I will never forget when I first met

Gavin Baker. It was early days of the

podcast and he was one of the first

people I talked to about markets outside

of my area of expertise which at the

time was quantitative investing about

the incredible passionate experience

that he's had investing in technology

across his career. I find his interest

in markets, his curiosity about the

world to be about as infectious as any

investor that I've ever come across. He

is encyclopedic on what is going on in

the world of technology today. and I've

had the good fortune to host him every

year or two since that first meeting on

this podcast. In this latest

conversation, we talk about everything

that interests Gavin. We talk about

Nvidia, Google and its TPUs, the

changing AI landscape, the changing math

and business models around AI companies.

This is a life ordeath decision that

essentially everyone except Microsoft is

failing it. We even discussed the crazy

idea of data centers in space which he

communicates with his usual passion and

logic.

>> In every way, data centers in space from

a first principles perspective are

superior to data centers on earth.

Because Gavin is one of the most

passionate thinkers and investors that I

know, these conversations are always

amongst my most favorite. I hope you

enjoy this latest in a series of

discussions with Gavin Baker.

I would love to talk about how you like

in the nitty-gritty process new things

that come out in this whole like AI

world because it's happening so

constantly. I'm extremely interested in

it and I find it very hard to keep up

and I you know I have a couple blogs

that I go read and friends that I call

but like maybe let's take Gemini 3 as

like a recent example when that comes

out. What like literally like take me

into your office like what are you

doing? How do you and your team process

an update like that given how often

these things are happening?

>> I mean, I think the first thing is you

have to use it yourself.

>> And I would just say I'm amazed at how

many famous and August investors are

reaching really definitive conclusions

about AI. Well, no, based on the free

tier.

>> The free tier is like you're dealing

with a 10-year-old

>> and you're making conclusions about the

10-year-old's capabilities as an adult.

And you could just pay and I do think

actually you do need to pay for the

highest tier whether it's Gemini Ultra,

you know, um, Super Grock, whatever it

is, you have to pay the $200 per per

month ti whereas those are like a

fullyfledged 30 35y old. It's really

hard to extrapolate from an eight or a

10-year-old to the 35y old and yet a lot

of people are doing that. And the second

thing is there was a insider post about

open AI and they said to a large degree

open AI runs on Twitter vibes

>> and I just think AI happens on X and you

know there have been some really

memorable moments like there was a giant

fight between the PyTorch team at Meta

and the Jax team at Google on X and the

leaders of each lab had to step in

publicly say

>> no one from my lab is allowed to say bad

things about the other lab and I respect

them and that is the end of that.

>> Yeah,

>> the companies are all commenting on each

other's posts. You know, the research

papers come out. You know, if on planet

Earth there's 500 to a,000 people who

really really understand this and are at

the cutting edge of it and a good number

of them live in live in China. Um I just

think you have to follow those people

closely

>> and I think there is incredible signal

to me. Everything in AI is just

downstream

>> of those people.

>> Yeah. Everything Andre Carpathy writes,

you have to read it three times.

>> Yeah.

>> Minimum.

>> Yeah. He's incredible.

>> And then I would say anytime at one of

those labs, the four labs that matter,

you know, being uh OpenAI, Gemini,

Anthropic, and XAI, which are clearly

the four leading labs. Like anytime

somebody from one of those labs goes on

a podcast, I just think it's so

important to listen. And then for me,

for me, one of the best use cases of AI

is to keep up with all of this. You

know, just like listen to a podcast and

then if there are parts that I thought

were interesting, just talk about it

with AI. And I think it's really

important to like have as little

friction as possible, I'll bring it up.

You know, I have it like um you know, I

have I can either press this button and

pull up Gro or I have this.

>> Oh, wow. I don't touch that. That just

get brings it right up.

>> Yeah, it brings it right up. What do you

think of Patrick Oanaughy?

>> Oh man, Patrick Oshanaugh is one of my

favorite voices in investing. His Invest

Like the Best podcast is straight fire.

Does deep dives with folks like Bill

Gurly or

>> Girly? Yes.

>> It's so Can you believe we have this?

>> I know. It's like we have Yeah. I think

somebody said on on on X, you know, like

we imbued these rocks with crazy spells

and now we can summon super intelligent

genies

on our phones over the air. You know,

it's crazy

>> crazy.

>> So something like Gemini come 3 comes

out, you know, the public interpretation

was, oh, this is interesting. It seems

to say something about scaling laws and

the pre-training stuff. What is your

frame on like the state of prog general

progress in frontier models in general?

Like what are you watching most closely?

>> Yeah. Well, I do think Gemini 3 was very

important because it showed us that

scaling laws for pre-training are

intact. They, you know, stated that

unequivocally and that's important

because no one on planet Earth knows how

or why scaling laws for pre-training

work. It there it's actually not a law.

It's an empirical observation and it's

an empirical observation that we've

measured extremely precisely and has

held for a long time. But our

understanding of scaling laws for

pre-training and maybe this is a little

bit controversial with 20% of

researchers but probably not more than

that is kind of like the ancient British

people's understanding of the sun are

the ancient Egyptians understanding of

the sun. They can measure it so

precisely that the east west axis of the

great pyramids are perfectly aligned

with the equinoxes and so are the east

axises of Stonehenge. Perfect

measurement.

>> But they had they didn't understand

orbital mechanics. They had no idea how

or why, you know, it, you know, rose in

the east, set in the west, and, you

know, kind of moved across the horizon.

>> The aliens.

>> Yeah. Our god in a chariot. And so it's

really important every time we get a

confirmation of that.

>> Um, so Gemini 3 was very important in

that way. But I'd say I think there's

been a big misunderstanding maybe in the

public equity investing community or the

broader more generalist community based

on the scaling laws of pre-training.

There really should have been no

progress in 24 and 25.

>> And the reason for that is, you know,

after XAI figured out how to get um

200,000 hoppers coherent,

>> you had to wait for the next generation

of chips.

>> Um because you really can't get more

than 200,000 hoppers coherent. And

coherent just means you could just think

of it as each GPU knows what every other

GPU is thinking. They kind of are

sharing memory. You know, they're

connected. They scale up networks and

scale out. and um and they have to be

coherent um for during the pre-training

process. And I think there's a lot of

misunderstanding about Gemini 3 that I

think is really important. So everything

in AI has a struggle between Google and

Nvidia and Google has a TPU and Nvidia

has their GPUs and each of I mean Google

only has a TPU and they use a bunch of

other chips for networking. You know

Nvidia has the full stack and Blackwell

was delayed. Blackwell was Nvidia's next

generation chip and

the first iteration of that was the

Blackwell 200. A lot of different SKs

were cancelled and the reason for that

is it was by far the most complex

product transition we've ever gone

through in technology. Going from Hopper

to Blackwell, first you go from air

cooled to liquid cooled. Um the rack

goes from weighing round numbers 1,000

lb to 3,000 lb. goes from round numbers

30 kilowatts which is 30 American homes

to 130 kilowatts which is 130 American

homes you know. So I I analogize it to

imagine if to get a new iPhone you had

to change all the outlets in your house

to you know 220 volt put in a Tesla

power wall put in a generator put in

solar panels that's the power you know

put in a whole home humidification

system and then reinforce the floor

because you know the floor can't handle

this. So it was a huge product

transition and then just the rack was so

dense it was really hard for them to get

get the heat out. So Blackwells have

only really started to be deployed and

really scaled deployments over the last

3 or 4 months. Had reasoning not come

along, there would have been no AI

progress

from mid 2024

through essentially Gemini 3. there

would have been none. Everything would

have stalled and can you imagine what

that would have meant to the markets

like for sure we would have lived in a

very different environment. So reasoning

kind of bridged this like 18month gap.

