Nvidia JUST GOT DESTROYED | WTF

Wow, there has been some major

artificial intelligence news with now

reports that Google is planning to sell

billions of dollars of chips. Yeah,

chips, tensor processing units, you

know, like their version of the GPU, but

Google's version and it's different.

Okay, we're not here to get into the

technicals on it. But now to Meta, which

is potentially going to lead to billions

of additional dollars to well, Google's

bottom line. On top of that, Nick T just

posted an article telling us about the

Fed's next rate cut. And Nick T forgot a

really important piece along with Nvidia

literally lashing out at Michael Bur.

You can't make this up. We have all

three of those to cover. So, let's knock

them out in order. Okay, so first things

first. Yes, Google is now in talks per

the information to sell TPUs to Nvidia

or sorry, to Meta, which is great. I

mean that's fantastic because what it

does is it creates competition in the

chip war and it gives Google another

stream of revenue. Uh in fact these TPUs

could be used in meta data centers as

soon as 2027.

Now Google is actually not just pitching

Meta but they're also trying to pitch

quote big financial institutions maybe

suddenly banks or you know who knows

what on using TPUs

instead of just relying on Nvidia chips.

Now this is a little bit of a middle

finger not just to Nvidia who makes 89%

of their revenue from data center sales.

It's also a middle finger to AMD where a

lot of people are like keeping their

fingers crossed that we're finally at

the AMD phase of the cycle.

And now Google is showing up to take

some of the cake. Well, this per the

information suggests that Google's cloud

unit could potentially see as much as

10% of Nvidia's market share going to

Google in terms of annual revenue. So,

if we assume Nvidia's data center sales

are going to be somewhere around $8

billion, and we assume Google's able to

produce these at a 70% gross margin, and

let's say they've already done the

research for the chips because they do

them in-house, so they don't really have

that much more in incremental operating

expenses. Let's assume they just pay

taxes on that. Works out to another

about 3% to EPS. Okay, that's really

quick math, but it works out to about 3%

in earnings per share. Now the stock

market hears that and the stock market

at least in after hours says loud and

clear Kevin stock up 3%. I kid you not

in after hours in the overnight market

right now. Google is literally up after

being up 6% on the day to $318

which is phenomenal already. Okay, we're

sitting at a an almost $4 trillion

market cap. We're basically going to be

over $4 trillion tomorrow at this rate

with a $3.44% 44% increase in the uh

overnight here. Now, so basically all of

that is already priced in, but what's

not priced in is the momentum cycle. And

see, that's what gets really fun here

because while Google is down 3.4%.

Nvidia, uh, sorry, Google is up 3.44%.

Too, sorry, it's late. [laughter] I've

been I've been up for way too long

working. Anyway, uh Nvidia is down 2% in

the after hours and AMD is down 1.77%.

If you look at the Q's were basically

flat, which is interesting because after

a day where we went lineto line right

here, 595 to 607, pretty remarkable call

in the alpha report. Here we are with

some major movements in the after hours

and the cues aren't moving at all. Now,

another reason that Nvidia could be

down, mind you. Quick finishing thought

there on Google. It's obviously bullish

for Google. I mean, congratulations to

people investing in Google or who have

Google shares or whatever. Like, this is

great. And Gemini is a great product. I

started, it's weird like on this

channel, if you go through back through

my channel, it was about 3 or 4 weeks

ago, I started saying, man, Gemini's

gotten really good. And that was based

on what my experience was with 2.5

because I usually I'll use multiple LLMs

just to kind of keep my finger on the

pulse of who's ahead. And so I'll use,

you know, Perplexity or GPT or Google

and I'll test the pro subs or whatever.

