TRANSCRIPTEnglish

The 30,000 Amazon Job cuts are just an early WARNING.

21m 50s3,891 words568 segmentsEnglish

FULL TRANSCRIPT

0:00

the redhead crossing the Bloomberg

0:01

terminal. Reuters reporting that Amazon

0:02

plans to cut up to 30,000 corporate

0:05

jobs. Those jobs represent nearly 10% of

0:08

corporate workers. Those jobs cuts jobs

0:10

cuts are set to start on Tuesday. That

0:12

again according to Reuters.

0:14

>> The irony of these companies cutting

0:16

jobs is that it is exactly what the AI

0:19

revolution does. the stock actually

0:22

falling slightly on that news. But the

0:24

opposite should be what happens because

0:27

when these companies cut jobs, guess

0:30

what happens, folks? Operating margins

0:32

go up. That is the biggest risk factor

0:35

for artificial intelligence right now.

0:37

And there was a great interview that we

0:39

went through in the information that

0:41

really talks about this

0:44

transformation of artificial

0:46

intelligence really benefiting the

0:48

companies that are AI native style

0:51

companies like AI first companies

0:53

because they're most capable of

0:56

profiting off of artificial intelligence

0:59

because right now you have a lot of

1:01

people that use AI but they say it's

1:04

kind of like giving Joe Blow a race car

1:07

and saying go drive it. And so the non

1:10

AI first companies will probably profit

1:13

the least. But when you have a company

1:15

like Amazon, this is a company that's

1:17

clearly at the forefront of the

1:19

artificial intelligence revolution. And

1:20

they recognize that, hey, artificial

1:23

intelligence is going to let us cut a

1:25

vast majority of low-level work that we

1:28

just don't need at the business. So is

1:30

it a surprise that corporate staff are

1:31

going to get cut, especially more

1:33

entry-level positions? No, not at all.

1:35

In fact, that's what's happening. across

1:37

the board at companies across the entire

1:39

country or even across the world and

1:41

this is normal. So guess who wins? Is it

1:45

people that win on net? No. People don't

1:48

win. It's corporations that win.

1:51

Corporations end up winning because they

1:53

can lay off 10% of their corporate

1:56

workforce, let's say. And then what does

1:58

that end up doing? It props their

1:59

margins. You've already seen this happen

2:01

at Amazon. Look at this. Amazon

2:03

year-over-year went from an operating

2:05

income, so an operating margin of 9.9%

2:09

to an operating margin of 11.4%.

2:12

Just in the last year, before even this

2:14

announcement, they've already reduced

2:16

their GNA expenses from over $3 billion

2:19

to under $3 billion in a quarter. And

2:23

that is likely now to accelerate as they

2:25

lay off even more people. Now, who wins?

2:29

And this is the big thing about well

2:31

frank frankly

2:34

this artificial intelligence revolution.

2:36

Who wins? The corporations win and that

2:40

sucks for people on net. Look at this

2:43

particular chart here. This was an

2:44

interesting one. So this particular

2:46

chart shows here rage against machines

2:49

US change in employment 3 or a month

2:53

after the release of uh GPT3.5.