Reasoning kind of saved AI because it

let AI make progress without Blackwell

or the next generation of TPU which were

necessary for the scaling laws for

pre-training to continue. The reason

we've had all this progress, maybe we

could show like the ARC AGI slide where

you had, you know, you went from 0 to 0

to 8 over four years, 0 to 8%

intelligence

>> and then you went from 8% to 95% in 3

months when the first reasoning model

came out from OpenAI is, you know, we

have these two new scaling laws of post-

training, which is just reinforcement

learning with verified rewards. Verified

is such an important concept in AI. Um,

like one of Karpathy's great things was

with software, anything you can specify,

you can automate. With AI, anything you

can verify, you can automate. It's such

an important concept and I think an

important distinction. And then test

time compute. And so all the progress

we've had, and we've had immense

progress um, since October 24th through

today was based entirely on these two

new scaling laws. And Gemini 3 was

arguably the first test since Hopper

came out of the scaling law for

pre-training and it held. And that's

great because all these scaling laws are

multiplicative. So now we're going to

apply these two new um reinforcement

learning with verified rewards and test

time compute um to much better base

models. Google came out with the TPU v6

in 2024 and the TPU v7 in 2025.

And in semiconductor time, it's like

almost like imagine like Hopper is like,

you know, it's like a World War II era

airplane. And it was by far the best

World War II era airplane. It's P-51

Mustang with the Merlin engine. And two

years later in semiconductor time,

that's like,

>> you know, you're an F4 Phantom. Okay.

Because Blackwell was such a complicated

product and so hard to ramp, Google was

training Gemini 3 on 24 and 25 era TPUs,

which are like F4 Phantoms. Like

Blackwell, it's like an F-35.

>> It just took a really long time to get

it going.

>> So, I think, you know, Google for sure

has this temporary advantage right now.

Um, from a pre-training perspective, I

think it's also important that they've

been the lowest cost producer of tokens.

Okay. And this is really important

because AI is the first time in my

career as a tech investor that being the

lowcost producer has ever mattered.

Apple is not worth trillions because

they're the lowcost producer of phones.

Microsoft is not worth trillions because

they're the low lowcost producer of

software. Nvidia is not worth trillions

cuz they're the lowcost producer of AI

accelerators. It's never mattered. And

this is really important because what

Google has been doing has the lowcost

producer is they have been I would say

sucking the economic oxygen out of the

AI ecosystem which is an extremely

rational strategy for them and for

anyone who's a lowcost producer you know

let's just let's make life really hard

for our competitors. Um and so what

happens now I think this has pretty

profound implications. One, we will see

the first models trained on Blackwell in

early 2026.

>> Y

>> I think the first Blackwell model will

come from XAI. And the reason for that

is just it's a according to Jensen, no

one builds data centers faster than

Elon. Yes, Jensen has said this on the

record. Even once you have the

Blackwells, it it takes 6 to9 months to

get them performing at the level of

Hopper

>> cuz the Hopper is finally tuned.

Everybody knows how to use it. The

software is perfect for it. engineers

know all its quirks. You know, everybody

knows how to architect a Hopper data

center at this point. And by the way,

when Hopper came out, it took 6 to 12

months for it to really outperform AER,

which was generation before. So, if

you're Jensen or Nvidia, you need to get

as many GPUs deployed in one data center

as fast as possible in a coherent

cluster so you can work out the bugs.

And so this is what XAI effectively does

for Nvidia because they build the data

centers the fastest. They can deploy,

you know, black wells that scale the

fastest and they can help work with

Nvidia to work out the bugs for everyone

else. So because they're the fastest,

they will they'll have the first

Blackwell model. We know that scaling

laws for pre-training are intact and

this means the Blackwell models are

going to be amazing. Blackwell is um I

mean it's not an F35 versus an F4

Phantom, but from my perspective it is a

better chip, you know, maybe it's like

an F-35 versus a Raphael. And so now

that we know pre-scaling holding, we

know that these Blackwell models are

going to be really good.

>> And you know, kind of based on the raw

specs, they should probably be better.

>> Then something even more important

happens.

>> So the GB200 was really really it was

really hard to get a coin. Um,

the GB300

is a great chip. It is drop in

compatible in every way with those GB200

racks. Now, you're not going to replace

the GB200s. No new power walls. Yeah.

>> Yeah. Just any data center that can

handle those. You can slot in the

GB300s. And now everybody's good at

making those racks and you know how to

get the heat out. You know how to cool

them.

>> You're going to put those GB300s in and

then the companies that use the GB300's,

they are going to be the lowcost

producer of tokens.

particularly if you're vertically

integrated. If you're paying a margin to

someone else to make those tokens,

you're probably not going to be. I think

this has pretty profound implications

because it ch I think it has to change

Google's strategic calculus. If you have

a decisive cost advantage and you're

Google and you have search and all these

other businesses, why not run AI at a

negative 30% margin?

>> It is by far the rational decision. You

take the economic oxygen out of the

environment. You eventually make it hard

for your competitors who need funding

unlike you to raise the capital they

need. And then on the other side of

that, maybe have an extremely dominant

share position. Well, all that calculus

changes once Google is no longer

>> the lowcost producer, which I think will

be the case. The black wells are now

being used for training. And then when

the that model is trained then you shift

you start shifting blackwell clusters

over to inference and then all these

cost calculations and these dynamics

change

>> and I do think it's this it's very

interesting like during the strategic

and economic calculations between the

players. I've never seen anything like

it. You know everyone understands their

position on the board, what the prize

is, you know what play their opponents

are running. Um, and it's really

interesting to watch. So, I just think

if Google changes its behavior, cuz it's

going to be really painful for them as a

higher cost producer to run that

negative 30% margin, it might start to

impact, you know, their stock. That has

pretty profound implications for the

economics of AI. And then when Reuben

comes out, we'll know the gap the gap is

going to expand significantly

>> versus TPUs.

>> Versus TPUs and and all other AS6. Now,

I think tranium 3 is probably going to

be pretty good. train for are going to

be good.

>> Why is that the case? Why why won't TPU

v8 V9 be every bit as good?

>> A couple of things. So one um for

whatever reason um Google made more

conservative design decisions.

I think part of that is so Google round

numbers Google like let's say the TPU

Google is so there's front end and back

end of semiconductor design and then

there's you know uh dealing with Taiwan

semi. You can make an ASIC in a lot of

ways. What Google does is they do mostly

the front end for the TPU and then

Broadcom does the back end and manages

Taiwan mean everything. It's a crude

analogy but like the front end is like

the architect of a house.

>> Yep.

>> They design the house. The back end is

the person who builds the house and the

managing Taiwan Simmyi is like stamping

out that house like LAR or you know Dr.

Horton and for doing those two ladder

parts broadcoms a 50 to 55% gross

margin. We don't know what on TPUs.