And I kind of put, you know, here, give

me your thoughts on this, give me your

thoughts on this, and then I'll take

like the thoughts of one and put them

into the thoughts of the other. I have a

lot of fun with it. And I'm like, damn,

you know, Gemini is actually getting

really good. It's it's slower, but it's

good. Uh, and dang, that would have been

the perfect time to go buy a Google

stock, wouldn't it been? [laughter] But

anyway, Nvidia is now lashing out at

Michael Bur. And this is really weird

because at the same time that Google is

starting to take a slice of the chip

market, even though that's for, you

know, data centers 2 years from now,

because keep in mind, you've got to

actually build these chips. Nvidia is

lashing out at Michael Bur, which is

really interesting because wait a

minute, why is Nvidia taking the time to

respond

to Michael Bur?

That's weird. Why are we stooping to

that level? Because some of the

allegations

are kind of not getting even great

answers here. You would think if Nvidia

is going to respond to one random

analyst on the street, you would think

they'd actually have good responses, but

I hate to say it, and I'm not trying to

be an Nvidia bear here, but the

responses aren't that good. Now, before

we go through some of the responses, I

want to show you a visual representation

of my thoughts about artificial

intelligence. So, I think that there are

two trains of thought here. Train of

thought number one is on the left. I

call it probably a fairy tale on the

left. So the fairy tale vision of

artificial intelligence is that we are

on the path to artificial general

intelligence and we're basically just

this straight line up right here. So

when you look at this exponential curve

right here, that's what we're on. And at

the top we're going to AGI. That's the

AI bull thesis. In my opinion, the

realistic AI thesis is that we're

actually on this scurve where we have

this exponential real step change up and

then what happens over time is you end

up getting to this logarithmic curve

out, this flattening out where progress

is approaching some form of a limit. So

the LLMs aren't getting that much

better. We're getting incrementally

better for the amount of money we're

spending, right? Remember back in the

days of math, you're approaching the

limit, right? you're basically

approaching this limit. And so the more

money you spend on compute, the slower

and slower your progress is. It's kind

of like full self-driving in Tesla when

I got it in 2017 was probably at the A

on this chart, like over here. I

couldn't technically use it on street

roads, but I did anyway. It just

wouldn't stop for red lights, so I'd

have to take over and stop for red

lights, you know. Then a few years later

when they actually introduced FSD

supervised, we're probably at the B

stage. And now we're probably at like

the C or even D stage of autopilot where

progress is happening so slowly that

it's almost indistinguishable

uh in my opinion at least where when I

go, man, did version 14 get better than

13 or did it get worse? Right? Like I'm

at that point where I'm like 13's

already really good. It's not 100% but

it's like 98% there. It's like did 14 go

to like 97.9 or did it go to 98.1? I I

don't know. I'll just say it went to

98.1 cuz I want to be bullish on it cuz

I can't wait for it to be at 100%. But

my point is more compute equals very

slow progress. So the question is where

is Google coming in at this race in 2027

when they actually potentially it's not

a for you know it's not guaranteed yet

that they're going to sell these chips

to Meta in 2027 are we going to be at C

or are we going to be at you know the

FSD D stage for example right like

usually where you want to get in on a

startup kind of like phase of these

cycles is you usually want to get in

like the a phase or even earlier than

that. Like think about it like this.

You know, I have a real estate startup.

You've heard me mention it before. We're

going to launch an AI product next month

in December. Uh we'll have more updates

on it probably this week, but we're

going to launch an AI product over here

at the ground floor. And my vision is

that that AI product is going to speed

us up to IPO so that knock on wood, no

guarantees, I could exceed people's

expectations because I've always said I

want to IPO between 28 and 2030 absent

some kind of big recession, right?

Hopefully I could exceed people's

expectations and we could IPO maybe in

27, maybe even at the beginning of 27,

right? And so what I got to do is I got

to boom that AI. I got to make sure that

AI crushes it. at real estate netw worth

boosting AI and the time you want to get

in on something like that is usually the

very earliest days. The problem is

that's when typically AI startups have

the biggest risk. Now I think that's

where we're unique cuz we don't have any

bank debt and we're backed by real

estate. You know if our AI doesn't work

out like there's still residual value of

the company that's really good and

strong. That's why I think we have some

advantages there. The point is like I

think where we are with how far our AI

has gone, our technology is so good.

