HomeVideos

Jack Dorsey Bets Big on AI Bubble Pop, 4,000 Workers Lose Jobs

Now Playing

Jack Dorsey Bets Big on AI Bubble Pop, 4,000 Workers Lose Jobs

Transcript

440 segments

0:01

Jarvis, immediately lay off 4,000 of my

0:04

employees. Let's do it by tweet and

0:06

let's make sure that that tweet is all

0:08

lowercase letters.

0:13

Jack Dorsey just laid off 4,000

0:15

employees from his company Block. He

0:18

framed it as AIdriven efficiency in the

0:21

markets response. That stock ripped up

0:24

almost 25% immediately after the layoff

0:27

announcement. I got the stock ticker up

0:30

here and man, that thing is still

0:33

ripping. It's cooled a little bit, but

0:35

I'm still seeing up about 18% as of open

0:38

business this morning. We're going to

0:40

talk about how this underscores the

0:42

Catrini hypothesis that's going around

0:45

right now, why they did the layoffs and

0:47

what that looks like. That's going to be

0:48

my perspective as somebody who's been an

0:51

engineering director and been in the

0:52

room where these decisions get made. And

0:54

then we're going to talk about what it

0:55

means for you and what's coming down the

0:57

pipeline as part of this AI bubble pop.

0:59

This is the cleanest, most real world

1:01

test we have of a concept going around

1:04

called phantom GDP. So what does phantom

1:07

GDP mean? I promise it's not a tricky

1:09

concept. You already understand it. You

1:11

probably just haven't heard the buzzword

1:13

term for it. What that means is when AI

1:16

massively improves worker productivity,

1:19

it improves company output, uh, value,

1:22

the type of products they're able to

1:23

provide, but it doesn't translate into

1:26

earnings for the employees. It doesn't

1:29

translate to increased wages and

1:31

prosperity for the people in that

1:33

society. If you want a really boiled

1:35

down way to think of it, it's when the

1:38

economy becomes unlin from reality as a

1:41

result of the AI productivity gains that

1:44

we're getting. What is really stunning

1:46

about this layoff in particular, we have

1:48

not really seen this yet. Most of the

1:51

time we see a layoff, it's phrased as

1:54

innovation or AI tooling is making it,

1:57

you know, less essential to have

1:58

employees. And what we'll see with a lot

2:00

of those companies, uh, we certainly see

2:02

this with American Express, we see it

2:04

with Capital 1, we see it with a lot of

2:06

other companies as well, is they they

2:08

prep with this innovation sort of

2:10

conditioning in their PR speak in the

2:12

releases and the earnings calls. They

2:15

lay off American employees and they do

2:17

one of two things, usually both things.

2:19

They staff out their domestic operations

2:21

with more H-1B visa recipients so they

2:24

can treat them like indentured servants

2:25

and just give them in general a horrible

2:28

time. uh the people coming here uh get

2:31

the raw end of the deal on that. Um it's

2:33

really unfortunate. I've talked about it

2:35

in other other videos. Um really crappy

2:37

thing to do on the part of the

2:38

corporations to treat people that badly.

2:41

And then the other thing, this is big,

2:42

is you know every quarter you're seeing

2:44

another announcement of we're building a

2:46

new campus in Hyereabad, in Bangalore,

2:50

over in India, over in other lowcost

2:52

centers in Southeast Asia as well.

2:56

you're starting to see a lot of those

2:57

crop up. And that's like I said, I've

2:59

I've literally been in the room where

3:00

these conversations are had at American

3:02

Express where the thinking is we can

3:05

just replace one American developer with

3:07

like five developers in India. It's

3:10

lower priced and we can do that and

3:12

maintain the same level of product

3:14

quality which as you're seeing across

3:16

the board, this isn't a dig against

3:18

India it. This is like a communication

3:21

gap. It happens when you fragment

3:23

people. It's a from Phoenix at least,

3:25

it's a 12 1/2 hour time zone difference,

3:28

which makes communication

3:29

extraordinarily difficult with those

3:31

teams. So, you just you get what you

3:33

might expect. You get what you might

3:35

expect. But more often than not, if

3:37

there's an AI layoff, these companies

3:39

are just using it as an excuse to

3:41

offshore and to import foreign labor.

3:44

And it's a continued concerted effort of

3:46

the executive class to drive down the

3:49

wages of the American employee. Very

3:52

interesting to note that that is not

3:54

what I see going on with the block

3:57

layoff. Jack Dorsey in his layoff tweet

4:00

talks about the regular talking points

4:02

AI tools are increasing efficiency.