Okay, let's say in 2027

TPU I think it sits estimates maybe

somewhere around 30 billion again who

knows I mean

>> yeah yeah yeah but I 30 billion I think

is a reasonable estimate 50 to 55% gross

margins so Google is paying Broadcom $15

billion

>> okay that's a lot of money

>> and at a certain point it makes sense to

bring a semiconductor program entirely

in house so in other words Apple does

not have an ASIC partner for their chips

>> they do they do the front end themselves

the back end and they manage Taiwan semi

and the reason is they don't want to pay

that 50% margin so at a certain point it

becomes rational to renegotiate this and

just as perspective the entire opex of

Broadcom's semiconductor division is

round numbers $5 billion so it would be

economically rational now that Google's

paying if it's 30 billion we're paying

them 15 Google can go to every person

who works in Broadcom Smi double their

comp

>> and make an extra extra 5 billion. You

know, in 2028, let's just say it does 50

billion. Now it's 25 billion. You could

triple their comp. And by the way, you

don't need them all.

>> Yeah.

>> And and of course, they're not going to

do that because of competitive concerns.

>> But with TPUv8,

all of this and V9, all of this is

beginning to have an impact because

Google is bringing in MediaTek. This is

maybe the first way you send a warning

shot to Broadcom. were really not happy

about

>> all this money we're paying

>> but they did bring MediaTek in and the

Taiwanese ASIC companies have much lower

gross margins so this is kind of the

first shot against the bow and then

there's all this stuff people say oh but

has the best certiscom

has really good certis and certis is

like an extremely foundational

technology because it's how the chips

communicate with each other you have to

serialize and do serialize but there are

other good certis providers in the world

a really good certis is not at a certain

Maybe it's worth 10 or 15 billion a

year, but it's probably worth about

worth 25 billion a year. So because of

that friction, um, and I think

conservative design choices on the part

of Google and maybe the reason they made

those conservative design choices is

because they were going to a bifurcated

supply. You know, TPU is slowing down. I

would say has kind of the GPUs are

accelerating. This is the first, you

know, the com the competitive response

of Lisa and Jensen to everybody saying

we're gonna have our own ASIC is, hey,

we're just going to accelerate. We're

going to do do a GPU every year and you

cannot keep up with us. And then I think

what everybody is learning is like, oh

wow, that's so cool. You made your own

accelerator has an ASIC. Wow, what's the

nick going to be? What's the CPU going

to be? You know, what's the scaleup

switch going to be? What's the scaleup

protocol? What's the scale out switch?

what kind of optics are you going to

use? What's the software that's going to

make all this work together? And then

it's like, oh I made this tiny

little chip and you know, like whether

it's admitted or not, like you know, I'm

sure the GPUs don't GPU makers don't

love it when their customers make AS6

and try and compete with them

>> and like whoops

what what did I do? I thought this was

easy. How do you know? And it also it

takes at least three generations to make

a good chip like the TP TPU V1. I mean

it was an achievement and that they made

it.

>> Yeah.

>> Um it was really not till TPU V3 or V4

that the TPU started to become like even

vaguely competitive.

>> Is that just a classic like learning by

doing thing

>> 100%.

>> Yeah. And even if you've made like the

first from my perspective, the best ASIC

team at any semiconductor company is

actually the Amazon ASIC team.

>> You know, they were the first one to

make the gravitron CPU. They have this

nitro. Um it was the first, it's called

Supernick. They've been extremely

innovative, really clever. And like

Tranium and Infantry One, you know, they

maybe they're a little better than the

TPUV1, but only a little. Trannium 2,

you get a little better. Trium 3 it's I

think the first time it's like okay and

then you know I think tradeium 4 will

probably be good. I will be surprised if

there are a lot of AS6 other than

tranium and TPU

>> and by the way and tranium and TPU will

both run on customerowned tooling at

some point. We can debate when that will

happen but the economics of success that

I just described mean it's inevitable.

Like no matter what the companies say,

just the economics make it and reasoning

from first principles make it absolutely

inevitable.

>> If I were to zoom all the way out on

this stuff, because sometimes I just

it's I I find these details unbelievably

interesting and it's like the grandest

game that's ever been.

>> That's what I mean. It's crazy.

>> It's so crazy and so fun to follow.

Sometimes I forget to zoom out and say,

"Well, well, so what?" Like, okay, so

project this forward three generations

past Reuben or whatever. What what is

like the global human dividend of all

this crazy development? Like we keep

making the loss lower on these, you

know, pre uh pre pre-training scaling

models like who cares? Like it's been a

while since I've asked this thing

something that I wasn't kind of blown

away by the answer for me personally.

What are the next couple of things that

all this crazy infrastructure war allows

us to unlock because they're so

successful? If I were to posit like an

event path, I think the Blackwell models

are going to be amazing. The dramatic

reduction in per token cost enabled by

the GB300 and probably more the MI450

than the MI355, you know, will lead to

these models being allowed to think for

much longer, which means they're going

to be able to, you know, do new things.

Like I was very impressed Gemini 3 made

me a restaurant reservation.

>> It's the first time it's done something

for me. And I mean, other than like go

research something and teach me stuff,

>> but you know, if you can make a

restaurant reservation, you're not that

far from being able to make a hotel

reservation and an airplane reservation

and order me an Uber and

>> all of a sudden you got an assistant.

>> Yeah. And you can just imagine,

everybody talks about that, but you can

just imagine it's on your phone. I think

that's that's pretty near-term, but you

know, it's you know, it's some big

companies that are very tech forward.

you know 50% plus of customer support is

already done by AI and that's a $400

billion dollar industry and then if you

know what AI is great about is

persuasion that's sales and customer

support

>> and so of the functions of a company if

you think about them them they're to

make stuff sell stuff and then support

the customers so right now maybe you're

in late 26 you're going to be pretty

good at two of them um I do think it's

going to have a big impact on media like

I think robotics you know we talked

about the last time are going to finally

start to be real. You know, there's an

explosion in kind of exciting robotic

robotic startups. I do still think that

the main battle is going to be between

uh Tesla's Optimus and the Chinese

because, you know, it's easy to make

prototypes. It's hard to massproduce

them. But then it goes back to that what

Andre Karpathy said about AI can

automate anything that can be verified.

So any function where there's a right or

wrong answer, a right or wrong outcome,

you can apply reinforcement learning and

make the AI really good at that. Yeah.

>> What are your favorite examples of that

so far or theoretically?

>> Does the model balance? They'll be

really good at making models. Does you

know do all the books globally

reconcile? They'll be really good at

accounting because it you know was you

know double entry bookkeeping. It has to

balance. There's a verifiable you got it

right or wrong supporter sale. Did you

make the sale or not? That's very clear.

I mean that that's that's just like

AlphaGo.

>> You know, did you win or you lose? Did

the guy convert or not? Did the customer

ask for an escalation during customer

support or not?

>> It's like it's most important functions

are important because they can be

verified,

>> right? So I think if all of this starts

to happen and starts to happen in in in

26

like there'll be an ROI on Blackwell and

then all this will continue

>> and then we'll have Reuben and then

that'll be another big quantum of spin

Reuben and the MI450 and the TPU V9 and

then I do think just the most

interesting question is what are the

economic returns to artificial super

intelligence because all of these

companies in this great game they've

been in a prison prisoners dilemma.

They're terrified that if they slow

down,

>> they're just gone forever.

>> And their competitors don't, it's an

existential risk. And you know,

Microsoft blinked for like six weeks

earlier this year.

>> Yeah.

>> And like I I think they would say they

regret that.

>> Yeah.