We're in like this a bubble over here on

the left. And so we've got such a huge

road map to cover. You could see, you

know, some of our screenshots and

there'll be more in the in the next few

days at reinvest.co or houseack.com.

Read the solicitation. This is not a

solicitation. Read the offering circular

on the web page. Right? That's that's

where you want to be on the S-curve

because you got so much growth ahead of

you. Eventually, we'll be here. Now,

hopefully that's when, you know, knock

on wood, we're like a $50 billion

company and we're, you know, we're we're

bringing real estate AI global.

I could dream, right? That those are my

goals. Those are my ambitions. That's

what I like. That's what I think about

every day. But that is all just an

analogy to saying, you know, goo is is

Google getting in at B, C, or D? Well,

in 27, there's probably a good chance

we're going to be knocking on the door

of D.

Okay, so shortterm, this is really great

for Google. Long-term, I don't know,

could be could be a little late. Okay,

now what happened here with Nvidia? So,

the first thing that Nvidia does is they

start lashing out at Michael Bur's

comment about the stockbased

compensation that Nvidia has paid.

They're like, "Oh, you actually

calculated the math wrong. we spent less

on stockbased compensation than you said

we did because you forgot to account for

taxes. Okay. But what Nvidia doesn't

realize is when they're basically

complaining by saying, "Oh, well, you

know, the market undervalued us back

then and and that's why the average

share price of stock comp was lower and

and you know, uh, you forgot taxes."

What Nvidia really is doing here is

being petty. Nvidia because the fact is

Bur's right. you know, the more of this

dilutive stock comp you issue, the more

of that upside you did take away from

other existing shareholders. Now, in

fairness, there's a flip side to that

argument because maybe one of the

reasons Nvidia did so well is because of

the stock comp, right? So I think this

is splitting hairs that Nvidia is really

responding to this because every company

that's public is doing stock comp even

non-public companies right now under

accounts receivable Bur's like oh you

know days of sales outstanding go from

46 to 53 mind you literally mentioned

Bur right here see Michael Bur on

Twitterx and they're like they basically

say nah bro you got the numbers wrong

we're fine everything's in line with

historical averages nothing has changed

here fine I'm okay with saying

somebody's right, somebody's wrong about

the days there. This is where things get

interesting. So, I wrote that this was

from my video. I'm not trying to like

full ego here, pat myself on the back. I

realize like other people can realize

this as well, but you know, we realized

it live together. We went into their

financial statements and we're like,

"Hey, you know, why did you say that

your inventory shot up because you guys

were trying to get, you know, material

together to make sure you have enough

for for future sales, but then you just

conveniently ignored the fact that

finished good inventories also

skyrocketed." Like, why would you not

also mention that and then at least talk

about that? It's kind of like you sort

of expected people not to look at that

and that just came across as a little

disingenuous, right? That was in the 10

Q. I'll show you right here. So, the 10

Q I probably should just Oh, there it is

right here. So, we wrote this out and we

said that, hey, you know, I get it. Your

raw materials are up 23% in 9 months.

But your finished goods are up 109% so

double and your work in progress is up

150 6% so a double and a half right and

so what they do in this letter is they

basically say subscribe to me Kevin's

alpha report at me Kevin okay they don't

say that uh they they basically say hey

hey listen listen you know growing

inventory doesn't necessarily mean

demand is weak it could but but we're

not saying that it is or isn't it you

know it could not answer that we're just

going to say that you Growing inventory,

just so you know, also includes

inventory for raw materials and work in

progress. Okay, mind you, they didn't

mention anything about the fact that raw

materials only went up 23% in 9 months

and finished goods went up over 100%.

And then they say, oh, by the way, uh,

growing inventory doesn't mean that

customers aren't accepting delivery

without payment. You know, we recognize

revenue upon shipment and only once we

deem that people can actually pay us.

