TRANSCRIPTEnglish

DeepSeek STRIKES Earnings | Tesla, Microsoft, Meta YIKES

20m 12s3,571 words508 segmentsEnglish

FULL TRANSCRIPT

0:00

make sure you go to meetkevin.com coupon

0:02

code expires in 2 days will include

0:05

lifetime access to the meetkevin alpha

0:08

report every morning we will start

0:10

charging for that at the end of February

0:13

so get in now before the price really

0:15

goes up and more content comes out check

0:18

it out at meetkevin.com any questions

0:20

always email us at staff at

0:21

meetkevin.com but for now we got to talk

0:25

deep SE well what did we just hear from

0:27

Microsoft meta and and Tesla and why the

0:32

heck is Microsoft down

0:34

5.4% in the after hours whereas Tesla is

0:38

up 4% in after hours and Nvidia down

0:41

about 1 and a quarter and all these

0:43

numbers obviously could shift and flip

0:45

and flop all over the place but a lot of

0:47

people were wondering what what's the

0:48

news what's going on so let's get

0:50

through it so first of all uh Microsoft

0:53

missed slightly on cloud now they're

0:55

blaming not having enough data servers

0:58

available for them to satiate the Azure

1:01

demand that there is but some people are

1:03

scratching their head going wait a

1:05

minute H how do you'all miss on

1:07

intelligent Cloud now they only missed

1:10

by about

1:13

1.35% so it's not that big of a deal in

1:16

terms of the Miss but I think the

1:19

markets were a little bit more spooked

1:22

because remember we went into this

1:25

expecting Microsoft and meta to say

1:27

something like hey uh we we have no

1:30

plans to spend or change our Capital

1:32

expenditures but we will see what we can

1:35

learn from Deep seek and incorporated

1:38

into our efficiency models at the

1:40

company right and guess what Facebook

1:44

which is up 2% right now and after hours

1:47

said literally exactly this it's it's

1:50

almost like they listen to my video

1:53

before earnings here's what meta is

1:55

likely to say and they copied it

1:57

Microsoft didn't listen to the video

1:59

meta said hey right now we still

2:02

anticipate our cap backs to be be

2:04

between 50 and 65 billion that's up from

2:07

the 51 they previously projected uh and

2:10

they quote hope to implement DC advances

2:13

at meta they see a business

2:16

opportunities from Artificial

2:17

Intelligence coming next year in terms

2:20

of being able to actually monetize

2:22

artificial intelligence but it's too

2:23

soon for us to make any kind of changes

2:25

to our capital expenditure guidance you

2:28

know just two days after the release of

2:30

deep seek essentially So Meta sort of

2:32

looked at this and was like oh yeah like

2:35

we can't make any changes yet on our

2:36

capex but oh absolutely we're going to

2:39

end up spend like we're going to be more

2:41

efficient we're going to learn what we

2:42

can and open source is the future yeah

2:46

Okay cool so that was meta's take to

2:49

deep seek well what did Microsoft say

2:52

Microsoft instead of saying this came in

2:55

and said well so we missed on cloud uh

2:59

and Kathy Wood by the way is complaining

3:01

that Microsoft is seeing uh you know

3:03

growth Decline and so we can kind of

3:05

explore that more deeply separately but

3:08

uh when it came to nadela speaking

3:10

satian Nadella speaking about deep seek

3:11

he says deep seek has some real

3:13

Innovations and we actually see capex

3:16

growth being lower this year than it was

3:20

last year and there are definitely

3:22

significant efficiency gains to be had

3:24

from artificial intelligence now this is

3:26

really interesting because you kind of

3:28

have this like Duality here

3:30

and my take is markets are looking at

3:32

Microsoft going okay like y'all are high

3:35

margin y'all obviously would be a

3:38

beneficiary of more efficient AI but

3:41

only to the extent that you can actually

3:43

grow on share and I think they're

3:45

getting beat up here a little bit

3:46

unfairly but I think they're getting

3:48

beat up on missing that cloud growth I

3:51

think markets might be looking at

3:53

Microsoft going okay well if y'all are

3:55

going to be more efficient we want you

3:57

to sell more but you're not selling and

4:00

more so we'll go somewhere else but

4:02

honestly it's not just Microsoft getting

4:04

beaten up after these earnings its

4:06

Salesforce is down 3% after that

4:08

Microsoft misses well possibly in

4:10

sympathy and they've been running quite

4:12

a bit over the last few days meta is

4:14

still holding on to its