>> But with Blackwell and for sure with

Reuben, economics are going to dominate

the prisoners dilemma from a decision-m

and spending perspective just because

the numbers are so big. And this goes to

kind of the ROI on AI question. And the

ROI on AI has

empirically, factually, unambiguously

been positive.

>> Like I just always find it strange that

there's any debate about this because

the largest biders on GPUs are public

companies. They report something called

audited quarterly financials. And you

can use those things to calculate

something called a return on invested

capital. And if you do that calculation,

the ROIC of the big public spenders on

GPUs is higher than it was before they

ramp spinning. And you could say, well,

part of that is, you know, opex savings.

Well, at some level that is part of what

you expect the ROI to be from AI. And

then you say, well, a lot of is actually

just applying GPUs, moving the big

recommener systems that power the

advertising and the recommendation

systems from CPUs to GPUs, and you've

had massive efficiency gains. And that's

why all the revenue growth at these

companies has accelerated. But like, so

what? the ROI has been there. Um, and it

is interesting like every big internet

company,

>> the people who are responsible for the

revenue

>> are intensely annoyed at the amount of

GPUs that are being given to the

researchers.

>> It's a very linear equation. If you give

me more GPUs, I will drive more revenue.

Give me those GPUs, we'll have more

revenue, more gross profit, and then we

can spend money. So, it's this constant

fight at every company. One of the

factors in the prisoners dilemma is

everybody has this like religious belief

that we're going to get to ASI and at

the end of the day what do they all

want? Almost all of them want to live

forever. Okay. And they think that ASI

is going to help them that

>> right good return.

>> That's a good return. But we don't know.

And if as humans we have pushed the

boundaries

of physics, biology and chemistry, the

natural laws that govern the universe.

I'm very curious about your favorite

sort of throw cold water on this stuff

type takes that you think about

sometimes. One would be like the things

that would cause I'm curious what you

think the things that would cause this

demand for compute to change or even the

trajectory of it to change. There's

there's one really obvious bare case and

it is just edge AI and it's connected to

the economic returns to ASI. in three

years on a bigger and bulkier phone to

fit the amount of DRAM necessary, you

know, and the battery won't probably

last as long, you will be able to

probably run a pruned down version of

something like Gemini 5 or Gro 4, Gro

4.1 or you know, Chat GPT at um

I don't know 30 60 tokens per second

>> and then that's free. And this is

clearly Apple's strategy. It's just

we're going to be a distributor of AI

>> and we're going to make it privacy safe

and run on the phone and then you can

call one of the big models, you know,

the the god models in the cloud whenever

you whatever you have a question. And if

that happens, if like 30 to 60 tokens at

like a one whatever it is a 115 30 60

tokens a second at a 115 IQ is good

enough.

I think that's

>> a bare case

>> other than just the scaling laws break,

you know. But in terms of if we assume

scaling laws continue and we now know

they're going to continue for

pre-training for at least one more

generation and we're very early in the

two new scaling laws you know for post-

training mid-training RLVR whatever

people want to call it and then test

time computed inference we're so early

in those and we're getting so much

better at helping the models hold more

and more context in the in their minds

as they do you know this test time

compute and that's really powerful

because you know everybody's like well

you know the how's the model going to

know this? Well, eventually if you can

hold enough context, you can just hold

every Slack message and Outlook message

and company manual in in a company in

your context.

>> Yeah.

>> And then you can compute the new task

>> and compare it with your knowledge of

the world, what you think, what the

model thinks, all this context. And you

know, it may be that like, you know, in

just really really long context windows

are the solution to a lot of the current

limitations. Um, and that's enabled by

some all these cool tricks like KV cash

offload and stuff. But I do think like

other than scaling laws slowing down,

other than there being low economic

returns to ASI, edge AI is to me by far

the most plausible and scariest bare

case.

>> I like to visualize like different

S-curves. invested through the iPhone

and I love to like see the visual of the

iPhone models as it as it sort of went

from this clunky bricky thing up to the

what we have now where like each one's

like a little bit, you know, obviously

we we've sort of petered out on its form

factor. If you if you picture something

similar for the Frontier models

themselves, does it feel like a like

it's at a certain part of that of that

natural technology paradigm progression

to you? If you're paying for Gemini

Ultra or Super Grock and you're getting,

you know, the good AI, it's hard to see

differences. Like, I have to go really

deep on something like, do you think PCI

Express or Ethernet is a better protocol

for scale up networking and why? Show me

the scientific papers. And if you shift

between models and you ask a question

like that where you know it really

deeply,

>> know that then

>> you know the answers. Yeah. then you see

differences. I do play fantasy football.

Um, winnings are donated to charity,

>> but it is like, you know, these new

models are quite a bit better at helping

like who should I play?

>> Yeah.

>> You know, and and they they think in

much more sophisticated ways.

>> Um, and by the way, if you're if you're

a historically good fantasy football

player and you're having a bad season,

it's why this is why because you're not

using it, you know. And I think we'll

see that in more and more domains. But I

do think they are already at a level

where unless you are a true expert or

just have an intellect that is beyond

mind

>> um it's hard to see um the progress and

that's why I do think we need to shift

from getting more intelligent to more

useful

>> unless more intelligence starts leading

to these massive scientific

breakthroughs and we're curing cancer in

26 and 27. Yeah,

>> I don't know that we're going to be

curing cancer, but I do think from a ROI

almost an ROIS curve, we need to kind of

hand off from intelligence to usefulness

>> and then usefulness will then have to

hand off to scientific breakthrough, you

know, just that creates whole new

industries.

>> What are the building blocks of

usefulness in your mind?

>> Just being able to do things

consistently and reliably. And a lot of

that is keeping all the context. Like

there's a lot of context if someone

wants to plan a trip for me. Like you

know I've I've acquired these strange

preferences. Like I follow that guy

Andrew Huberman. So I like to have an

east facing balcony so I can get morning

sun. You know the AI has to remember,

you know, being on a plane with Starlink

is important to me. Okay. Here are the

resorts I've historically liked. Here

are the kinds of areas I've liked. Here

are the rooms that I would really like

at each. That's a lot of context and to

keep all of that and kind of weight

those it's a hard problem. So I think

context windows are a big part of it.

You know, there's this meter task

evaluation thing like

>> how long it can work,

>> how long it can work for. And you could

think of that being related and in some

way to to context. Um although not

precisely, but that just task length

needs to keep expanding because you know

booking a restaurant and booking is

economically useful but you know it's

not that economically useful. But

booking me an entire vacation and

knowing the preferences of my parents,

my sister, my niece, and my nephew, and

what it means that like that's a much

harder problem. And that's something

that like a human might spend three or

four hours on optimizing that. And then

if you can do that, that's that's

amazing. But then again, I just think it

it has to be good at sales and customer

support relatively soon. I do think

we're going to see an kind of an

acceleration in the awesomeness of

various products just because engineers

are using AI to make products better and

faster.

>> We both invested in Forell, the hearing

aid company, which is just absolutely

remarkable. I think something I never

would have thought of.

>> And we're going to see I think something

like that in every vertical and that's

AI being used for the most core function

>> Yeah. of any company which is designing

the product and then it will be you know

there's already lots of examples of AI

being used to help manufacture the

product and distribute it more

efficiently you know whether it's

optimizing a supply chain you know

whether it's you know having a vision

system watch a production line you know

so I I think a lot of stuff is happening

the other thing I think is really

interesting in this whole ROI part is

Fortune 500 companies are always the

last to adopt a new technology they're

conservative they have lots of

regulations lots of lawyers Startups are

always the first. So let's think about

the cloud which was the first which was

the last like really truly kind of

transformative new technology for

enterprises. Being able to you know do

have all of your um compute and the

cloud and use SAS. So it's always

upgraded. It's always great etc. etc.