But then the counter-argument to that is

okay so hence your finished inventory

went up because you didn't ship it yet

because you're not sure people can pay.

Like they actually did a really bad job

of answering the questions. I would be

pissed if I were Jensen and my a like

first of all this note. You think they

sent this to us normal humans? No. They

send it some jerkoff Wall Street

analyst, you know, probably working at

some, you know, ripoff bank

and we have to hear about it through

leaks.

And then what the hell? We don't even

get good answers. Like the document kind

of sucks. It's sort of embarrassing. I

think Nvidia did a bad job on this

document. I feel like they they raised

more questions than they answered. Look

at this. They go, "By the way, we're not

like Enron. Nvidia doesn't hide using

special purpose entities to hide debt

and inflate revenue. Okay. Well, you

know who does? Meta. And they do it to

buy Nvidia chips.

Meta literally through Blue Owl has

structured a special purpose entity to

keep lease payments, long-term lease

liabilities, and the debt off of Meta's

balance sheet. So they could use a $30

billion data center and not have it on

their balance sheet. That's insane.

I don't like it. It's smart by Meta.

It's very smart. I'd rather them not be

responsible for the debt because if the

bubble blows, Meta doesn't have to pay

that crap back. So it's it's great for

Meta, but it's brutal for the whole

industry because Meta's like, "We'll

peace out, bros." Like, they're being

blunt. We we'll we'll peace out if we

need to. And then who'll be left holding

the bag would be like Nvidia with lower

growth forecasts and obviously Blue Owl,

right? So it's ironic to me that Nvidia

would literally say the words Nvidia

doesn't use S, you know, special purpose

entities to hide debt and inflate

revenue. Meanwhile, your customers are

literally doing that. [laughter]

Uh, okay. All right, whatever. Then

Nvidia's like, "Oh, also we don't do

vendor financing arrangements to

artificially grow revenue." Okay, fine.

But you do lease back or guarantee that

you will lease back the very garbage

you're selling, which is basically

vendor financing because you are now

providing a stable stream of cash flows

to these businesses that they can then

go to the bank with and go, "See, I have

a stable form of cash flows coming from

Nvidia. Can I get cheap financing?" So,

it's sort of like it's it's almost like

why Okay, watch this. Watch this. You

ready? What this is like? It's almost

like Mercedes. You going into the

Mercedes dealership and you walk in,

you're like, "Bro, SL300 or 500,

whatever it is. I'll take the SL500. I

think that's a car. That's that's that's

the the big one, right? I think it's

500. I can't remember what the uh letter

I'm pretty sure. SL500. Yeah, baby. I

want the SL500.

The 2006 SL500,

according to the first result here on

Google, will cost you 300 grand.

[laughter] Okay, I want the SL500. You

walk in and the car dealer is like,

"We'd love to sell you the SL500, but

unfortunately, you don't qualify because

your income is too low." And then you're

like, "Damn, I really want the SL500.

Like, I'm a ready willing and able

buyer. I want like I'm going to door

dash with it. And they're like, "Hold on

a moment.

Hey man, can we help this guy out?"

Okay. Okay. Okay. Got it.

We will invest in your Door Dash

business and provide you $20,000 a month

of Door Dash revenues.

Huh, sounds great. And now you can

qualify for the SL500. Hell yeah, boys.

[laughter]

I mean, that's basically what it is. I

think it's a great analogy. I hate to

say it, but that's basically what it is.

You put unqualified dunes in the Nvidia

sales office. Nvidia's like,

they can't qualify for dirt. No bank's

going to lend to them. Well, let's just

give them a steady stream of cash flow

and then people will lend to them.

[laughter]

Don't worry, boys and girls. It is not a

bubble. It is fine. [laughter]

Oh my gosh. All right. Anyway, so you

know, again, a lot of this to me, this

this lashing out about the appreciation

curves or whatever. To me,

this is this is really

a sign to me of why like why is Nvidia

going to this level of responding to

this stuff? And then they don't even

give us good answers. Like we already

know Michael Bur's depreciation curve

argument is down the road. Michael Bur's

depreciation curve argument will happen.