2:56

And so what you end up finding is that

2:58

AI adopting firms and nonAI adopting

3:01

firms had no real net change in their

3:04

senior employees. So the ones with as

3:07

Andrew Carpathy says agency are the ones

3:11

that stay. What does he mean when he

3:13

talks about agency? What he means is the

3:15

people who have the ability to make

3:17

decisions to allocate resources. The

3:21

people who take the benefits of

3:23

encyclopedic or resource artificial

3:25

intelligence, the pattern recognition

3:28

that we get from the models of AI, those

3:30

people who then apply the results of AI,

3:33

the senior developers, the executive

3:35

staff, the corporations, they benefit

3:38

from artificial intelligence. But the

3:40

entrylevel workers who don't have

3:43

agency, who don't have the ability to

3:45

allocate resources, they lose. And so

3:49

this Amazon announcement about laying

3:51

off 30,000 employees, this should come

3:53

as no surprise and it's expected to

3:55

continue to happen across our entire

3:57

economy. This individual in this

4:00

artificial intelligence piece by the

4:01

information, they actually talk about

4:04

potentially a a third of workers. See,

4:07

they say the US's economy has a $15

4:09

trillion labor force and suggests that

4:12

maybe you could replace in the next 5 to

4:14

10 years 1/3 of the labor force with

4:17

artificial intelligence. Replacing $5

4:20

trillion of the labor force with

4:22

artificial intelligence. There's no

4:24

question that yes, AI will somewhere

4:27

create new jobs, new startups or

4:29

whatever. for example, hey, you know, if

4:31

house hacks AI, Reinvest AI keeps

4:34

booming the way so far our technology

4:36

has been booming. I mean, we raised over

4:38

we're at almost a half million dollars

4:39

raised in 36 hours now since we put out

4:42

sort of our reinvest.co artificial

4:44

intelligence uh roadmap over here, which

4:46

is really cool at reinvest.co.

4:48

Obviously, this video is not a

4:49

solicitation. Read the offering circular

4:51

here, but you can see all of our goals

4:52

and milestones. Just go to houseack.com

4:54

or reinvest.co. It's the same company

4:55

with a different DBA. But yes, startups

4:59

are going to hire people. Like as our AI

5:02

booms, which I expect it will, we are

5:03

going to hire people. But I think on

5:05

net, you're going to lose. I mean, we're

5:06

not gonna hire 30,000 people like, you

5:08

know, Amazon just laid off, right? And

5:10

we'll hire like three of those. Like,

5:12

how many great startups could you have

5:14

uh to pick up those workers? And so this

5:17

idea about, oh, but we're going to

5:19

create jobs with AI, yes, but very

5:21

slowly. So at first you could see this

5:24

massive loss of employment from

5:25

artificial intelligence that takes

5:28

decades to recover.

5:30

But in the meantime, who profits?

5:33

Big companies. People always wonder

5:35

like, oh, who's going to win from

5:37

artificial intelligence? It's big

5:39

companies. In fact, we also already know

5:42

that the cost of developing chips is

5:44

going to go down. The cost of buying

5:47

chips is going to go down. The cost of

5:49

using chips is going to go down. You're

5:51

going to see deflation in chips,

5:54

deflation in service provider, white

5:56

collar job pricing. The benefits of

5:59

those are the aggregators of all of

6:01

that. Big corporations that now don't

6:04

need as many employees on the white

6:05

collar side that could benefit most from

6:07

cheaper AI chip prices. Like, sure, do

6:11

small companies benefit or do you and I

6:13

benefit from cheaper AI prices? Sure, I

6:15

guess I could use GPT more and maybe it

6:18

doesn't cost $200 a month for the pro

6:20

version. Maybe it'll cost $99 a month.

6:22

Well, whatever, right? Who cares? What

6:24

really matters is that at scale, a

6:26

company like AI or Amazon can now

6:29

utilize AI to blow off 30,000 jobs and

6:33

start firing people. Understand what

6:35

30,000 jobs is. Okay? 30,000 jobs on a

6:38

white collar job. Call it $125,000 a

6:41

year. Plus, add in 20% for benefits

6:44

work. Actually, probably more than 20%.

6:46

add in 20% for workers comp and uh uh

6:50

retirement and regular stock comp. Now

6:53

add in additional competitive benefits

6:55

to work at Amazon. Uh you're probably

6:57

looking at another 40% for employee

6:59

related costs, perks, vacations, and

7:01

otherwise, right? That means per per

7:04

30,000 worker, you're probably looking

7:06

at $175,000 package. Multiplied by

7:09

30,000 people, that works out to

7:15

$5 billion. $5.25 billion a year. See,

7:19

that's Yeah. $5.2 billion a year. That

7:24

money can basically go straight to

7:26

Amazon's cash flow. They're they're

7:29

still going to provide our our crap

7:31

every single day. They're still going to

7:33

end up replacing people with symbotic

7:35

robots probably in their factories uh

7:38

and eventually, you know, maybe humanoid

7:40

robots to actually deliver our packages.

