and you know you can see in the
01:44:12
wake of this Genius Act the stable coin bill that the banks have now talked
01:44:17
about getting into stable coins they're going to issue one and then also Tether will under this act will have three
01:44:24
years to come on shore but the bottom line is they will have to operate in the United States and that's a good thing
01:44:31
for consumers it's a good thing for they three years to get compliance they have three years but they have to move on
01:44:36
shore now all stable coin issuers under this bill will have to be audited
01:44:41
quarterly and by a real audit not this attestation nonsense like real audits by
01:44:46
American real audits and it will verify that every stable coin that's been
01:44:52
issued is backed or fully reserved on a onetoone basis with real dollars in an
01:44:58
American bank accounts that are in US T bills or money market accounts and so it
01:45:04
what it does is by the way I'm not saying there's anything wrong with Tether but this does provide additional certainty and confidence because you
01:45:12
know that all the companies are onshore and they've been fully audited and we
01:45:18
know that they're fully reserved so that when you want to redeem and cash out your stable coin tokens there's a real
01:45:24
dollar waiting there to cash out you prevent the undercolateralization issue yeah and and by the way I'm not
01:45:31
saying that there is but but what I'm saying is now we create total certainty and confidence which is good for the
01:45:37
market what happens if a stable coin issuer does not like can you issue US
01:45:42
dollar stable coins and not be governed under this system or no you're saying because the US dollar is a US government
01:45:48
instrument then no matter where you are or no matter where you issue from yeah all the issuers will be governed by this
01:45:55
and if you're a legacy offshore issuer you're given this time period to bring yourself into conformity but yeah
01:46:01
otherwise what happens if if they don't Well it's a good question i mean I guess
01:46:06
the exchanges won't be able to carry their their tokens and they won't be able to set foot in
01:46:12
the US they'll be in violation of US law it's just not a good place to be yeah i mean you don't have to guess um there
01:46:17
have been dozens of actions and accusations like legitimate ones against
01:46:22
Heather new York's Attorney General did a major settlement with him in 2021 they've been banned from many
01:46:27
jurisdictions and uh in Senate hearings yeah tether should just Tether should
01:46:33
just go public in America and be done with it well and the issue was there was deep concerns that they didn't have the
01:46:39
deposits and now they're they're really trumpeting the fact that they're massively profitable obviously so
01:46:45
there's been tons of uh you can just search Tether and allegations and you'll find all that stuff we don't trash it
01:46:50
here tether founder is Italian sax I got to give you a lot of credit we knew that you would bring an efficiency level and
01:46:58
some expertise to this administration but I got to give you your flowers we're 5 months into this administration can
01:47:03
disagree about many things one thing we can't disagree about is that this piece of legislation uh is here and we're
01:47:10
we're only 5 months in so maybe you could speak to the velocity at which things are getting done and then uh any
01:47:16
other clothing closing thoughts i know you got to get back to your day job jake how a lot of people deserve credit for this i just want to give out a couple of
01:47:22
shout outs so Senator Bill Hagerty from Tennessee was the principal author of the legislation he did an amazing job
01:47:27
getting Democrat votes and also bringing the Senate bill into greater alignment with the House bill so hopefully this
01:47:33
can pass the House very quickly chairman Tim Scott who's the chairman of the banking committee was also incredible
01:47:39
the majority leader uh John Thun and then we had uh a few co-sponsors of the legislation Cynthia Lemus from Wyoming
01:47:46
and then two Democrats actually were really important kirsten Gillibrand from New York and Angela also Brooks from
01:47:51
Maryland all them did a great job and we've got great leaders on the House side as well french Hill who's the chairman of the House Financial Services
01:47:57
Committee Tom Emmer who's the whip and Mike Johnson who's the speaker so kudos
01:48:03
to all of them because I think that it really is a pretty incredible achievement that they've been able to
01:48:08
get this bill through again just a huge sea change from where we were a year ago where crypto was basically under attack
01:48:15
it was being driven offshore and now we have it as one of the first major piece of legislation by this new Congress and
01:48:21
again that's all because of President Trump's leadership and prioritization of this issue so thank you to all of them
01:48:26