4:16

2.2% but that's probably because meta

4:19

makes most of their money from

4:20

advertising and frankly they're crushing

4:23

it with advertising I mean look at M

4:26

meta numbers for example the average

4:28

price paid for for an advertisement at

4:30

meta is up 14% year-over-year that is

4:34

phenomenal growth uh over at meta uh

4:37

you've also got to remember that even

4:39

though they missed on their guidance by

4:42

2.5% their their numbers across the

4:45

entire world are frankly great you look

4:47

at ad Revenue growth at 28% at the rest

4:50

rest of the world 48% operating margin

4:53

they costs only up about 15% well

4:55

they're really exploding with not only

4:58

advertising but also the potential side

5:01

of innovation Mark Zuckerberg for

5:03

example started by talking about the

5:05

rayb band metag glasses and how hey the

5:08

market will tell them if this is

5:09

something that billions of people are

5:11

going to want to use or if it's going to

5:12

be a slow schlog or grind so it seems

5:15

like for some reason markets were really

5:17

resonating with meta here a little bit

5:19

more uh service now was another one of

5:21

those software businesses that just

5:24

disappointed with the subscription

5:26

Outlook not great and the stock down

5:29

about 8 % in after hours after they had

5:31

a Miss on their guidance so it seems

5:33

like markets today are saying look

5:36

that's cool y'all can be more efficient

5:38

because of deep seek you all are

5:40

reiterating the innovations that are

5:43

happening with the Deep seek technology

5:45

but we need to see big dollars com into

5:47

the bottom line and growth and if you're

5:49

not going to grow like Kathy Wood said

5:52

regarding Microsoft then maybe this

5:53

isn't the place to be and so I think

5:55

that's kind of why you started seeing

5:57

Microsoft trade down a bit uh not only

6:00

that Miss but also that mention that

6:02

we're going to focus more on short-lived

6:04

assets so basically focusing on more

6:06

near-term spending that could be like a

6:08

marketing campaign right like oh we'll

6:10

Market our business and we'll use that

6:12

marketing expense this year and that's

6:14

it as opposed to like a longer term

6:18

asset which would be like another data

6:19

center so they're kind of in this

6:22

Duality where they're like on one hand

6:23

we're data and compute constrained but

6:25

on the other hand things are getting

6:27

more efficient and we don't want to keep

6:28

building data centers because maybe

6:30

we'll just sort of converge into that

6:32

efficiency now that's kind of a

6:34

complicated idea right there so I want

6:36

to draw that one out and and I'm not

6:38

saying that you can't understand it I

6:40

just I think the way that I'm conveying

6:41

it is a little bit complicated but I

6:43

think it's actually really important for

6:45

us to think about so let's grab a page

6:47

on the iPad here and uh let's draw this

6:51

out this this thesis and then I want you

6:53

to tell me what you think about this

6:55

thesis let's see if we're on the same

6:57

page on this puppy all right so and then

6:59

of course we'll do our test learnings so

7:01

I think there's a chance with what

7:03

Microsoft is doing is they saying look

7:05

we have spent uh a lot of money on

7:09

artificial intelligence right so

7:11

spending on AI infrastructure let's call

7:15

it this right that's spending on AI

7:18

infrastructure and then let's go ahead

7:20

and draw a uh AI

7:23

needs curve so we'll do uh uh we'll kind

7:27

of draw this logarithmically

7:31

AI needs curve actually it's kind of

7:33

like logarithmically and then it turns

7:36

more what is that parabolic upside down

7:41

Parabola so uh let me label this uh and

7:45

then I'll I'll sort of put

7:47

like where we are now I I'll call

7:51

this the

7:53

now it feels like uh and I could be

7:57

wrong about this but it it feels feels

7:59

like this is what they're talking about

8:02

so spending on AI infrastructure is this

8:06

that's the now and this right here is AI

8:10

Demand right so they're at a place where

8:14

right now there's a shortage this

8:17

section right here is the

8:20

shortage but the problem is because AI

8:23

is getting more efficient they actually

8:26

see in the future a collapse in the need

8:30

for data centers that they have and so

8:33

if they stop building let's say they

8:36

stop building right here eventually what

8:40

this section right here is going to be

8:42

is uh excess uh

8:46