Not today. It's not happening right now,

but when we get to that D level on the

right, that's when Michael Bur will be

right. Right here is when Michael Bur

will be right. Now, when Michael Bur

will be wrong is if we get another

S-curve on top of that. So, that's

called stacking S-curves. When you

basically get another S-curve that booms

up right here and then you have another

growth cycle. That's always a dream.

It's almost always a dream. Like

companies make most of their money on

the first S-curve and then they're like

I mean it's kind of honestly like the

Model 3 right on Tesla Model 3 and FSD

first S-curve and then they're like

trust us bro we'll get to robo taxis and

Optimus that's the hope S-curve

okay like is it going to happen we don't

know that's the hope S-curve right

usually the subsequent S-curves are an

order of magnitude harder to actually

achieve like the probability is way

lower of that other S-curve happening.

It's weird because you'd think like, oh,

but they have enough money, they should

be able to pull it off hopefully. Uh,

but anyway, it it that's just

historically what we find. So, that's

Google, that's Nvidia, and then you have

Nick T.

He kind of forgot something pretty

important. So Nick T uh wrote this

article right here and he's like Fed

Chair Powell allies provide opening for

December cut and basically he says it's

going to come down to Powell. He's like

look this is a divided committee. The

last time we were in this divided

committee it was uh 2019 and it came

down to Powell. You know three rate cuts

is usually the last they do three rate

cuts and they're like okay now we're

going to pause blah blah blah blah blah.

Okay. My opinion is what Powell is going

to do is actually what Nick T did not

mention in this. I was so disappointed I

actually ended up leaving a comment uh

right here in the Wall Street Journal.

Okay, look. I got one like. I liked it.

Our boy meet Nick totally neglected to

mention Powell's likely reliance on ADP

data coming up. Maybe nobody knows what

that is in the Wall Street Journal

comment section. both the monthly and

two weekly data sets before PAL's

meeting. We get an ADP weekly tomorrow.

We get a monthly next week, Wednesday,

that's December 3rd, and then we get a

weekly on December 9th, right before the

FOMC meeting. So, we're literally going

to get three ADP data prints before that

meeting. So, stop speculating now. Just

pay attention to the ADP prints. It's

not that damn hard.

>> [laughter]

>> That's my take, you know. I I don't

know. So, anyway, um if any of this

makes you interested in uh the what

we're doing over at Reinvest, by the

way, we are going to be updating this

website a lot uh probably over the next

48 hours, so you can kind of refresh it,

but we're going to be posting uh new

pictures, pricing,

um details on like the release schedule.

This is the rough release schedule right

now. Uh we expect all of our features to

be live by the end of 26 and of course

we'll think of more by then I'm sure but

uh we expect all of our features to be

live by the end of 26 and then we'll get

to sort of our our full and normal

pricing. Uh and so we want to you know

we'll do some kind of like really low

foundation intro pricing but we're

really excited about it not only

releasing the AI app but we hope you all

love it and uh and then we improve it

over time together. I want you to see

the progress on it, too. Like, you know,

there'll be issues, there'll be bugs,

but then I want you to see the progress

on it because then if you're like, "Oh

my gosh, you know, it started like that

in December." And then in February, it

looks like this. Holy smokes, look at

that progress. You know, when you see

that, you'll see where we're going with

this and and what we're growing and what

we're cooking. So, uh yeah, we could we

couldn't be more excited. Uh anyway, uh

that's all I got for us. Thank you so

much and we'll see you in the next one.

Why not advertise these things that you

told us here? I feel like nobody else

knows about this.

>> We'll we'll try a little advertising and

see how it goes.

>> Congratulations, man. You have done so

much. People love you. People look up to

you.

>> Kevin Praath there, financial analyst

[music] and YouTuber. Meet Kevin. Always

great to get your take.