7:42

That'll all come in the future. We're

7:44

still going to get our packages. But

7:46

this white collar workforce reduction,

7:49

that's kind of the problem that we don't

7:51

really think of yet. And who benefits?

7:53

I'll tell you exactly where that $5

7:55

billion goes. So that $5 billion per

7:57

year works out to about $1.2 billion

8:00

over here uh in the GNA side, which

8:03

basically goes straight to profit,

8:04

right? Straight into margin. So where is

8:06

that going to show up? It's going to

8:07

show up in their cash flow statement,

8:09

which they actually put it on top. So

8:10

it's going to show up right here.

8:13

[snorts]

8:14

See, right now they're cash flow neutral

8:16

right now. Cash flow equals zero on free

8:20

cash flow at Amazon right now because

8:22

they're blowing it all on chips. Okay,

8:25

that cash flow explodes. Not only

8:27

positive again as you lay off white

8:29

collar workers, but it also as you stop

8:32

buying as many chips, your cash flow

8:34

machine is insane at Amazon. Yeah,

8:37

you're going through a a capital

8:39

intensive investment period right now,

8:41

but how lean these companies are

8:43

getting. This is this company is

8:44

becoming massively lean. I mean,

8:47

honestly, it's almost an investment. I

8:49

don't own any Amazon, but this is almost

8:52

like as I'm describing this, this is

8:54

almost becoming an investment thesis to

8:56

invest in Amazon. Ironically, the firing

9:00

of of white collar workers, which will

9:02

be very bad for humans, which I guess in

9:04

a circular manner means people will buy

9:06

less on Amazon if they can't get new

9:07

jobs, right? Which is also then bad for

9:10

Amazon. But, uh, for right now, let's

9:12

see what that looks like. So, Amazon has

9:15

a share price of about, you know, 226

9:18

bucks. get some details here. 226 bucks.

9:21

We've got EPS of 833

9:25

uh for the end of the year. Okay. So,

9:28

what we're going to do is we're going to

9:29

go 226 divided by 8.33.

9:33

That gives me a 27

9:35

PE ratio right now on 2025 numbers. Now,

9:39

growth of earnings, I'm looking at 10.4

9:42

+ 22.7 + 15.6 + 15.4 equals divid 4.

9:47

I've got growth of earnings of 16 16%.

9:52

Uh Wall Street growth next four years

9:55

averaged, right? So if I take 27 divided

9:58

by 16, holy smokes, they're trading for

10:02

1.68 peg, that's cheap, you know, for

10:06

what Amazon is. I mean, they're

10:08

basically selling. I mean, they're a

10:10

very labor intensive business with their

10:12

fulfillment services, but that's a fair

10:13

valuation in this economy. That's a fair

10:16

price for Amazon. Uh and and frankly,

10:20

their EPS growth could be substantially

10:22

greater if they do keep laying off HR

10:25

related staff or white collar related

10:26

staff. Uh so I actually I mean like

10:29

really when you put all of this

10:30

together, I hate to say it because it's

10:32

like so antihuman and like I have major

10:35

empathy for like people losing their

10:37

jobs. I have major empathy for just the

10:40

human condition. Like what is AI going

10:41

to do for us? Um, and and I believe and

10:45

I I say these things on my channel not

10:47

because they're popular. Uh, like I

10:49

think a lot of people when they hear me

10:51

say things like, "Oh, like like this

10:53

whole LLM thing, this isn't that big of

10:54

a deal. Like this is really pattern

10:56

recognition." You know, really, you

10:57

should watch this in full. Okay, I did a

11:00

breakdown on this in the Me Kevin app.