You can get it on every device. I mean

it's those were dark days before the

cloud. You know

>> the first AWS reinvent I think it was in

2013. Every startup on planet Earth ran

on the cloud.

>> Yeah. The idea that you would buy your

own server and storage box and router

was ridiculous. And that probably

happened like even earlier that that had

probably already happened before the

first reinvent. And then like you know

the first big Fortune 500 companies

started to standardize on it like maybe

5 years later. You see that with AI I'm

sure you've seen this in your startups

and I think one reason VCs are more

broadly bullish on AI than public market

investors is VCs see very real

productivity gains. There's all these

charts that for a given level of

revenue, a company today has

significantly lower employees than a

company of two years ago.

>> And the reason is AI is doing a lot of

the sales, the support and helping to

make the product. And I mean there is,

you know, Iconic has some charts. A6Z,

by the way, David George is a good

friend, great guy. You know, he has his

model busters thing. So there's very

clear data that this is happening. So

people who have a lens into the world of

venture see this. And I do think it was

very important in the third quarter,

this is the first quarter where we had

Fortune 500 companies outside of the

tech industry give specific quantitative

examples of AIdriven uplift. So C

Robinson went up something like 20% on

earnings. And should I tell people what

C Robinson does?

>> Let's just say a truck goes from, you

know, Chicago to Denver. And then, you

know, the trucker lives in Chicago. So

it's going to go back from Denver to

Chicago. There's an empty load. And CH

Robinson has all these relationships

with these truckers and trucking

companies and they match shippers demand

with that empty load supply to make the

trucking more efficient. You know,

they're a freight forwarder. You know,

there's there's there's actually lots of

companies like this, but kind of they're

the biggest and most dominant. So, one

of the most important things they do is

they quote price and availability. So,

somebody a customer calls them up and

says, "Hey, I urgently need three

18-wheelers from Chicago to Denver." You

know, in the past they said it would

take them, you know, 15 to 45 minutes

and they only quoted 60%

of inbound requests. With AI, they're

quoting 100% and doing it in seconds.

>> And so they printed a great quarter and

the stock went up 20% and it was because

of AIdriven productivity that's

impacting the revenue line, the cost

line, everything. And so I actually

think that's pretty important because I

was I was actually very worried about

like the idea that we might have this

Blackwell ROI air gap because we're

spending so much money on Blackwell.

Those Blackwells are being used for

training and there's no ROI on training.

Training is you're making the model. The

ROI comes from inference. So I was

really worried that you know we're going

to have

>> you know maybe this threequarter period

where the capex is unimaginably high.

>> Those black wheels are only being used

for training bars staying flat eyes

going up.

>> Yeah. Yeah. Exactly. And so ROIC goes

down and you can see like Meta Meta they

printed you know because Meta has not

been able to make a frontier model. Meta

printed you know a quarter where ROIC

declined and that was not good for the

stock. So I was wor really worried about

this. I do think that those data points

are important in terms of suggesting

that maybe we'll be able to navigate

this potential air gap and ROIC.

>> Yeah it makes me wonder about in this

market I'm like everybody else. It's the

10 companies at the top that are all the

market cap more than all of the

attention. There's 490 other companies

500. You studied those too. Like what

what do you think about that group? Like

what what is interesting to you about

the group that now nobody seems to talk

about and no one really seems to care

about because they don't they haven't

driven returns and they're a smaller

percent of the overall.

>> Well, I think that people are going to

start to care if you have more and more

companies print these CH Robinson like

quarters that companies that have

historically been really wellrun.

The reason like they have a long track

record of success is they have a long

you cannot succeed without using

technology well and so if you have a

kind of internal culture of

experimentation and innovation I think

you will do well with AI

>> you know so like I would bet on the best

investment banks to be early

you know earlier and better adopters of

AI than maybe some of the trailing banks

you know just sometimes

past prologue one thing that I strong

opinion I

you know, all these VCs are setting up

these holding companies and, you know,

we're going to use AI to make

traditional businesses better and, you

know, they're really smart VCs and

they're great track records. But that's

what private equity has been doing for

50 years.

>> You're just not going to beat private

equity at their game.

>> What Vista did in the early days, right?

>> Yeah. Private equity's maybe had a

little bit of a tough run. You know,

just multiples have gone up. Now,

private assets are more expensive. The

cost of financing has gone up. It's

tough to take a company public because

the public valuation is 30% lower than

the private valuation. So PE's had a

tough run. I actually think these

private equity firms are going to be

pretty good at systematically applying

AI. We haven't spent much time talking

about meta, anthropic or open AI. And

I'd love just like your impression on

everything that's going on in this

infrastructure side that we talked

about. These are three really important

players in this in this grand battle,

this grand this grand game. How does all

of this development that we've discussed

so far impact those players specifically

do you think? The first thing let me

just say about frontier models broadly.

>> Yeah.

>> You know in in 2023 and 24 I was fond of

quoting Eric Visria and Eric Fishria's

statement our friend um brilliant man

and Eric would always say foundation

models are the fastest appreciating

assets in history.

>> And I would say he was 90% right. I

modified the statement. I said

foundation models without unique data

and internet scale distribution are the

fastest appreciating assets in history.

But reasoning fundamentally changed that

in a really profound way. There was a

loop, a flywheel to quote Jeff Bezos

that it was at the heart of every great

internet company and it was you made a

good product, you got users, those users

using the product generated data that

could be fed back into the product to

make it better. And that flywheel has

been spinning at Netflix, at Amazon, at

Meta, at Google, you know, for over a

decade. And that's an incredibly

powerful flywheel. And it's why those

internet businesses were so tough to

compete with. It's why they were

increasing returns to scale. You

everybody talks about network effects

much more and you know network effects

are they were important for social

networks. I I don't know to what extent

meta is a social network anymore. It's

more like a content distribution

>> but they just had increasing returns to

scale because of that

>> flywheel. And that dynamic was not

present in the pre-reasoning world of

AI. You pre-trained a model, you let it

out in the world, and it was what it

was. And it was actually pretty hard.

They would do RLHF, reinforcement

learning with human feedback, and you

try and make the bot model better, and

maybe you'd get a sense from Twitter

vibes that people didn't like this, and

so you tweak it, and you know, there

were the little up and down arrows, but

it was actually pretty hard to feed that

back into the model. with reasoning.

It's early but that flywheel has started

to spin and that is really profound for

these frontier labs. So one reasoning

fundamentally changed the industry

dynamics of Frontier Labs. just explain

why specifically that is like what what

what is going on

>> because if a lot of people are asking a

similar question and

they're consistently either liking or

not liking the answer, then you can kind

of use that like that as a verifiable

reward. That's a good outcome. And then

you can kind of use feed those good

answers back into the model. and we're

very early at this flywheel spinning

>> like it's hard to do now,

>> but you can see it beginning to spin.

