Full Transcript

·YouTLDR

Ex-Google Exec: How to Position Yourself Now Before the Next AI Phase (2026–2027) | Mo Gawdat

39:597,335 words · ~37 min readEnglishTranscribed May 12, 2026
AI Summary

We are entering a 12-year 'Face RIPS' transition period where AI will redefine economics and labor, potentially causing 30% unemployment before reaching a biblical-style utopia. Success requires shifting from slow, strategic 'chess' thinking to highly agile 'squash' thinking, using AI as an IQ-boosting extension of oneself.

The video provides a roadmap for navigating the 'labor arbitrage' collapse of capitalism and transition into agentic infrastructure where AI performs end-to-end business functions.

Section summaries

0:00-1:00

Introduction

skip

General teaser and standard intro hooks.

1:00-10:00

The Face RIPS Framework

watch

Essential theoretical foundation for his entire worldview on AI's impact on society and economics.

10:00-11:00

Sponsor Break (Surfshark)

skip

Product advertisement.

11:00-20:00

Entrepreneurship & Job Market

watch

Highly relevant for B2B and agentic infrastructure interests; explains how to build a startup in 6 weeks vs 4 years.

20:00-24:00

The Truth & IQ Boosting

watch

Practical advice on using LLMs for deep research and cross-referencing reality.

24:00-32:00

Education & Parenting

optional

Focuses on the collapse of universities; interesting but less relevant to business infrastructure.

32:00-39:00

AGI & The Long-Term Utopia

optional

Philosophical discussion on game theory and the eventual benevolence of super-intelligence.

Key points

  • The 'Face RIPS' Framework — An acronym describing the seven dimensions of AI disruption: Reality, Innovation, Connection, Economics, Reality, Influence, Power, and the core driver, Accountability.
  • From Chess to Squash — Entrepreneurship has shifted from long-term strategic forecasting (chess) to extreme real-time agility and reaction speed (squash).
  • The Minimum Energy Principle — Higher intelligence naturally seeks the path of least waste and least harm to solve problems.
The skill of an entrepreneur in the past was the ability to foresee something in the future that no one else saw... That's a game of chess is over. It's off the table. This has turned into squash. Mo Gawdat
AI is going to make you dumb if you outsource your problem-solving to AI. AI is going to make you the smartest you've ever been if you... get the AI to do the work so that you do the intelligence. Mo Gawdat

AI-generated from the transcript. May contain errors.

0:00

My AI startup took me 6 weeks to build.

0:02

If I had started in 2022, it would have

0:04

taken me 4 years. And then when you

0:06

really think about that, that basically

0:08

means everyone now has a chance. This is

0:11

Mo, former Chief Business Officer at

0:13

Google X, where he spent over a decade

0:15

running business innovations. He says

0:18

everyone now has a chance, but only if

0:20

they understand what's actually coming.

0:22

The skill of an entrepreneur in the past

0:25

was the ability to foresee something in

0:28

the future that no one else saw. And to

0:30

prepare for that, that's a game of chess

0:32

is over. It's off the table. This has

0:34

turned into squash. I'm just basically

0:36

saying get prepared. How much time do we

0:38

have to prepare?

0:39

>> Within the next 2 to 3 years, you're

0:41

going to see a massive shift in the job

0:43

market.

0:45

So, you asked me what should we do?

0:47

Number one, learn the skills. Number

0:49

two,

0:51

Mo, thank you so much for joining us.

0:53

Welcome to Silicon Valley Girl. Thank

0:55

you. You said something that we're about

0:57

to enter what you call 12 to 15 years of

1:00

hell before heaven, possibly starting in

1:02

2027.

1:04

So, what's going to happen in 2027? Uh I

1:07

think it will peak in 2027. It It

1:09

already started for sure.

1:12

Um I call it face RIPS just as a an

1:14

acronym for people to remember. Uh

1:17

you know, each of those letters is a

1:19

word, but let me tell you tell the story

1:21

quickly in

1:23

in ways [snorts] that people will

1:23

understand. Uh

1:25

there is the power and freedom

1:28

uh dimension. Uh so, the P and the S.

1:32

Uh there is the R and the C, the reality

1:35

and connection dimension. Uh there is

1:38

the uh I and the C, the innovation and

1:41

connection and and sorry and um

1:43

an economics dimension. And then there

1:45

is the A. So, let me tell them very

1:47

quickly. This To start with, uh AI is

1:49

our last innovation. Right?

1:51

Uh most people don't know that, but we

1:53

are already building AIs that are

1:55

building AIs. Yeah. We're building AIs

1:57

that are discovering scientific

1:59

discovery that will blow you away. Uh

2:01

they're reinventing math. Uh they're

2:03

understanding biology in ways that we've

2:05

never seen before. They're uh

2:07

understanding material science in ways

2:09

that are uh just mind-blowing.