11:02

Remember, you can get the Me Kevin app

11:03

for free. Like this weekend I was

11:05

sitting watching AI you know videos and

11:07

I'm reading AI books and you know I

11:09

break down some of this on the

11:11

difference between uh here I mean you

11:13

could see on my on the meet Kevin app

11:16

patterns of AI pattern fitting the

11:18

mental sandbox reinforcement land

11:21

learning supervised learning LLM scaling

11:24

whatever you you could see all this kind

11:26

of stuff uh along with projections and

11:28

otherwise in the M Kevin app which is

11:29

kind of cool you know you get the app

11:31

for free you can also get the daily

11:32

wealth in there which is sort of a daily

11:34

notification I send out if you want it.

11:35

You can enable it or disable it. Uh and

11:38

uh you can decide to get that if you

11:40

want. I think it's a cool little

11:41

newsletter we do daily. See, if you

11:42

actually go to the app, you could go to

11:44

the little hamburger menu at the top and

11:45

then pick, oh, I don't want the daily

11:47

wealth reports or I do or I want Kevin's

11:49

alpha reports, but that's for course

11:51

members. Sale ends Wednesday, by the

11:52

way, for that. But anyway, uh you know,

11:56

I I do worry

11:58

dramatically that we're going to have

12:00

over the next 5 years this dramatic

12:03

decline in labor in the labor force.

12:06

It's going to really suck. Like we're

12:08

going to have a renter nation.

12:09

Basically, people aren't going to own

12:11

real estate. You will own nothing and be

12:13

happy, right? And I think the people who

12:15

will get the most screwed are going to

12:17

be people who are really heavily

12:19

leveraged. So, you have a lot of margin

12:22

debt, personal lines of credit, SoFi

12:24

personal loans, buy now pay later,

12:27

you're living the least lifestyle, you

12:29

have high expenses. Uh, and then what

12:31

ends up happening is you lose your job

12:33

and you're like, well, this sucks, you

12:35

know, and you lose your job and then

12:36

you're like, crap, I can't get another

12:38

job. And, you know, that's where

12:40

eventually you get this like slow bleed

12:42

economy that really just hurts people

12:44

before we actually get the generation

12:46

of, you know, the new wave of of jobs.

12:50

It could be decades, frankly. So, this

12:52

revolution we're going through in the

12:53

labor market. It doesn't have to be we

12:55

lose jobs now, we get jobs tomorrow. It

12:57

could be we bleed jobs for the next 5

12:59

years and we get jobs in 15 to 20 years.

13:03

And so, if if you're not mentally

13:05

prepared for how are you going to be

13:06

different? How are you going to be the

13:09

allocation of resources agent versus

13:12

just the person who's replaced because

13:14

your knowledge job is replaced? You

13:16

know, if you're not thinking about that,

13:17

then you're thinking about AI wrong.

13:18

Like if you're thinking I'm going to

13:20

wake up every morning and have AI

13:21

summarize my life and tell me what to

13:23

do, you're going to get replaced. And

13:25

this is what I say regularly. Like like

13:27

somebody left a a video or a comment. Uh

13:31

I'll I'll I'll share this because I

13:33

thought it was very interesting. I

13:35

replied to this on X and I thought it

13:36

was very very valuable. So, I did a

13:40

40minute video breakdown on, you know,

13:43

the the aircraft

13:45

and uh somebody here writes, "Love the

13:49

AI summary on a 45minute video, lol."

13:54

And like I gave what I thought are

13:56

millions of dollars of life lessons,

13:58

right? So, I wrote, "Wow." My response

14:00

was, "Wow, the loss of context is

14:02

insane. Imagine someone giving millions

14:04

of dollars worth of perspective for free

14:07

to be boiled down into a oneliner that

14:10

loses context completely. It will create

14:13

a two-tier division in learners, right?

14:17

This two tier division in learners

14:19

because you're going to have people that

14:21

like there are going to be people who

14:23

watch this video and go, "Wow, I learned

14:25

a lot if I ever want to be a pilot. I

14:27

learned a lot if I want tax benefits. I

14:29

learned a lot. If I ever want to own an

14:31

asset like an aircraft, I I learned a

14:33

lot about, you know, different aspects

14:35

of life, right? And so, they're going to

14:37

be the people who learn and then avoid

14:40

making very expensive mistakes or take

14:42

advice from this and go, "Wow, I can

14:44

make really good decisions because of

14:45

the advice in this." And then there are

14:47

going to be other people who do an AI

14:49

summary and be like, forget about it.