>> So, this is important fact number one

for all of those dynamics. Second,

>> I think it's really important that Meta,

you know, Mark Zuckerberg at the

beginning of this year in January said,

you know, I anticipate, you know, I'm

highly confident, I'm going to get the

quote wrong, that at some point in 2025,

we're going to have the best and most

performant AI. I don't know if he's in

the top hundred. Okay,

>> so he was as wrong as it was possible to

be. And I think that is a really

important fact because it suggests that

what these four companies have done is

really hard to do because Meta threw a

lot of money at it and they failed.

Yamakun had to leave. They had to have

the famous billion dollar for AI

researchers. And by the way, Microsoft

also failed. They did not make such an

unequivocal prediction but they hire but

they bought um inflection AI and you

know there were a lot of comments from

them that we anticipate our internal

models quickly getting better and we're

going to run more and more of cop you

know our AI on our internal models

>> Amazon they bought a company called

Adept AI

>> they have their models called Nova

>> no I don't think they're in the top 20

>> so clearly it's much harder to do than

people thought a year ago and there's

many many reasons for that like it's

actually really hard to keep a big

cluster of GPUs coherent. A lot of these

companies were used to running their

infrastructure to optimize for cost

>> right

>> instead of performance

>> complexity and performance

>> complexity and keeping the GPUs

running at high utilization rate in a

big cluster. It's actually really hard

and there are wild variations in how

well companies run GPUs.

>> Yeah. And if you're running if the most

anybody because the laws of physics, you

know, maybe you can get two or 30

hundred,000 black wells coherent, we'll

see. But if you have 30% uptime on that

cluster and you're competing with

somebody who has 90% uptime,

>> you're not even competing. So one,

there's a huge spectrum in how well

people run GPUs. Two, then I think there

is, you know, these AI researchers, they

like to talk about taste. I find it very

funny. You know, oh why do you make so

much money? I have very good taste. You

know what taste means is you have a good

intuitive sense for the experiments to

perform. And this is a this is is why

you pay people a lot of money because it

actually turns out that as these models

get bigger, you can no longer run an

experiment on a thousand GPU cluster and

replicate it on 100,000 GPUs. You need

to run that experiment on 50,000 GPUs

and maybe it takes, you know, days.

>> And so there's a very high opportunity

cost. And so you have to have a really

good team that can make the right

decisions about which experiments to run

on this. And then you need to do, you

know, all the reinforcement learning

during post- training well and the test

time compute. Well, complicated.

>> It's really hard to do. And everybody

thinks it's easy, but all those things,

you know, I used to have this saying

like, hey, I was a retail analyst long

ago. Pick any vertical in America. If

you can just run a thousand stores and

have them clean, well lit, stocked with

relevant goods at good prices and

staffed by friendly employees who are

not stealing from you, you're going to

be a $20 billion company, a $30 billion

company. Like 15 companies have been

able to do that. It's really hard. And

it's the same thing. Doing all of these

things well is really hard.

and then reasoning with this flywheel.

This is beginning to create more

separation.

>> And what's even more important, every

one of those labs, XAI, Gemini, OpenAI,

and Enthropic, they have a more advanced

checkpoint

internally of the model. Checkpoint is

just um you're kind of continuously

working on these models and then you

release kind of a checkpoint and then

the reason these models get fast

>> the one they're using internally is for

>> better and they're using that model to

train the next model

>> and if you do not have

>> that latest checkpoint it's

>> you're behind

>> you're it's getting really hard to catch

up. Chinese open source is a gift from

God to meta

>> because you can use Chinese open source

>> to try and that can be your checkpoint

and you can use that

>> as a way to kind of bootstrap this and

that's what I'm sure they're trying to

do and everybody else. Um the big

problem and the big a giant swing factor

I think China's made a terrible mistake

with this rarest thing you know I think

China because you know they have the

Huawei Asin and it's a decent chip and

verse something you know like you know

the the deprecated hop preserving

something it looks okay so they're

trying to force Chinese open source to

use their Chinese chips uh their

domestically designed chips. The problem

is Blackwell is going to come out now

and the gap between these American

frontier labs and Chinese open source is

going to blow out because of Blackwell

and actually DeepSeek in their most

recent technical paper v3.2 said like

one of the reasons we struggle to

compete with the American Frontier Labs

is we don't have enough compute. That

was their very politically correct,

still a little bit risky way of saying,

you know, cuz China said, "We don't want

the black wells, right?" And they're

saying, "Guys, that might be a big

mistake. That might be a big mistake."

And so, if you just kind of play this

out, these four American labs are going

to start to widen their gap versus

Chinese open source, which then makes it

harder for anyone else to catch up

because that gap is growing. So, you

can't use Chinese open source to

bootstrap. And then geopolitically,

China thought they had the leverage.

They're going to realize, oh, whoopsy

daisy. We do need the black wells. And

unfortunately, they'll probably for them

um they'll probably realize that in late

26. And at that point, there's an

enormous effort underway. DARPA has

there's all sorts of really cool DARPA

and DoD programs to incentivize really

clever technological solutions for rare

earths, you know, like using enzymes to

refine them or there's all sorts of

really cool things happening, you know,

and then, you know, there's a lot of

rare earth deposits in countries that

are very friendly to America that, you

know, don't mind actually refining it in

the, you know, traditional way. So, I

think rare earths are going to be solved

way faster than anyone thinks. You know,

they're obviously not that rare. They're

just misnamed. they're rare because, you

know, they're really messy to refine.

And so geopolitically, I actually think

Blackwell is pretty significant. Um, and

it's going to give America a lot of

leverage as this gap widens. And then in

the context of all of that, going back

to the dynamics between these companies,

XAI will be out with the first Blackwell

model and then they'll be the first ones

probably using Blackwell for inference

at scale. And I think that's an

important moment for them. And by the

way, it is funny like um you know if you

go on open router you can just look they

have dominant share now open router is

whatever it is it's 1% of of API tokens

but it's an indication

>> they process 1.35 trillion tokens Google

did like eight or 900 billion this is

like whatever it is last 7 days or last

month you know anthropic was at 700

billion like XAI is doing really really

well and the model is fantastic I highly

recommend it but you'll see XAI you know

come out with this open AAI will come

want faster. OpenAI's

issue that they're trying to solve with

Stargate is because they pay a margin to

people for compute

>> and maybe the people who run their

compute are not the best at running

GPUs. They are a high-cost producer of

tokens. Um, and I think this kind of

explains a lot of their

>> code red recently.

>> Yeah. Well, just the 1.4 $4 trillion in

spending commitments. And I think that

was just like, hey, they know they're

going to need to raise a lot of money.

Um, particularly if Google keeps its

current strategy of sucking the economic

oxygen out of the room and, you know,

you go from 1.4 trillion rough vibes

code red like pretty fast, you know, and

the reason they have a code red is

because of all these dynamics. So then

they'll come out with a model but they

will not have fixed their per token cost

disadvantage yet relative to both XAI

and Google and almost and anthropic at

that point. Anthropic is a good company.

You know they're burning dramatically

less cash than openai and growing

faster. So I think you have to give

anthropic a lot of credit and and a lot

of that is their relationship with

Google and Amazon for the TPUs and the

trainiums. And so Anthropic has been

able to benefit from the same dynamics

that Google has. I think is very

indicative in this great game of chess.

You know, you can look at Daario Jensen

maybe have taken a few there have been a

few public comments, you know, that

were, you know, made between them.

>> Jousting,

>> a little bit of jousting. Well,

Anthropic just signed the $5 billion

deal with Nvidia.