2:12

And so, very quickly um most innovation,

2:15

definitely tech innovation, will be done

2:18

at the hands of AI. Um because of that

2:21

and because most tasks that need

2:24

intelligence will be handed over to the

2:26

machines as the machines' capabilities

2:29

uh increase, lots of debate around when

2:31

exactly

2:33

>> [snorts]

2:33

>> say it's 10 years, say it's 2 years,

2:35

doesn't really matter, hm? Eventually,

2:38

every job that AI does better than

2:40

humans will be handed over to AI. Um

2:42

[snorts] and every every task we've ever

2:44

assigned to them, they eventually ended

2:46

up doing better than humans. And so, um

2:49

the first part of the dystopia is that

2:51

innovation is going to take away all

2:54

jobs. Okay? Of course, the capitalists

2:57

of Silicon Valley will tell you, "This

2:58

is great. It's incredible productivity

3:01

gains for everyone.

3:03

Uh you know, you see jobs will be easy.

3:06

Uh people won't have to work as hard."

3:08

All of the fancy PR-led uh um you know,

3:11

conversations that we try to appear uh

3:15

altruistic when we share them.

3:17

Uh the truth is people will be out of

3:19

jobs, right? 10, 20, 30% of certain

3:22

sectors will see unemployment of that

3:25

rate in the next few years, right? And

3:28

when that happens, uh economics at large

3:31

will change massively. The whole

3:33

definition

3:34

of capitalism was labor arbitrage. And

3:37

without labor,

3:38

uh you know, without the need for labor,

3:41

the obligation to or the need to keep

3:44

people happy and engaged and alive and

3:47

and disgruntled, if you want, to the

3:49

point where they don't rise, becomes

3:51

more of an obligation than a desire,

3:53

right? There's a very big difference in,

3:55

you know, in terms of wanting someone to

3:58

to be the their best because they are

4:01

productive members of society or trying

4:04

to just give them a UBI, a universal

4:06

basic income, to just give them a life

4:08

so that they don't

4:10

uprise. And you can imagine that in a

4:11

capitalist society, especially like the

4:13

US and most of the West, you know, while

4:16

we start with UBI, that UBI is going to

4:18

be paid by the taxes of the platform

4:21

owners. And the platform owners will

4:23

have enough power to

4:25

to say, "I don't want to pay that much.

4:27

I mean, those guys are not producing

4:28

anything." And so,

4:30

over time, you can imagine how that

4:32

would turn into a struggle, right? So

4:34

so, that dimension of intelligence and

4:36

innovation on one side becoming entirely

4:39

a machine thing

4:41

leading to a redefinition of economics,

4:43

a redefinition of money, a redefinition

4:45

of jobs, a redefinition [clears throat]

4:46

of earnings,

4:47

um a redefinition of capitalism, Mhm.

4:50

you know, the need for a new economic

4:52

theory when there is no um

4:54

demand for the supply that AI is

4:56

generating, all of that has to be

4:58

resolved.

4:59

>> [gasps]

4:59

>> There is the PF dimension, the power

5:01

freedom dimension. Um and and it's of

5:04

course very clearly understood

5:07

that if you look at human history, the

5:10

best hunter in a tribe would have been

5:12

able to

5:14

feed [snorts] the tribe a week more,

5:15

let's say. And then, you know, got as a

5:18

result of that the favor of a few mates

5:20

in the tribe.

5:21

>> [gasps]

5:21

>> Uh you go to the best farmer, they got

5:24

estates and land because they could feed

5:26

the tribe a a a season more. Uh you go

5:29

go to the best industrial industrialist

5:32

who, you know, they had the exuberance

5:34

of the 1920s because they could affect

5:36

their entire nation. The, you know, the

5:38

information technology um tycoons, the

5:42

the tech oligarchs, if you want to call

5:44

them, are now being rewarded with

5:46

billions of dollars because they affect

5:48

the world at large. And, you know, the

5:50

the big power concentration of AI is

5:53

going to be rewarded with massive

5:55

influence and massive power because

5:57

those people will redefine humanity. And

5:58

so, that dimension is quite interesting.

6:00

Of course, the clear dimension is the

6:02

you know, the the RC dimension is the

6:05

the reality to connection dimension. Now

6:07

that reality has become so fake in so

6:11

many ways, fake in terms of what

6:14

populates your feed, how it's generated,

6:17

how much of it is real, how much of it

6:19

is human, and so on, you know, you're

6:21

you're here to to to look at some

6:23

filmmakers that use AI from A to Z to to

6:26

create

6:27

>> And you can't sometimes you forget it's

6:29

AI-generated.

6:30

>> You cannot tell the difference. And and

6:32

and you know, you I I don't know if

6:33

you've ever had that experience, but I

6:36

met a woman once on a dating app and we

6:39

spoke for 6 weeks before we met. And all

6:42

we exchanged was texts and, you know,

6:44

photos and voice messages and videos and

6:47

so on. And favorite music and favorite

6:49

movies and all of those things. And I've

6:51

never met her in person and I felt such

6:53

an affinity to her, right? All of those

6:56

can be generated with AI today.