14:51

That's the other thing that happens with

14:52

AI summaries. You forget. Remember an AI

14:56

summary is basically skimming an article

14:59

and then it doesn't embed into your

15:01

brain like when I read a book like I'm

15:04

reading this machine learning and

15:06

finance right and I go to

15:10

even just this right here this book is

15:12

written for advanced graduate students

15:14

in academics and financial economic

15:16

econometrics

15:18

management science and applied science

15:20

statistics in addition to quants and

15:22

data scientists in the field of quantum

15:23

uh quantit ative finance, bro. I could

15:25

barely even read the sentence because

15:27

it's so high level. Like, and I'm not

15:28

saying like I'm at these high levels.

15:30

Like, I I try my best to read this and

15:32

and there's there's a lot of work

15:33

involved in this. Uh and and I enjoy

15:36

what I'm reading. Uh especially since

15:38

this is an older book and learning

15:40

about, you know, gajing processes or

15:42

whatever. Like, I love this kind of

15:43

stuff, but uh it's hard. But it's like

15:47

if we don't challenge ourselves to that

15:48

kind of stuff, we're just going to get

15:49

replaced. Uh and so you know you skim AI

15:53

everything you're going to get replaced.

15:55

So it's sort of like a war like this

15:58

Amazon layoff is really a warning to

16:01

everyone to wake up. And I think really

16:04

if we look at you know this this AI

16:07

interview here in the information this

16:08

is called

16:10

uh Venod talks AI power struggle blah

16:13

blah blah blah. Uh most enterprises who

16:15

are executing AI are doing it with their

16:17

people who are not qualified to execute

16:18

it. Well, the non-qualified people,

16:20

those are the people who are going to

16:21

get fired.

16:23

Uh, and so you will see these I mean,

16:26

they talk a little bit about increasing

16:27

margins at AI companies. Basically, they

16:29

make this argument like there's so much

16:30

profit in artificial intelligence that

16:32

like don't worry about AI margins.

16:35

There's so much money that even if

16:36

prices come down, companies will still

16:38

be able to make AI margins. This is

16:40

really true, by the way. Like, you know,

16:43

if you're at a 90% margin on AI, who

16:46

cares if you go down to a 50% margin?

16:48

it's still really good. So there's so

16:49

much that's the benefit of companies

16:51

with huge pricing power with huge PP

16:53

because there's so much money you can

16:55

make.

16:57

You know, you look at like uh Six Flags

17:00

for example, we were doing a deep dive

17:01

fundamental analysis on this. You know,

17:03

they have like an eight or 9% margin.

17:06

There's no room for you to cut prices

17:08

because you have no margin. You have no

17:10

pee. You have small pee pee. Uh, so you

17:14

know, here if you're paying $100 to $300

17:15

an hour and your cost is $1 to $3, you

17:17

have massive room to cut prices.

17:19

Exactly. But that all cuts into labor,

17:21

too. You know, like the attorney today

17:24

who could do so much more work because

17:26

they could hire so much more staff. It's

17:28

kind of impressive. The question is how

17:29

much work is there for that person? Uh,

17:32

so it's uh it's it's kind of a weird

17:34

time. We'll put it that way. But I think

17:36

this uh Amazon layoff is a little bit of

17:39

a warning. And so it's not that it's,

17:41

you know, a fraction of their 1.5

17:43

million people that work there. You what

17:45

is that 2% or whatever. It's that it's

17:46

like 10 to 15% of their corporate

17:48

workforce fired because of AI. And

17:52

that's the corporate workforce. They're

17:53

the ones who get the stock comp. So it's

17:55

actually less dilutive to Amazon stock,

17:58

right? Because usually it's the

17:59

corporate workers that get the stock

18:00

comp, not the fulfillment drivers,

18:02

right? So you're paying less out in

18:03

stock comp. Kind of scary. Kind of kind

18:06

of crazy. Uh so uh argument is not that

18:12

you shouldn't use AI summaries. The

18:15

argument is when you use AI summaries,

18:17

you are much more likely to forget what

18:19

you read. So sometimes what I'll do is

18:22

I'll read a book or read a passage. I'll

18:25

I'll read a whole thing and then

18:27

afterwards I'll do an AI summary, right?