4:03

We're going to get into that tweet in a

4:04

minute, but what I do not see from the

4:07

numbers, a lot of people have been,

4:08

you've been sending me messages on X.

4:10

Uh, by the way, if you don't follow me

4:11

over there, I'm dr_jc

4:14

Simmons on X. A lot of people are

4:16

sending me messages. They're tweeting at

4:18

me. They're saying, "Dr. J, you got to

4:20

cover, you know, the block layoffs. I've

4:21

been getting texts yesterday night

4:24

saying, you hey, how how does this

4:25

translate into offshoring? And I'm here

4:27

to tell you, I I did an early dive into

4:30

the numbers uh last evening and I have

4:33

not found evidence to indicate that this

4:35

is going on at Block. I simply I don't

4:38

see it in the numbers. If you have proof

4:40

of that or if you think, you know, I'm

4:42

getting it wrong, post some evidence

4:44

there in the comments because like I

4:45

said, I I couldn't find it. no large

4:47

footprint in India and Southeast Asia.

4:50

Uh more than usual. Um no mass

4:52

importation of H1B visas. They're

4:54

sponsoring a few, but on average

4:56

compared to the rest of the industry,

4:58

it's quite low. So I don't really see

5:00

that as a valid thesis. It's it's really

5:02

important we don't get carried away with

5:04

our preconceived notions and apply that

5:05

to everything. We got to stay looking at

5:07

the data. That's what we do on this

5:09

channel. So this seems like a real case

5:12

where Jack Dorsey lays off these people

5:14

because of AI efficiency. And the reason

5:16

is AI efficiency. The tweet is extremely

5:19

verbose, but we're going to look at just

5:22

one section here that's relevant. He

5:24

says, "We're not making this decision

5:26

because we're in trouble. Our business

5:28

is strong. Gross profit continues to

5:30

grow. We continue to serve more and more

5:32

customers and profitability is

5:33

improving. But something has changed.

5:37

We're already seeing that the

5:39

intelligence tools we're creating and

5:41

using paired with smaller and flatter

5:43

teams are enabling a new way of working

5:45

which fundamentally changes what it

5:48

means to build and run a company and

5:50

that's accelerating rapidly. My heart

5:53

goes out to you if you were one of the

5:54

affected in the layoffs. He goes on to

5:57

say that his two main options were to

5:59

cut gradually over the next few years

6:00

which is the tack that American Express

6:02

is taking. That's why you never see in

6:04

the news anything about their layoffs is

6:06

they just quietly squeeze people out

6:08

through performance improvement plans.

6:10

They don't backfill them. They fire one

6:12

off. You know, a small department goes,

6:14

but it's always under the amount where

6:16

they have to report a layoff. There's a

6:17

lot of other companies doing that. So, I

6:19

do respect Block for the fact that they

6:21

at least just made the cut all at once.

6:23

If that's the business decision and the

6:26

differences between, you know, slow

6:27

dripping that to avoid a news cycle

6:29

versus just saying like we we effed up,

6:31

we overhired, things are different now,

6:33

AI tools are making it so we don't need

6:35

so many engineers. I hate the layoff,

6:37

but I do respect that. I think

6:39

ultimately that's going to be better for

6:41

morale, which is what he hypothesizes in

6:43

the tweet. And again, I can tell you

6:46

from American Express, I I still talk to

6:48

a bunch of employees there all across

6:50

the company and morale just gradually

6:53

decreases when you do these silent

6:54

layoffs. All of a sudden, you know, this

6:56

department got cut, but hey guys, don't

6:58

worry about it. It's just a it's a

6:59

one-off. You know, this is business as

7:01

usual. Um, don't even worry about it.

7:03

Morale just slips. And eventually your

7:05

engineers, all of your talent at the

7:07

company starts to think, you know, if

7:09

they're going to do these kind of

7:11

rolling layoffs every year. There's not

7:12

just one big slash and it's done. If

7:14

they're going to do this every year, why

7:16

why am I working hard? Like, why am I

7:18

not just doing the bare minimum to get

7:20

by, applying for other jobs on the side,

7:22

trying to get a side hustle going? It's

7:24

not immediately obvious to me why I'm

7:27

going to why I should try it all,

7:28

honestly. Back to Phantom GDP. I want to

7:31

talk about that a little bit. This

7:33

hypothesis comes from Catrini research.

7:35

They coined the term. They talk about it

7:37

and it has another side that I didn't

7:39

get into earlier. AI boosts the

7:42

productivity and the profits. But at the

7:44

same time, the displaced workers like

7:46

these people laid off from block spend

7:49

less because they're not making any

7:51

money. If you were a software engineer

7:53

at block, we have the comp here from

7:55

levels.fyi.