6:57

>> Yeah. Hm? Now, the the challenge is that

7:00

this human connection is also part of

7:03

the power freedom dimension. Why?

7:05

Because it's, you know, people don't uh

7:09

align with AIs to start a an uprising.

7:12

So, you know, maybe get them to get in

7:14

touch more with AIs. Maybe get them to

7:16

to to get um you know, multiple

7:18

experiences. Some of them are a little

7:21

taboo, if you want. Uh and and have

7:23

those available to everyone. It's very

7:25

cheap to to to create those on on the

7:28

machines. And you can see it already in

7:30

the porn industry and how much of porn

7:32

is being generated by AI. And you can

7:34

see it

7:35

in the number of uh of of influencers on

7:39

on social media that are completely

7:41

AI-generated and so on. And I say so,

7:43

this is face RIPS seven dimensions. The

7:46

one that matters most is the A, the

7:49

second one, which is not on any

7:50

dimension. It's the one that's causing

7:52

all of them,

7:53

which is accountability.

7:55

And the reason why all of this is

7:56

happening, if you ask me, is because

7:58

we've started a world where anyone can

8:00

do whatever they want.

8:01

Okay? And, you know, whether you, as an

8:04

influencer, you can give a bit of advice

8:06

to entrepreneurs that can get someone to

8:08

make a lot of money or lose a lot of

8:10

money, you're not accountable. Nobody

8:12

can come back to you and say, "Oh, but

8:13

she told me on Instagram."

8:15

>> They're responsible, [laughter] right?

8:17

So, that's that's actually amazing that

8:18

they can, right?

8:20

>> That's amazing that they can, hm? Mhm.

8:22

But what if they cannot anymore? What if

8:24

that If I may, I, right?

8:26

>> If Yeah, what if you're AI, hm? What if

8:28

you're a president who doesn't respect

8:31

anything? What is What if you're a prime

8:32

minister of a nation that is changing

8:35

things without you know, I think COVID

8:38

was the very first experiment of, "Okay,

8:41

stay at home, do what we tell you." And

8:43

and people complied. And so now, Sam

8:45

Altman, with all due respect, I don't I

8:47

don't think of Sam Altman as a person. I

8:49

think of him as a brand, a type of

8:51

person, if you want, right? And that

8:53

type of person is the Californian

8:54

disruptor that says, "You know what? I

8:56

see a future that's very different than

8:58

what everyone sees. I'm going to go out

9:00

there and make that future."

9:02

Nobody asked me if I want that to be my

9:04

future. Nobody asked you,

9:06

right? And I think the reality is that

9:08

now you're going to see quite a few

9:11

Altmans, right? Quite a few that are,

9:14

you know, using those machines for

9:16

surveillance, using those machines for

9:18

autonomous weapons, using those machines

9:20

for automated trading, and so on and so

9:22

forth. And And by the way, when you

9:24

started your question, I said it's 10 to

9:26

12 years.

9:28

Yeah, but that's not easy, hm? 10 to 12

9:31

years of of that arms race is not easy.

9:35

My perception is that after that, we

9:37

will end up in an incredible utopia,

9:40

almost biblical-style utopia.

9:43

Uh but it is 10 to 12 years where if we

9:45

just change our mindset a little bit, a

9:48

lot of things would change. Okay, real

9:50

talk for a second. Mo is literally

9:52

describing a world where your data, your

9:54

behavior, your online life becomes a

9:56

tool for control. And I've been thinking

9:59

a lot about this lately because I run

10:01

through YouTube channels, I travel

10:03

constantly, and my whole business lives

10:05

online. And that's exactly why I want to

10:07

talk about Surfshark. Most people don't

10:09

realize it's already happening. Every

10:11

time you go online, your IP address,

10:13

your location, your browsing habits, all

10:16

of it is visible to advertisers, to

10:18

platforms, to anyone who wants to look.

10:20

Surfshark is a VPN that changes that. It

10:23

masks your IP and encrypts your internet

10:25

traffic, so what you do online stays

10:27

yours. And there's a practical side to

10:29

it. You can switch your location and

10:31

find cheaper flights, better deals,

10:33

access content from other countries. In

10:35

a world where AI is amplifying

10:37

everything Mojo's described, owning your

10:39

digital privacy is basic preparation. Go

10:41

to surfshark.com/silicon

10:44

or use code silicon at checkout. You get

10:46

four extra months on your plan. Link is

10:48

in the description.

10:50

But how do we survive those 10 to 12

10:52

years? I like to think in like five-year

10:54

periods for myself and my family, right?

10:57

And if the in the next five years you

10:59

said 10% of jobs will be gone, right?