compute uh and I think when we get into

8:48

the Tesla earnings you're going to see

8:50

some of that as well though Tesla didn't

8:52

get punished as much because Tesla has

8:55

the argument of well Optimus robots are

8:58

going to need way more compute so just

9:01

stay tuned and I'll talk about Tesla in

9:03

just a moment

9:05

but to me I think like I I personally

9:10

would would consider taking a screenshot

9:13

of this right here I'm not saying that's

9:15

because this is like beautiful art it's

9:17

because I I really think if you're an

9:19

investor you want to look at this

9:22

critically and you want to make an

9:24

opinion about this you want to say do I

9:26

agree with this so so again the dotted

9:29

line is where we are now the orange

9:31

vertical line is spending on AI

9:33

infrastructure and the reason it's not

9:35

exponential is because there's a limit

9:37

to how quickly you can get chips and

9:39

Machinery Machinery just sort of grows

9:42

linearly it doesn't really grow

9:44

exponentially uh whereas demand can

9:47

right demand can grow overnight like we

9:49

saw during the pandemic but actually

9:51

manufacturing stuff uh doesn't usually

9:55

grow exponentially it usually does this

9:58

stair stepping thing

9:59

where it probably practically looks like

10:02

this uh and and then the linear line is

10:05

really just the average of it right but

10:08

the point is if AI demand is so

10:10

explosive to where yes we're compute

10:12

constrained but then you have Microsoft

10:14

saying we actually don't need this much

10:16

like we're not going to invest that much

10:18

anymore into capex like we're going to

10:19

slow our capex growth I think that's

10:21

because they actually see those

10:23

efficiencies at some point putting them

10:26

in a place where they're going to have a

10:28

substantial excess of supply of compute

10:31

uh like think about this like take this

10:33

out I don't know 3 years when it's all

10:35

built okay so you really need new demand

10:40

sources for this stuff uh look go go out

10:44

into this direction over here and what

10:46

you end up with is this entire section

10:49

right here all of this widening this all

10:52

becomes excess compute what you need are

10:55

more purposes more purposes to fulfill

10:59

this excess compute and so what's an

11:01

example of excess compute that can

11:03

actually push that red line up or that

11:05

that sort of demand curve up what can do

11:08

that well honestly Optimus is something

11:11

uh that can push this curve up and so I

11:15

think that's why Tesla is uh reacting a

11:19

lot more positively in the after hours

11:22

at least right now uh you know Stock's

11:24

up over 3% it was down as much as 5%

11:27

when it first missed on its numbers

11:29

we're going to talk um uh Tesla here in

11:31

just a moment but uh really when you

11:34

listen closely to what Elon said what

11:36

was different this time was that they

11:39

talked in their shareholder deck about

11:41

how they have this uh

11:43

50,000 unit server cluster uh in fact I

11:47

have a picture of it right here where is

11:50

it right here right here it's the cortex

11:53

50,000 GPU training cluster that was

11:56

different you know something else that

11:57

was different was them recognizing $600

11:59

million of revenue from Bitcoin uh as a

12:02

mark to Market which also means if

12:03

bitcoin's going to go down in the future

12:05

they'll have to take the hit but the

12:07

fact that they took a

12:08

600,000 uh sorry $600 million bonus like

12:12

to their income and they still missed

12:14

earnings per share on an adjusted basis

12:16

by 2 cents they missed Revenue uh by

12:19

like $1.5 billion dollar uh and uh their

12:24

gross margin came in at

12:25

16.3% versus the estimate of 18.9% with

12:29

margin excluding regulatory credits at

12:31

just 13.6% these numbers were not good

12:34

but the thing that actually kept them

12:36

going was that they have more compute

12:38

but not only do they have more compute

12:40

they have the purpose and so that is the

12:43

opposite of what Microsoft is feeling

12:45

because again compare this Microsoft

12:48

equals uh more compute uh uh uh but uh

12:53

Miss On Demand right uh and then at

12:58

Tesla what you actually have is uh more

13:03

compute more demand coming via Optimus

13:07

so this is a a very different artificial

13:11

intelligence story they're not really

13:13

relyant on an outside source for keeping

13:16

their servers busy and I think that's

13:18