7:57

I'm going to say senior is probably

7:59

level five here. So, your base salary

8:01

203,

8:03

136,000 uh in stock every year. Uh which

8:07

man, if you're able to hold on to that,

8:08

that's ripping. So, you got a 25% boon

8:11

on that stock since the layoff

8:12

announcements. Uh and and it looks like

8:14

they don't have a bonus structure set

8:15

up. So, that's 339K allin. Um

8:19

engineering jobs are just we're not

8:21

seeing those compensation packages right

8:23

now, at least for for companies that are

8:25

not uh hyperfocused in the Bay Area. I'm

8:28

sure of a lot of very talented people

8:30

got laid off. If they got into block,

8:31

it's very, very hard interview process.

8:33

They got laid off, they get another job,

8:35

they're making maybe 160k a year, maybe

8:37

200. You know, that's a 33% to 50%

8:41

haircut. They're not going to be, you

8:43

know, they're they're not going to be

8:44

frivolously spending on anything that

8:46

they might have been spending on while

8:47

they were working at Block. So, it

8:50

really fs the economy up in a number of

8:53

ways. And nobody has a great answer for

8:55

this yet. I've talked a little bit

8:57

before on the channel about, you know,

8:58

maybe universal basic income. Don't get

9:01

me wrong, I am an unrepentant

9:03

capitalist. I I very much disagree with

9:05

and don't like socialism, but it's the

9:08

best proposed alternative that I've

9:10

heard. Again, in the comments, if you

9:12

are like, I got the answer to this, uh,

9:15

this is what we should do with all of

9:17

these displaced workers due to AI

9:18

efficiency, po post that in the

9:20

comments. I would love to hear it

9:21

because right now the best argument I've

9:23

heard is is you we need some kind of uh

9:26

universal basic income to replace this

9:28

which has a ton of downsides. So why is

9:31

Wall Street rewarding this? Why did the

9:34

stock just shoot to the moon? Well, it's

9:37

quick gains, right? You and I know it

9:40

takes a lot of time and a lot of effort

9:43

to maintain a massive codebase, a lot of

9:45

software. But in the short term, these

9:48

products are already built. they already

9:50

exist. And so if you say, "Hey, we're

9:53

just not going to add any new features,

9:55

no new features to the cache app, we're

9:58

not going to make any changes to Square

10:00

or the cash app." Uh, going forward, all

10:02

of the remaining engineers, they are

10:04

going to shift to just maintenance mode

10:06

on keeping this thing alive. Guess what?

10:08

That's still generating a ton of money,

10:11

and your operational expenditures have

10:13

just cut dramatically. So, I did some

10:16

really rudimentary math. If I am Jack

10:20

Dorsey and I lay off 4,000 employees,

10:24

like I said, that senior engineer

10:25

salary, total comp package, we're

10:26

looking at about 300 350K. Let's be

10:29

really conservative and let's say, you

10:31

know, they get a lot of people that are

10:32

making more towards the median for the

10:34

company that are maybe non-technical.

10:36

That's let's say a $200,000 total

10:39

compensation package is like kind of

10:41

average, right? Uh if I lay off 4,000,

10:44

that's 800 million annually saved. So

10:46

that you're saving instantly today

10:49

almost a billion dollar in operating

10:52

expenses. And that's not even accounting

10:55

for taxes, benefits, anything like that.

10:59

So in the short term, financially, any

11:02

massive layoff like this, at least to

11:04

Wall Street, looks like a huge win

11:06

because you've just saved almost a

11:09

billion dollars a year. Now, this gets

11:10

really interesting because you might

11:12

say, "But Dr. J to maintain their

11:14

competitive edge. To maintain their

11:16

competitive edge, apps like Square and

11:20

the Cash App need to be cranking out new

11:22

features. They need to be competing with

11:24

Stripe. Um Dr. J on his show gets those

11:27

two confused all the time. They need to

11:30

differentiate theel themselves obviously

11:32

and ordinarily you'd be right. But if we

11:34

look at the strat treaty hypothesis,

11:36

remember the second part of that is that

11:38

the people that are displaced, the

11:40

people that were earning 300 350k a year

11:43

that are now earning 150 200k a year,

11:46

they're not spending as much money. So

11:49

they're not going to use those apps as

11:52

much. It's fintech and fintech is

11:54

downstream of consumer spending. and

11:57

you've just nerfed 4,000 high earners

12:00

and made them take massive pay cuts or

12:03

potentially just go on unemployment for

12:04

an indefinite amount of time. That is

12:06

the dark underside of this. So, what

12:08

does this mean for workers, builders,

12:10

and investors through the rest of 2026

12:13

and moving forward? the people that are

12:15

going to be safest while this is

12:18

unsettled, let's say, while we're in

12:20

transition to whatever is next. Anything

12:22

that is high risk, uh, linked to

12:25

something with a lot of regulation, that

12:28

work is going to be fairly safe. It's

12:30

very difficult to replace that with AI.