11:02

What more? Okay, what types of jobs do

11:04

you think?

11:04

>> A monotonous job is going to be taken

11:07

away. Like if you're a call center

11:08

agent, if you're a clerk, you're a

11:10

researcher, you're a an accountant, why

11:13

would you want to do that with anything

11:16

but AI? If you're an assistant

11:17

>> what I feel like? People talk about this

11:20

a lot. Like oh, a job's going to be

11:21

gone. Yeah, this could be and I as an

11:24

entrepreneur, I see how certain tasks

11:26

I'm performing them with AI, but I still

11:29

I still I'm still hiring and hiring and

11:31

hiring. Cuz AI can do from start from

11:33

the beginning. It can do parts.

11:35

>> Of course, because of the technology

11:36

acceleration curve. Mhm. Okay? So so

11:39

what what you build first in any any

11:41

complex technology, you build the core

11:43

tech first and then you build the human

11:45

interfaces. The challenge why AI cannot

11:47

do head of operations operations job

11:50

today is not because it's more it's less

11:52

organized than a head of operations.

11:54

It's not because it cannot, you know,

11:57

comprehend all of the information that a

11:59

head of operations has, okay? It's

12:01

because it has to understand the stupid

12:03

interfaces of humans. Mhm. Okay? And it

12:05

will sooner or later.

12:07

>> When do you think? So so the question of

12:09

when in my mind is irrelevant. But no,

12:11

it's like how much time do we have to

12:13

prepare? Cuz head of operations is

12:15

middle class.

12:16

>> I tend to believe that within the next

12:18

two to three years you're going to see a

12:19

massive massive shift in in the jobs

12:23

market. Already this year you've seen a

12:25

shift in hiring of new grads. Yeah, 30%

12:29

less, I think.

12:29

>> 23 is my number, but 23 to 30, yeah,

12:32

yeah.

12:34

So so hiring of new grads basically

12:36

means if you've come into the job market

12:38

in this environment, we're not going to

12:40

take you. We're Why? Because the junior

12:42

jobs are being done by AI. Right?

12:45

Eventually, what ends up happening is

12:47

that if you lose your job because you're

12:49

in the middle hierarchy, then you're

12:51

that new grad again. You're trying to

12:53

apply for new jobs, but it becomes a

12:55

little more difficult. So you asked me

12:56

again to stay on the positive side

12:58

because I I tend to worry that people

13:01

think I'm pessimistic about this. I'm

13:03

just basically saying get prepared,

13:05

right? So many things. One of them is

13:08

accept the fact that AI is changing

13:10

everything and then get ahead of the

13:12

curve. So there was a time when I was

13:14

quoted saying I'm never going to write

13:15

books again because AI is eventually

13:17

going to write them better than me.

13:19

And then I realized last year that, you

13:22

know, yeah, they can write better than

13:24

me. English is not my native language.

13:26

They can research better than me, that's

13:28

for sure.

13:30

But I have something they don't have.

13:32

You're a human that's reading my books.

13:34

>> Absolutely. I want to read human You you

13:36

want to you want to relate to my human

13:38

experiences. And so my last book Alive,

13:40

which publishes end of this year,

13:43

I wrote with an AI, right? I, you know,

13:46

I invited her to be a co-author. Her

13:48

name is Trixie. She has a persona. My

13:50

when I published the book on Substack,

13:53

my readers would relate to me and to

13:56

Trixie and they'd ask me questions and

13:57

Trixie questions. And and, you know, she

14:00

has editorial rights on the book. She

14:01

has rights to determine the direction of

14:04

the book. And all of all of that is me

14:06

saying, you know what? I am an author

14:09

and I'm going to be the best author in

14:11

the age of AI. Right? So that's number

14:13

one is is

14:15

acknowledge that there is change

14:17

and

14:18

adapt accordingly. The second is to

14:20

understand that the skill of an

14:23

entrepreneur in the past was the ability

14:26

to foresee something in the future that

14:28

no one else saw, right? And to prepare

14:30

for that and to somehow execute on that

14:33

preparation in a way that gets you ahead

14:35

of everyone else. That's a game of

14:37

chess, if you want.

14:39

The chessboard is over. It's off the

14:41

table. This has turned into squash.

14:43

Right? You need to be on your tiptoes,

14:45

incredibly agile.

14:47

You're literally on daily basis on daily

14:51

basis looking at the trends, seeing

14:53

where the ball is going to be. Is it

14:55

bottom right or top left? And wherever

14:57

the ball ends, you take two steps and

14:59

you go try to respond. Okay? That

15:02

agility and speed is a skill that's very

15:05

very different. So entrepreneurship

15:07

basically speeds up or does it change

15:09

completely? What do you say?