>> That is because Daario is a smart man

and he understands these dynamics about

Blackwell and Rubid relative to TPU. And

so Nvidia now goes from having two of

the fighters,

two fighters, XAI and OpenAI to three

fighters. So that that helps in this

Nvidia vers Google battle. And then if

Meta can catch up, that's really

important. And so I'm I am sure Nvidia

is doing whatever they can to help Meta,

you know, whatever. Like let us you're

running those GPUs this way. May maybe

we should maybe we should twist the

screw this way or turn the dial that way

and then it will be also if Blackwell

comes back to China which it seems like

it probably will happen that will also

be very good because then Chinese open

source will be back. What other I'm I'm

always so curious about the polls of

things like one poll would be the other

breakthroughs that you have your your

mind on things in the data center that

aren't chips that we've talked about

before as as one example. I think the

most important thing that's going to

happen in the world in this world in the

next 3 to four years is data centers in

space

>> and this has really profound

implications for everyone building a

power plant or a data center on planet

earth. Okay. And there is a giant gold

rush into this.

>> I haven't heard anything about this so

please.

>> Yeah. You know it's like everybody

thinks like hey AI is risky you know uh

but you know what I'm going to build a

data center. I'm going to build a power

plant that's going to do a data center.

We will need that. But if you think

about it from first principles, data

centers should be in space. Okay.

What are the fundamental inputs to

running a data center? There are power

and there are cooling

>> and then there are the chips.

>> That's like the total if you think about

it from a total cost perspective.

>> Yeah. And just the the inputs to making

the tokens come out of the magic

machines.

>> Yeah.

So in space you can keep a satellite in

the sun 24 hours a day

>> and the sun is 30% more intense. You

know you can keep it in the sun just

because like if the sun's here's this

you know you can have the satellite you

know always kind of catching

>> catching the light

>> catching the light. The sun is 30% more

intense and this results in six times

more irradiance in outer space than the

high than on planet earth. So you're

getting a lot of solar energy. Point

number one. Point number two, because

you're in the sun 24 hours a day, you

don't need a battery. And this is a

giant percentage of the cost. So the

lowest cost energy um available in our

solar system is solar energy and space.

Okay. Second, for cooling in one of

these racks, a majority of the mass and

the weight is cooling.

>> And the cooling in these data centers is

incredibly complicated. You know, I

mean, the HVAC, the CDUs, the liquid

cooling.

In space, cooling is free. You just put

a radiator on the dark side of the

satellite.

>> It's gold.

>> And it's as close to absolute zero as

you can get.

>> So, all that goes away and that is a

vast amount of cost. Okay, let's think

about um how this these, you know, maybe

each satellite is kind of a rack. It's

one way to think of it. Maybe some

people make bigger satellites that are

three racks. Well, how are you going to

collect connect those racks? Well, it's

funny. In the data center, the racks are

over a certain distance um connected

with fiber optics. And that just means a

laser going through a cable. The only

thing faster than a laser going through

a fiber optic cable is a laser going

through absolute vacuum. So, if you can

link these satellites in space together

using lasers, you actually have a faster

and more coherent network than in a data

center on Earth. Okay, for training

that's going to take a long time

>> just because it's so big.

>> Yeah, just because it's so big. But for

inference, but I think even training

will eventually happen. But then for

inference, let's think about the user

experience when I when we asked when you

know when I asked Gro about you and it

gave the nice answer. A radio wave

traveled from my cell phone to a cell

tower. Then it hit the base station,

went into a fiber optic cable, went to

some sort of metro aggregation facility

in New York, probably within like, you

know, 10 blocks of here. There's a small

little metro router that's routed those

packets to a big XAI data center

somewhere. Okay? And then the

computation was done and it came back

over the same path.

If the satellites can communicate

directly with the phone and Starlink has

demonstrated directto cell capability,

you just go boom boom. It's a much

better lowerc cost user experience. So

in every way data centers in space from

a first principles perspective are

superior to data centers on earth.

>> So if we could teleport that into

existence, I understand that that

portion. What are the frictions to that?

H like why will that not happen? And is

it launch cost? Is it launch

availability?

>> I mean, we need a lot of the space

starships. Like the Starships are the

only ones that can eomically make that

happen.

>> We need a lot of those Starships. Um,

you know, maybe China or Russia will be

able to land a rocket. Blue Origin just

landed a booster. It's an entirely new

and different way to think about SpaceX.

And it is interesting that you know Elon

posted yesterday or said in an interview

>> that Tesla, SpaceX and XAI kind of

converging

>> were converging and they really are. So

XAI will be the intelligence module for

Optimus made by Tesla with Tesla vision

has its you know perception system and

then you know SpaceX will have the data

centers in space that will will you know

power a lot of the AI presumably for XAI

and Tesla and the Octopuses and a lot of

other companies and it's just it is just

interesting the way that they're

converging and each one is kind of

creating competitive advantage for the

other you know so it's if If you're XAI,

it's really nice that you have this

built-in relationship with Optimus and

now, you know, Tesla's a public company,

so there's going to be like I cannot

imagine the level of vetting that will

go into that intercomp agreement, you

know, and then you have a big advantage

with these data centers in space. Um,

and then it's also nice if you're XAI

that you have two companies with a lot

of customers who you can use to help

build your customer support agents, your

customer sales agents with kind of

built-in customers. So, they really are

all kind of converging um in a neat way.

And I do think like it's going to be a

big moment when that first Blackwell

model comes out from XAI next year. Hm.

If I go to the other end of the spectrum

and I think about something that seems

to have been historically endemic to the

human economic experience that uh

shortages are always followed by gluts

in capital cycles. What if in this case

um the shortage is compute like Mark

Chen now is on the record as saying they

would consume 10x as much compute if you

gave it to them in like a couple weeks.

So so like there seems to still be a

massive shortage of compute which is all

the stuff we've talked about today. But

there also just seems to be this like

iron law of history that gluts follow

shortages. What do you think about like

that concept as it relates to this

>> technology be a glut?

>> Yeah.

>> You know, and AI is fundamentally

different than the software just in that

every time you use AI takes compute in a

way that traditional software just did

not. I mean it is true like I think

every one of these companies could

consume 10x more compute. Like what

would happen is just the $200 tier would

get a lot better. the free tier would

get like the $200 tier. Google has

started to monetize AI mode with ads

>> and I think that will give everyone else

permission to introduce ads into the

free mode and then that is going to be

an important source of ROI you know like

>> seems like OpenAI is tailor made to

>> Yeah. Absolutely. All of them and

actions like you know hey

>> you know here are your three vacations

would you like me to book one and then

they're for sure going to collect a

commission. Yeah.

>> You know here's you know there there

there's many ways you can make money. I

think we went into great detail on

maybe a prior podcast about how just

inventory dynamics made these inventory

cycles inevitable in semis. Um, and the

iron law of semis is just that customer

buffer buffer inventories have to equal

lead times. And that's why you got these

inventory cycles historically. We

haven't seen a true capacity cycle in

semis maybe arguably since the late 90s.

And that's because Taiwan Smi has been

so good at aggregating

and smoothing supply.

And a big problem in the world right now

is that Taiwan semi is not expanding

capacity as fast as their customers

want. And I think this is actually a

pretty big this they're they're in the

process of making a mistake just because

you know you do have Intel and with

these fabs and they're not as good and

it's really hard to work with their PDK

but now you have this guy Leapoo who's

who's a really good executive um and

really understands that business. I mean

by the way Patrick Elsinger I think was

was also a good executive and he put

Intel on the only strategy that could

result in su success and I actually

think it's shameful that the Intel board

fired him when they did it. But Leapoo

is a good executive and now he's reaping

the benefits of Patrick's strategy and

Intel has all these empty fabs and

eventually

given the shortages we have of compute

those fabs are going to be filled.