one of the reasons why Tesla got

13:20

rewarded here even in the face of crap

13:23

margin but then again remember when I

13:25

did my earnings preview we talked about

13:26

this with course members in the alpha

13:28

report we talked about this in videos on

13:30

the channel like I talked about this

13:31

everywhere okay this was I think I was

13:33

as clear as I could possibly be here

13:36

that the Tesla earnings today were never

13:38

going to matter about the vehicle

13:39

margins or vehicle deliveries we already

13:41

knew all this stuff was going to be crap

13:43

it was all going to be about the future

13:46

Optimus when are we going to get it oh

13:48

okay you guys are going to try to build

13:50

a few thousand by the end of the year to

13:51

have them functional in factories and

13:53

then you want to get to ideally 10,000

13:55

uh a month at some point in the future

13:58

your first line will be th000 a month

13:59

then the next line will be 10,000 a

14:01

month then you'll be at 100,000 a month

14:03

that's awesome that's really exciting

14:05

okay what else can you tell us oh you're

14:07

going to start doing cyber cab tests in

14:09

uh uh Austin uh potentially in June and

14:12

you're probably going to release uh full

14:15

self-driving

14:16

unsupervised later this year in

14:19

California and Texas and then other

14:21

parts of the United States this year

14:23

that's great hey what about Hardware uh

14:25

like Hardware 4 oh crap cars with hard

14:29

Ware 3 you're going to have to upgrade

14:31

to Hardware 4 to probably get

14:33

unsupervised oh no and then to that

14:35

elon's like good thing a lot of people

14:37

didn't buy the FSD package because he's

14:39

basically implying that they're going to

14:41

provide that upgrade for free because

14:43

yeah you're going to need a hardware

14:44

upgrade you're going to need a new

14:45

computer upgrade to actually get to to

14:48

this level so it'll be interesting

14:49

because I've got two cars I think

14:52

actually one of them might still be on

14:53

hardw War 2 but anyway I've got two

14:54

older ones and then I've got the Cyber

14:56

truck which is on hard War 4 so we'll

14:58

see I guess but anyway um they also

15:00

still sort of tease this idea about the

15:02

launch of a new more affordable model

15:06

but nobody's asking questions about it

15:08

or they're purposefully dodging

15:09

questions on it and not even having them

15:11

ask those questions uh they keep teasing

15:14

it they mention it in their prepared

15:16

remarks they mention it in their

15:17

shareholder deck that uh uh you know a

15:20

more affordable model should come in

15:22

20125 but they never talk about is this

15:25

going to be a whole new car is this

15:27

going to be like a refresh version of

15:29

some existing product is this just going

15:31

to be a low battery variant of something

15:34

that we already have like a lower

15:36

smaller battery model 3 uh they also

15:38

have a an interesting slide where they

15:41

they kind of suggest they're not going

15:42

to use the stainless steel ever again in

15:44

future products so maybe they kind of

15:45

learned a poopy doopy lesson there see

15:47

right here they say use in future

15:49

vehicles and they check mark steer by

15:51

wire adaptive air suspension custom

15:54

laminated glass ether Lube 800 volt

15:56

battery system 48 volt architecture

15:59

that's how you don't get the 12vt

16:00

cigarette lighter anymore it's actually

16:02

kind of annoying if you have old

16:03

cigarette lighter products but uh look

16:05

what's not going to be included the

16:06

stainless steel

16:08

exoskeleton is not going to be used in

16:11

future Vehicles so I thought that was

16:13

kind of interesting because that sort of

16:15

implies that they're not really planning

16:16

on a future version of like a cyber

16:19

truck or a cyber van and instead it'll

16:21

be more like the Roven the robo van or

16:24

or whatever they talked about coming out

16:26

with Apple watch support through their

16:28

app I guess I need to get that app

16:29

update cuz I'd love to have the key on

16:31

my watch instead of on my phone that'd

16:33

be really cool uh talked about how

16:35

they're producing 2 and a half th000

16:38

cyber uh truck worth cyber trucks worth

16:41

of

16:42

4680 uh and uh Mega sha Mega Factory

16:45

Shanghai was completed in Q4 that's

16:47

that's all fantastic then we've got uh

16:51

they've got the roof going on the Reno

16:54

facility for the semi- trucks they

16:56