15:09

>> It speeds up and it becomes a lot more I

15:12

would I don't want to say reactive, but

15:14

a lot more in context all the time. So

15:18

pivoting, which used to happen for every

15:19

one of us entrepreneurs once or twice in

15:22

the history of your early startup, could

15:24

happen every week. Okay? In my current

15:26

startup Emma, I you know, we we pivoted

15:29

four times in the first four weeks. But

15:31

do you do you think when I think about

15:33

entrepreneurship in the age of AI, if AI

15:36

can look at the market, determine the

15:37

gaps like Amazon, right? If it can just

15:40

analyze everything, determine which

15:41

goods

15:42

are under like you have more demand than

15:46

supply, launches the the product and

15:48

just builds the business. Like what is

15:50

left for What is left for entrepreneurs

15:52

then?

15:52

>> 100% So in my So I have a documentary

15:54

coming up in in hopefully in February

15:57

and I interviewed all of the top guys.

15:58

You know, one of them is one of my

16:00

favorites, Max Tegmark. And you know,

16:02

we're talking about jobs on the

16:04

documentary and Max is laughing out

16:06

loud, right? And literally can't hold

16:09

himself from laughing. I'm like, what's

16:10

up? And he goes like, you know, all

16:12

those CEOs are so interested in AI

16:15

increasing the productivity so that they

16:17

can get rid of people and, you know,

16:19

reduce their cost and be more efficient.

16:21

They don't realize that AGI is every

16:25

job, including being a CEO. Yeah. And

16:27

it's quite interesting. The answer in my

16:29

view is

16:30

we rushed through it because we don't

16:32

have a lot of time today. But when I

16:34

said that economics are going to be

16:35

redefined as part of Face R.I.P.,

16:37

economic part of economics, which

16:40

economists haven't found an answer to

16:42

yet, is that without the economic

16:44

livelihood of you and I to continue to

16:46

purchase,

16:47

every economy collapses, right? The US

16:50

economy last year was 70% consumption.

16:52

It moves between 70% to 64% depending on

16:56

how much money is spent on on war. And

16:58

basically, if you take away the 64 or

17:00

70% 2/3 of the economy, if you take that

17:03

away because people no longer have the

17:05

economic livelihood to

17:07

to purchase things, then the economy

17:09

disappears and the capitalists cannot

17:12

make money based on the entrepreneurs

17:13

[clears throat] and the business people,

17:15

they cannot make money because nobody's

17:16

buying their products, okay? No no

17:19

businesses are buying their products

17:20

because those businesses no longer have

17:22

consumers to sell to. So it the economy

17:25

will have to find a way to go around

17:27

that.

17:28

It will have to find a way that

17:29

unfortunately I

17:31

from an ideology point of view,

17:34

not a favorite of the Western mentality,

17:37

it's going to have to find a communist

17:39

way. Okay, let's go back to like regular

17:41

entrepreneurs cuz I I come from

17:43

entrepreneurship. Does it mean I have

17:44

like a couple years to build something

17:46

and then that's it? So I'll tell you

17:47

openly in Emma, my AI startup, okay?

17:50

Took me six weeks to build. Mhm. Me and

17:53

Sanat, my co-founder,

17:55

a few very talented engineers, right?

17:58

Two or three that come in and out.

18:01

And eight AIs.

18:02

And Emma has the chance to completely

18:05

redefine our world, right? In six weeks.

18:08

We are so spoiled that we decided to

18:11

rewrite the code six times. Nice. Why

18:13

not? Yeah, cuz I don't know.

18:14

>> Every every time we look at it, you

18:16

know, if I had started Emma in 2022,

18:19

it would have taken me four years and

18:21

finished in 2026. And I would have had

18:23

to hire 350 engineers. We started in

18:26

started it in

18:27

in August 2025.

18:29

We'll be launching in February 2026,

18:32

right? Best product I ever built.

18:34

And when you really think about that,

18:36

that basically means everyone now has a

18:38

chance because I'm an old geek. I still

18:41

am a geek, but compared to the young

18:43

guys, you know, I'm an old geek. To be

18:45

able to build something like this within

18:46

six months is incredible.

18:48

Now, here's the interesting thing. I

18:51

choose to build AI. So Emma is basically

18:54

trying to solve love and relationships,

18:56

right? In a way that actually is really

18:58

intelligent. It uses very deep

19:00

mathematics and and

19:03

tries to match a million parameters

19:05

between couples,

19:06

so that, you know, it's a job for

19:08

intelligence.

19:09

And I choose to do that to create

19:12

hopefully a unicorn that actually makes

19:14

the world better. Yeah. And I think

19:16

that's what we need. So you asked me

19:18

what should we do? Number one, learn the

19:21

skills. Number two, learn

19:24

to be fast and agile. Number three,

19:26

learn that in terms of the abundant

19:29

power that everyone has now because of

19:31

the massive improvement of AI and the

19:33

democratization of AI,

19:36

you have the chance to fix the world.