12:32

That's going to be, uh, KYC, like know

12:34

your customer, when you have to provide

12:36

your ID and stuff when signing up for a

12:38

new um, like Cash App account or

12:40

something like that. It's going to be

12:42

risk and fraud detection. So like the

12:45

department that can detect when your

12:47

card it city or whatever is being used

12:51

in a weird way and they flag it. It's

12:53

also going to be risk modeling. What's

12:55

the attack surface? How do we minimize

12:57

that with our application? Mission

12:58

critical hardware that is going to be in

13:01

life situations. That's going to be very

13:03

high stakes. I talked about that a

13:05

little bit with a manufacturing company.

13:08

Let me know if you guys want to see more

13:09

defense tech videos. I'd love to do

13:11

that. It just didn't really get views

13:12

last time I did it. But happy to cover

13:14

it because for my money, it's the safest

13:16

place to be working right now and it's

13:18

the safest place to be investing your

13:19

money if you're trying to park your

13:20

money somewhere. Builders, your your

13:23

operator class, they're going to need to

13:26

focus on how AI directly improves

13:29

outputs. This is going to need to be

13:30

measurable as opposed to AI

13:32

transformation. In 2026, AI

13:34

transformation without figures and facts

13:37

behind it is done. You cannot just say

13:39

we're adding AI and stock goes up. You

13:41

need to say we added this AI and it

13:44

improved this metric significantly

13:46

because investors are starting to say

13:47

where is my payday on this? Where is my

13:50

return on investment? So bottom line, I

13:52

think block is relatively early in this

13:54

trend, but this is something we're going

13:55

to continue to see through 2026,

13:57

probably ramping up over the near

13:59

future. As we've seen with these

14:01

layoffs, there's usually a shot across

14:03

the bow. somebody goes early, there's a

14:05

bunch of copycats because they want that

14:07

24% stock velocity so they can pump the

14:09

stock, especially if they're an

14:10

executive. They can pump the stock and

14:12

then they can get the hell out of there

14:13

with their nice stock package and do a

14:15

big sell-off and profit from it and then

14:17

say like, "Peace, I don't care about the

14:19

long-term ramifications of this." But

14:20

we're going to see this a lot,

14:21

especially with like new tech fintech

14:23

companies. So again, banks are going to

14:25

be a little bit more insulated, but

14:26

you're going to look at your Square,

14:28

your Stripe, Revolute is another one

14:30

that's kind of up and coming. You're

14:32

going to see some turmoil there in the

14:34

interim places like CLA, Afterpay, any

14:37

of the products like that. They're

14:39

probably going to do decent business

14:40

because people aren't going to just

14:41

change their lifestyle when they get

14:42

laid off all of a sudden. They're going

14:44

to want to maintain that at least for a

14:45

little while. And in order to maintain

14:47

that, they're going to need payment

14:48

plans. So, it's it's very dark. But

14:50

remember, I was in charge of line and

14:52

lending products at American Express.

14:54

So, I know what I'm talking about. The

14:55

projections, I think, personal opinion,

14:58

no company data. I think that space is

15:00

going to be hot for a little while. And

15:01

then it's going to cool as uh it hits

15:05

the fan, let's say. But here's your case

15:07

study for AI era labor redistribution.

15:10

Like it is it is happening. It is

15:12

happening. Nobody knows what the future

15:13

is going to be like. But this is

15:15

concerning, especially on the send side,

15:17

the layoffs. It's only 4,000 people, but

15:20

how much money were they spending on

15:21

dumb hobbies, you know, and they're not

15:23

spending on those dumb hobbies? And you

15:25

may think that's fine, but that's less

15:26

money moving through the economy. That's

15:29

less companies that need to exist. So,

15:31

it really just creates this uh seize of

15:34

the whole process. Everything kind of

15:37

jams up from every side. We don't know

15:39

where it's going to land exactly, but

15:41

we'll keep covering it on this channel.

15:42

Thank you so much for watching. If you

15:44

like the hat, that's my buddy's company,

15:45

Abandoned Wear. Link down there.

15:47

Subscribe, hit the bell to be notified,

15:49

and join my newsletter in the

15:51

description.

Interactive Summary

Loading summary...