for making this happen congratulations to you David hey uh one uh tactical question I forgot to ask you the float
01:48:32
on these this is like how Tether is making billions of dollars a year and this is how people anticipate they're going to make billions of dollars a year
01:48:38
are they able to split that with consumers yet because I I remember reading in early legislation that you
01:48:43
weren't allowed to pass on the interest made from a stable coin to like the consumers I guess so you wouldn't it
01:48:50
couldn't be an interest bearing account if you buy stable coins you can't get interest on it but the issuer like Circle that's their main business model
01:48:56
so did that make it into the final and and maybe you can give us some background on that no
01:49:01
no it did not the way the framework works is that the stable coin issuers cannot pass on interest why to the token
01:49:08
holders why is that look I mean I don't know if there's a great principled reason this was a compromise that was
01:49:13
necessary to get the support of the banking industry quite frankly ah they see it as competition i'm betting well
01:49:18
there was a lot of concern from community banks that if stable coins were paying 5% interest it would put them out of business personally I think
01:49:25
that that concern although understandable from them I don't think that that's what would have happened but
01:49:32
these are the types of compromises quite frankly that you need in order to pass legislation i hope that at some point in
01:49:37
the future we'll revisit that and allow stable coin issuers to kind of just do
01:49:42
what they want to do all right and that'll be easier once the banks get into the act and they're participating
01:49:47
in this industry got it but right now they're total outsiders and you can understand the fear factor all right Sax
01:49:53
we're going to drop you off man i wish we could have you on for the full show but uh you you're busy you got a lot of things to do love you David shed a
01:49:59
little tear and I'll miss my bestie see you soon thanks guys all right back at you we got two hours of classic Allin uh
01:50:05
in part two of the show we're going to do an hour and a half on the Israeli
01:50:10
conflict with Iran we've got 90 more minutes coming up for you and uh we've got Ukraine Ukraine Ukraine Mir Shimemer
01:50:17
and uh Jeffrey Saxs joining us in the second in the third and fourth hour of the All-In podcast how's the All-In
01:50:24
Summit going Friedberg how's the All-In Summit you know we might get uh Jason Shai wants to come from uh Alibaba who's
01:50:32
in touch with him i am you thanks to Phipe i just want to do one quick shout out to our friend and fellow bestie
01:50:38
Vinnie Lingum oh yes he has movies coming out a friend of ours did a
01:50:43
documentary on It's great Freeberg you're going to love this on only I denounce I denounce I love Vinnie i
01:50:49
denounce it so great anyways it's called Animal oh it's great
01:50:56
Doc and uh perfect can't wait where can where can people watch it i
01:51:01
think he's got a couple of deals come to your local slaughter house and put it on your phone and watch it at the slaughter
01:51:07
house while you're there here's the idea you're going to consume a certain number of calories from us humans were designed to eat meat that's the number one thing
01:51:13
we should be doing as a species is eating meat nick can you put the trailer in the show notes so that he can get a
01:51:19
little play actually play us out with the trailer you can play us out with the trailer on the show we'll do it myself all right guys i got to go eat i have a
01:51:25
photo shoot in 2 hours i love you got a photo shoot is it going to be you showing the legs or just the top this
01:51:30
time what are you showing what are you shooting i'm going to do blur out the anaconda you should do pixelate the anaconda
01:51:36
i hope it's Italian vogue what are you shooting thomas is in the general neighborhood i I can't comment but uh
01:51:42
tell us bleep it out nice tell I should bleep it out jam give me a call i got to talk to you about this weekend okay love
01:51:48
you guys talk to you guys i'll see you at Are you guys still doing the tequila launch yeah Saturday night we'll see you
01:51:53
Saturday night absolutely see you there go to allin.comyatta yada yada to sign up for the all-in
01:51:59
summit apply there for Thomas Lefant Shimatia Dave Freeberg and the Zar David
01:52:04
Saxs i am the world's greatest executive producer we'll see you next time jason atallin.com byebye adios play the
01:52:13
trailer we're too good of hunters we came out of the trees not to eat the
01:52:19
grass but to eat the grass eaters meat is the most nutrientdense food that
01:52:26
human beings can eat we're carnivores but we're not living as carnivores
01:52:32
we are just better designed and more efficient at getting nutrition from meat got to remember where we came from and
01:52:38
what our food should be it will change your life