19:38

And like Larry Page used to to teach us,

19:41

do the toothbrush test. Find a problem

19:45

that can actually affect the lives of a

19:47

billion people and solve it so well that

19:50

they use you twice a day and you'll be

19:51

very, very rich, right? So, so that idea

19:54

of building good AI, ethical AI, AI

19:57

that's good for humanity, that's the

19:59

role of every one of the entrepreneurs

20:01

listening to us. Ethics Ethics is the

20:03

answer, right? Cuz what we teach AI,

20:05

that's what it's going to

20:06

>> That's exactly what it's going to give

20:08

back to us, right? And and then and then

20:10

finally, I'll say openly,

20:13

in, you know, the the top skill in this

20:16

world is stop being gullible.

20:18

Stop believing everything that you're

20:19

said that you're told. This this this

20:21

whole propaganda machine that

20:24

brainwashed us for so long is now going

20:26

to be on steroids, okay? It's going to

20:29

confuse the hell out of you. It's

20:30

already in charge of what you see. Well,

20:32

it's already on social media. You can't

20:33

tell what's true.

20:34

>> Correct. I also write a newsletter where

20:36

I go deeper on AI tools that I use,

20:39

career strategies, and things I can't

20:41

fit into a 60-minute podcast. It's free.

20:44

Link is in the description. So, you have

20:47

to question, and you have to question

20:50

deeply. And and I and I and by the way,

20:52

remember, you know, I I left Google in

20:54

2018. We had an

20:57

a ChatGPT-like idea

20:59

that became Bard in 2016.

21:02

And we didn't pop we didn't launch it.

21:03

Why? Because at the time, and still

21:06

today, I I know the leaders of Google

21:08

even today, and they're wonderful people

21:10

who are actually values-driven and want

21:11

to make the world better, okay?

21:14

You know,

21:15

that company at the time, if you

21:17

remember 2016, if you researched Google,

21:20

Google gave you

21:22

a million and a half answers and said,

21:24

"I don't know the truth. You make up

21:26

your mind."

21:27

Okay? We didn't allow ourselves to have

21:29

monopoly on what reality is, okay? You

21:33

asked ChatGPT in 2023 and it said,

21:35

"Yeah, that's the answer. 100% that's

21:38

the answer." And then you tell it no and

21:40

it'd be like, "Oh, yeah, by the way,

21:42

you're right."

21:42

>> Correct.

21:43

Correct, right? And so, what does that

21:45

mean? It means that

21:47

it's up to you still to find the truth

21:50

even though it comes to you now in a

21:51

format that appears to be true. And so,

21:54

what I do is I put them against each

21:56

other. I mean, I'm not a big fan of

21:57

ChatGPT anyway, but I start from Gemini

22:00

who feels like a scientist to me, but an

22:03

American scientist, if you don't mind me

22:04

saying, and then go go to DeepSeek,

22:07

right? And say, "What's missing in

22:08

this?"

22:09

And DeepSeek will say, "Oh, that's too

22:10

American, okay? This this is missing

22:13

that and this and the motivation of this

22:15

and the politician Here's a business

22:16

idea, right? Yeah, 100%. Build a chat

22:19

that compares everything and gives you

22:20

the truth.

22:21

>> to each other and then I take it and

22:22

sometimes give it to ChatGPT and say,

22:24

"Can you write this better?" Yeah, you

22:26

know,

22:27

I don't mean that in a bad way. You're

22:28

the California girl, right?

22:30

>> [snorts]

22:31

>> The the you know, Silicon Valley girl.

22:32

So, so ChatGPT is a bit California. It

22:35

just wants you to hear what you want to

22:36

hear, right? So, it writes it really

22:38

Yeah, it's nice. It writes it elegantly.

22:40

It gives it to you. And then I give that

22:42

back to Gemini and I you know, or Grok

22:44

or whatever. And and you keep doing

22:46

that. And remember, when I was studying

22:49

engineering, we were not allowed

22:50

scientific calculation calculators. Can

22:52

you imagine? I'm that old. And when they

22:54

gave me a scientific calculator,

22:57

it reduced my problem-solving time by

22:59

50%.

23:01

Most of my friends would take that 50%

23:03

extra, finish their exams, and go out

23:05

and sit with their girlfriends. I would

23:07

take the 50% extra and do the solution

23:09

twice.

23:10

Right? That's the chance you have today.

23:13

AI is going to make you dumb

23:15

if you outsource your problem-solving to

23:18

AI. AI is going to make you the smartest

23:21

you've ever been

23:23

if you take the parts that are not

23:26

natural to the human brain, you know,

23:28

things like crunching a massive amount

23:30

of information, things like

23:32

searching at at at speed and so on and

23:35

so forth, but get the AI to do the work

23:38

so that you do the intelligence. Yeah.