expect to manufacture semi- trucks next

16:58

year Elon Musk as usual and almost I

17:01

mean frankly I predicted it on the live

17:03

stream you should go watch it just cuz

17:04

it's a meme at this point I'm like well

17:07

elon's going to get off playing Diablo

17:09

or whatever he's playing and then he's

17:10

going to get on the radio or you know

17:12

his his phone be like oh what are we

17:14

doing guys oh earnings call Tesla is

17:17

going to be the most valuable company in

17:18

the world uh and then today he actually

17:21

surprised me because he got on he goes

17:23

Tesla is not only going to be the most

17:25

valuable company in the world but uh

17:27

it's going to sound like Donald Trump

17:29

and it's probably going to be worth five

17:32

times as much as the most valuable

17:34

company in the world today it's going to

17:37

be huge it's going to go ballistic he

17:39

had a lot of

17:40

excitement at the beginning uh let's

17:42

just say everybody's so used to hearing

17:44

at this point it didn't move the stock

17:46

at all but that's okay because the stock

17:48

really shouldn't move off that because

17:49

it was kind of just like a pump we

17:51

really want to hear updates uh and so

17:55

again very very optimistic but uh he

17:59

kind of goes through the usual how

18:01

useful and autonomous car is going to be

18:03

in the future and blah blah blah but

18:04

really they want to tow into doing FSD

18:07

trials in June in Austin as an initial

18:10

operated by uh Tesla and hopefully

18:14

they'll have a few thousand Optimus

18:16

robots at the end of the year operating

18:17

in their facility and then next year we

18:20

get into more volume production for

18:21

Optimus and cyber cabs uh probably where

18:24

then they can expand the robo taxi

18:26

service so to speak so uh internal only

18:29

for now they do think that they can get

18:30

the Optimus cost down to $20,000 and

18:33

cost a good sold which actually pretty

18:35

good given that they're cogs right now

18:36

on a car is about $34 $35,000 but their

18:40

average selling prices have gone down

18:42

their average selling prices are so low

18:43

right now that uh they uh they're at the

18:46

lowest margin per vehicle right now that

18:49

they have seen uh since

18:51

2018 and that's before the EV tax

18:54

credits go away so you know some

18:57

potential pain on the ACT ual EV and the

19:00

like manufacturing side but again the

19:02

market is rewarding the enthusiasm

19:04

because yes as AI becomes more efficient

19:06

it'll actually be a Big Boon to Tesla so

19:09

the Deep seek model would be a huge Boon

19:11

to Tesla because really you're you're

19:13

able to likely operate artificial

19:16

intelligence at frankly a much more cost

19:20

effective uh level of energy or spend

19:23

essentially to accomplish hopefully uh

19:26

that 10x requirement that Optimus robot

19:28

are going to need since they're going to

19:30

be doing as Elon says so many more

19:32

things than just what a car does so

19:34

anyway that gives you a little bit of a

19:35

deep- seek impact on the Tesla earnings

19:38

currently up about

19:40

4.4% Microsoft earnings currently down

19:42

about

19:43

5.28% uh

19:45

and meta earnings uh currently up about

19:49

1.8% there you have it check out

19:51

meetkevin.com thank you so much for

19:52

watching and we'll see you in the next

19:53

one can not advertise these things that

19:55

you told us here I feel like nobody else

19:57

knows about this we'll we'll try little

19:58

advertising in see go congratulations

20:01

man you have done so much people love

20:02

you people look up to you Kevin P there

20:04

financial analyst and YouTuber meet

20:06

Kevin always great to get your take

UNLOCK MORE

Sign up free to access premium features

INTERACTIVE VIEWER

Watch the video with synced subtitles, adjustable overlay, and full playback control.

SIGN UP FREE TO UNLOCK

AI SUMMARY

Get an instant AI-generated summary of the video content, key points, and takeaways.

SIGN UP FREE TO UNLOCK

TRANSLATE

Translate the transcript to 100+ languages with one click. Download in any format.

SIGN UP FREE TO UNLOCK

MIND MAP

Visualize the transcript as an interactive mind map. Understand structure at a glance.

SIGN UP FREE TO UNLOCK

CHAT WITH TRANSCRIPT

Ask questions about the video content. Get answers powered by AI directly from the transcript.

SIGN UP FREE TO UNLOCK

GET MORE FROM YOUR TRANSCRIPTS

Sign up for free and unlock interactive viewer, AI summaries, translations, mind maps, and more. No credit card required.