18:31

And then I'll compare. I rarely do this.

18:34

Rarely, but sometimes I'll do. uh like

18:35

we can even do it right now on uh on

18:38

this u you know interview here summarize

18:40

okay so we'll do it together here so key

18:43

points enterprise AI struggles are

18:45

largely due to in-house staff lacking

18:48

qualifications to execute AI projects

18:50

margins for AI projects are expected to

18:52

improve both algorithmic in efficiency

18:55

and cheaper hardware will drive cost

18:56

reductions pricing for AI inputs will

18:58

rise uh when people derive greater value

19:01

circular financing is not inherently

19:03

alarming right and he makes the analogy

19:05

regarding GM, right? Uh, okay. So, like

19:09

that's great, but if I just read that

19:11

quick little summary, I might forget all

19:12

of it. But see, now if I put my context

19:15

hat on, what do I remember? Or like what

19:17

do I think of if somebody asked me about

19:18

this article? Well, I think about his

19:20

reference to which is from the actual

19:23

context of the interview because this

19:25

was an interview. Hey, is Nvidia

19:28

financing OpenAI deals bad? Well, he

19:31

gives a really good point. He says,

19:33

'Well, it's kind of like GM financing

19:35

their cars, isn't it? Their customer is

19:37

able to buy more car or more chips

19:40

because of financing. And I'm like,

19:41

well, I can't really dispute that.

19:43

That's actually a really good point and

19:45

I appreciate that. And so now I remember

19:48

that that log that codes into my brain

19:51

neural net more. I already forgot the AI

19:54

summary because I remember that context

19:56

way more than the scenario. Uh, circular

19:59

financing, not a big deal. That context

20:04

ingrains in my head this idea about

20:06

in-house staff lacking qualifications

20:08

that doesn't I didn't really get

20:09

anything out of that article regarding

20:11

this right but what I did help me was

20:15

thinking about Karpathy's post here

20:18

about agency it's the decision makers

20:21

who utilize AI who benefit from AI ah I

20:25

could see that okay so then I try to put

20:27

myself in those shoes so if two years

20:30

ago I had to hide hire two or three

20:32

people to help me go through, you know,

20:35

earnings calls to find really specific

20:37

niche things. I can now use AI to give

20:39

me a head start, then I can personally

20:42

go look through it. And I've basically

20:44

now replaced the human pre-ressearcher

20:47

with AI. Then I can go read it myself

20:50

and get my own context. But now I'm

20:52

spending my time reading where it's

20:55

potentially going to be most valuable

20:56

rather than, you know, going on a goose

20:58

chase, right? So uh

21:01

those sort of contexts I think get lost

21:05

substantially when we just use summarize

21:07

tools. So it's I I think it's how you

21:10

deploy it, right? If you uh summarize to

21:14

find where you want to spend your time

21:16

and then you get your context, that's

21:17

probably good. If you get your context

21:19

and then summarize to help you remember

21:21

some of the points or see if you missed

21:23

anything, that's probably good. If

21:25

you're only going around summarizing

21:26

everything and hoping that AI agents are

21:29

going to run your life for you, probably

21:30

gonna lose.

21:32

>> Why not advertise these [music] things

21:33

that you told us here? I feel like

21:34

nobody else knows about this.

21:35

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

21:37

see how it goes.

21:38

>> Congratulations, man. You have done so

21:39

much. People love you. People look up to

21:41

you.

21:41

>> Kevin Praath there, financial analyst

21:43

and YouTuber. Meet Kevin. Always great

21:45

to get your [music] 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.