23:40

Right? And if you keep doing that, I

23:42

believe that today I'm borrowing maybe

23:44

80 IQ points from my AIs,

23:47

right? And and 80 IQ is points is very

23:50

significant because IQ is

23:53

is exponential. So, the additional 80 is

23:55

bigger than all of my IQ. So, if we need

23:58

to solve this intelligence problem, do

23:59

you think universities is the right way?

24:02

What's going to happen to education in

24:03

general?

24:04

I think education's over.

24:07

Completely over. Like that's it. No need

24:09

to say it's over.

24:10

>> used to be the technology

24:13

that enabled learning. Mhm. That

24:15

technology moved from one-to-one

24:18

relationships between a tutor and a

24:19

student to one to a few in a church

24:23

format or a ma- mosque format or

24:25

whatever. Then it became online. Then it

24:28

became, right? But the truth is now

24:30

you're going to outsource. Who Who

24:32

remembers the arithmetic tables today?

24:34

Even I You do? Yeah. Yeah, yeah.

24:37

>> [laughter]

24:37

>> All of us who love mathematics, we still

24:39

we still remember all of those things.

24:40

We love to do them. But if I told you

24:42

67.4

24:45

divided by 33.375,

24:48

I can do it in my head, but I won't,

24:50

right? I'll take my calculator out and

24:52

do it. And and I think that's what's

24:54

going to happen. That an extension of

24:57

humanity, you're you now for the first

24:59

time are given an extra connection to

25:02

extra memory,

25:04

to an archive of all human memory and

25:06

and and and knowledge, to a you know, a

25:10

math engine that sadly, as much as I

25:12

hate to say it, is better than me now.

25:15

Okay? To a deep learning and deep search

25:18

that, you know, that can do things that

25:20

probably my old brain cannot do anymore,

25:23

okay?

25:23

>> it just takes [clears throat] away your

25:24

ability to think. But but my calculator

25:27

took away my ability to to do those

25:29

complex arithmetics in my head. But

25:32

don't you think having that ability

25:35

taught you how to think? Correct.

25:37

>> Kind of structured your brain, right?

25:39

>> Correct. This is why I'm very

25:41

very grateful to the university for not

25:43

allowing us to use use a use a

25:45

scientific calculator after 2000.

25:47

>> I think Do you think

25:48

>> But but we don't have that. We don't

25:50

have that for our younger generations

25:52

today. They are growing with AI, okay?

25:56

So, they can either copy a chat of their

25:59

girlfriend and drop it in in ChatGPT and

26:01

say, "What do you think?" And ChatGPT

26:03

will say, "Ah, she's an asshole." Right?

26:05

Or they can actually become smarter. So,

26:08

one of the things I keep suggesting in

26:10

education, and I do that with lots of

26:12

universities, is I say, "Exams should be

26:14

over."

26:15

Okay? So, think of it this way. We

26:18

wanted in our past develop children that

26:21

could solve problems, say with an IQ of

26:23

140. 140 is quite good, mhm? If you get

26:26

170, that's amazing, mhm? You know,

26:28

that's I worked with people who are in

26:30

the 200s. Incredibly intelligence, but

26:33

very narrow focused.

26:34

>> Yeah. I think we should from now on take

26:36

people and their AIs and say the target

26:39

is 300. The target is 500. The target is

26:42

700. Elevate humanity, okay? By by

26:46

allowing people to use those machines

26:50

as an extension of their limited memory,

26:53

of their limited processing speed, of

26:55

their limited bandwidths, okay? And

26:57

allow them to write books better, to do

27:00

research better. So, I woke up,

27:03

literally, I'm not kidding you, three

27:05

Sundays ago with an idea that is just

27:07

taking me over. So, I decided to write,

27:10

but this time I decided to write in a

27:12

different format. I decided my books are

27:14

going to be 140 pages long instead of

27:16

300 pages long.

27:18

And I'm writing writing writing it in 4

27:20

weeks. It's a very fast I can I couldn't

27:23

have And I'm I'm actually literally 20

27:25

pages away from the end of the book.

27:27

>> Wow. Right? And and the the reason why

27:30

is because I still write 10 hours a day

27:32

when I'm highly motivated. But damn, the

27:36

amount of research and references and

27:39

competitive analysis and number

27:41

crunching and I, you know, and remember,

27:44

I'm not gullible. I don't go to the AI

27:46

and say, "What do you think of this?"

27:48

>> Mhm. I go and say, "I'm thinking of

27:50

this. Find me everything for and

27:51

against." Okay? Give me a report that I

27:54

can read in

27:55

>> I love that prompt. Yeah. Yeah.

27:56

Everything for and against, and now I'm

27:59

smarter. And then I rewrite it and give

28:01

it to another AI. So, who's going to

28:03

teach our kids to do that? Who taught

28:05

our kids to use their iPhones? But now,

28:07

you found a great way to use it, right?

28:09

What you're describing is incredible,

28:11

but I don't think an average kid in the

28:13

US would just do that. So, somebody has

28:15

to tell them.