>> So I think Taiwan Sim is in the process

of making a mistake but they're just so

paranoid about an overbuilt. Yeah.

>> And they're so skeptical. You know

they're the guys who met with Sam Alman

and laughed and said he's a podcast bro.

He has no idea what he's talking about.

You know they're terrified of an

overbuild. So it may be that Taiwan

Simei

singlehandedly that they're cautious

>> the breaks on the bubble

>> is is is is the governor um and you know

and we do like I think you know I think

governors are good it's good that you

know it's good that power is a governor

it's good that Taiwan sim is a governor

if Taiwan semi opens up at the same time

when you know data centers in space

relieve all power constraints but that's

like I don't know five six years away

that data centers in space or majority

of deployed megawatt like yeah I think

you get it overbuild really fast but

just we have these two really powerful

natural governors

>> and I think that's good you know like

smoother and longer is good

>> we haven't talked about the power other

than alluding to it through the space

thing haven't talked about power very

much power was like the most

uninteresting topic because there's the

de demand and nothing really changed for

like a really really long time all of a

sudden we're trying to figure out how to

get like gigawatts here there and

everywhere how do you think about are

you interested in powers

>> I'm very interested I do feel lucky in a

prior life I was the sector leader for

the telecom and utilities team.

>> Okay,

>> I I do have some base level of

knowledge. So one, you know, having um

having watts as a constraint is like

really good for the most advanced

compute players because if watts are the

constraint,

>> the price you pay for compute is

irrelevant. The TCO of your compute is

absolutely irrelevant because if you

could get 3x or 4x or 5x more tokens per

watt, that is literally three or 4x or

5x more revenue.

And so, you know, it's just like if

you're going to build a like, okay, like

an advanced data center costs 50

billion. A data center with your ASIC

maybe costs 35 billion, but if that $50

billion revenue, if that $50 billion

data center pumps out 25 billion of

revenue and your ASIC data center at 35

billion is only pumping out 8 billion,

well, like you're, you know, you're

pretty bummed. It's good for like all of

the most advanced

technologies in the data center which is

exciting to me as an investor. So as

long as power is a governor the best

products are going to win irrespective

of price and have crazy pricing power.

Okay, I think that's that's the first

implication that's really important to

me. Second, it is in the only solutions

to this. We just can't build nuclear

fast enough in America. Like as much as

we would love to build nuclear quickly,

we just can't. We just can't. Yeah,

>> it's just too hard, you know. Um, NEPA,

all these rules, like it's just it's too

hard. Like a a rare ant that we could

move and it could be in a better

environment can totally delay the

construction of a nuclear power plant.

You know, one ant. It's crazy actually.

Um, like humans need to come first. We

need to have a humanentric view of the

world. But like the solutions are

natural gas and solar. And the great

thing is the great thing about these AI

data centers is apart from the ones that

you're going to do inference on, you can

locate them anywhere. So I think you

were going to see and you're this is why

you're seeing all this activity in

Abalene, you know, because it's in the

middle of a big natural gas basin and we

have a lot of natural gas in America

because of fracking. You I think we

we're going to have a lot of natural gas

for a long time. We ramp production

really fast. So I think this is going to

be solved. You know, you're going to

have power plants fed by gas or solar. I

think that's the solution. And you know,

you're already, you know, all these

turbine manufacturers were reluctant to

expand capacity. Caterpillar just said,

"We're going to increase capacity by 75%

over the next few years." So like the

system on the power side is beginning to

respond. One of the reasons that I

always so love talking to you is that

you do every like you do as much in the

top 10 companies in the world as you do

looking at brand new companies with, you

know, entrepreneurs that are 25 years

old trying to do something amazing. And

so you have this very broad sense of

what's going on. If I think about that

second category of young enterprising

technologists who now are like AI,

they're like kind of the first

generation of AI native entrepreneurs.

What are you seeing in that group that's

notable or surprising or interesting?

>> These young CEOs, they're just so

impressive in all ways and they get more

polished faster. And I think the reason

is is they're talking to the AI.

>> How should I deal with pitching this

investor? I'm meeting with Patrick

Oanaughy. What What do you think the

best ways I should pitch him are?

>> Yeah. And it works.

>> Do a deep research. And it's good. You

know, hey, I have this difficult HR

situation.

>> How would you handle it?

>> That's correct.

>> And it's good at that. How would you,

you know, we're struggling to sell our

product. What changes would you make?

And it's really good at all of that

today.

And so, and that goes to these, you

know, VCs are seeing massive AI

productivity in all their companies.

It's because their companies are full of

these, you know, 23, 24 or, you know,

even younger AI natives. I've been so

impressed with like young investment

talent

>> and it's just part of it. Like your

podcast is part of that. There's just,

you know, knowledge and very very

specific knowledge has became so

accessible, you know, through podcasts

and the internet. Impressive young

people

>> come in and they're just I feel like

they're where I was as an investor like

in my, you know, early 30s and they're

22 and I'm like, "Oh my god,

>> like I have to run so fast to keep up."

these kids who are growing up native in

AI, they are just proficient with it um

in a way that I am trying really hard to

become.

>> Can we talk about semi VC specifically

and like what is interesting in that

universe?

>> Oh, just the one thing I just think that

I just think is so cool about it and so

underappreciated is your average

semiconductor venture founder is like 50

years old.

>> Okay. and Jensen and what's happened

with Nvidia and the market cap of Nvidia

has like singlehandedly

ignited semiconductor venture but the

way it's ignited it's ignited in an

awesome way that's like really good for

actually Nvidia and Google and everyone

>> is like let's just say you were the best

DSP architect in the world you had made

for the last 20 years every two years

because that's what you have to do

semiconductors it's like every two years

you have to win run a race

>> and if you won the last race you start

like a foot ahead

>> and over time those compound um and make

each race easier to win but like maybe

that person and his team maybe he's the

head of networking at a big public

company he's making a lot of money and

he has a good life and then because he

sees these outcomes and the size of the

markets in the data center he's like wow

why don't I just go start my own company

but the reason that's important is that

you know I forget the number but I mean

there are thousands of parts in a

blackwell rack and you know and there's

thousands of parts in a TPU rack And in

the Blackwell rack, you know, maybe

Nvidia makes,

I don't know, two two or 30 hundred of

those parts. And, you know, same thing

in an AMD rack. And they need all of

those other parts to accelerate with

them.

>> So, they couldn't go to this one-year

cadence if the rest everything was not

>> keeping up with them. The fact that

semiconductor venture venture has come

back with a vengeance, you know, Silicon

Valley stopped being Silicon Valley long

ago. My little firm maybe has done more

semiconductor deals in the last seven

years than the top 10 VCs combined, you

know, but that's really really important

because now you have an ecosystem of

companies who can keep up and then that

ecosystem of these venture companies is

putting pressure on the public companies

that are also need to part of part of

this if we're going to go to this annual

cadence which is just so hard. Um, and

it's one reason I'm really skeptical of

these AS6 that don't already have some

degree of success. So, I do think that's

a super super important dynamic and one

that's

absolutely foundational and necessary