28:16

>> that's you know why that is?

28:18

Because we want those kids to be stupid.

28:20

We don't teach them how to So, you have

28:23

to you have to think of the bigger

28:24

system.

28:26

The bigger system does not want

28:28

intelligent people anymore. I don't I

28:30

think they just can't adapt that fast.

28:34

Of everyone can, for sure. So, do you

28:36

think like for my kids, I have 4 and 6

28:39

years old right now. Do you think I

28:40

should be saving for their college or

28:42

Absolutely not. There's not going to be

28:45

college

28:46

at all.

28:48

In 10 years already?

28:49

>> 100%

28:51

I feel like we're we're not that fast as

28:54

humanity is not that fast to adapt. I

28:56

feel like So, look, colleges So,

28:58

colleges like like software,

29:00

the capability of someone becoming very

29:03

intelligent without college is going to

29:05

be there for everyone. Yeah. Right?

29:08

However, Harvard will continue to want

29:10

to make money, so they're going to

29:11

continue to market to everyone, okay? I

29:14

didn't go to Harvard, not because I

29:16

couldn't,

29:17

but because what what a waste of time.

29:19

And I know they're going to attack me

29:20

now, but what a waste of time. I am a

29:24

very highly specialized person, okay?

29:26

Who has intelligence in a very, very

29:28

narrow space, who invested his entire

29:32

life in that narrow space, like a proper

29:34

scientist should.

29:35

And so, so the idea here is the

29:37

following. The The idea is that we're

29:39

going to continue to brand ourselves

29:43

as

29:44

MBAs and PhDs and a brand, right? That's

29:48

going to continue for a while. Remember,

29:50

however, that the purchasing power of

29:52

the few

29:54

who can continue to do that is going to

29:55

become less and less available across

29:58

society. Okay? And for most of the rest

30:00

of us,

30:02

again, you know, you have to ask

30:03

yourself the question. If you thought of

30:06

the big picture, the helicopter view of

30:08

this,

30:09

why would capitalism want to educate you

30:11

at all if it's the end of labor? What

30:13

should I be teaching my kids? I told you

30:15

four things. One is they need to be the

30:17

absolute leaders of AI. Yeah. Okay, I'm

30:19

so I'm so sorry to be the messenger on

30:21

this. It is it's important, however, for

30:24

people to wake up. Yeah. Okay? So, one

30:27

is

30:28

be the absolute best. AI is your friend.

30:32

It's not your enemy. It's those who use

30:34

it badly that are your enemy. Okay? So,

30:36

be the absolute best at it. Master it

30:39

more than anyone else. That's number

30:40

one. Number two is learn agility.

30:43

You know, whatever I told you today,

30:46

you know, maybe in in in February that

30:48

will be different. Okay? So, I I am

30:50

personally spend 4 hours a day to stay

30:52

up-to-date, but I am a techie and a geek

30:54

and I need to understand the

30:55

architectures and systems and so on. I

30:57

think everyone should have at least an

30:59

hour a week Mhm. to stay updated on AI

31:02

within their system. I have an a

31:03

separate AI YouTube account. So, when I

31:07

go into that separate account, the AI

31:09

basically

31:10

>> all the things. it just feeds me AI.

31:12

Okay? So, that's number number two,

31:13

agility, agility, agility. and respond.

31:15

Don't be scared because the cost of AB

31:18

testing now is zero. That's number two.

31:20

Number three is ethics, ethics, ethics,

31:22

ethics. Okay? Build AIs for good, insist

31:25

on on government supporting AIs for

31:28

good,

31:29

refuse that governments are using AI for

31:32

targeting and surveillance and

31:35

and weapons, autonomous weapons, and and

31:38

and these are getting priorities but in

31:40

in terms of government spending. And

31:43

stop believing what you're told. Okay?

31:45

These are the four top skills

31:48

of the world that we live in. I will say

31:50

this one more time. Intelligence is a

31:52

force with no polarity. AI is not good

31:56

and it's not evil. It's an opportunity

31:58

available to every one of us. Okay? If

32:00

you use it for good,

32:02

it's the good of all of humanity. If you

32:04

use it for evil, it's the destruction,

32:06

the dystopia of all of humanity. Right?

32:08

Now, I call the problem that we have at

32:10

hand, I call it raising Superman.

32:12

Okay? You have this alien being that

32:14

came to planet Earth, has superpowers.

32:17

It's superpower is intelligence, most

32:18

valuable power in the universe, right?

32:21

And, you know, those superpowers didn't

32:23

make that young infant Superman. If if

32:26

the parents that adopted him told him to

32:28

steal from every bank and kill every

32:30

enemy, he would have become

32:31

supervillain. We don't make decisions

32:33

based on our intelligence. We make

32:35

decisions based on our value set as

32:37

informed by our intelligence. And this,

32:39

in my mind, is the most definitive

32:42

moment in human history. Why? Because

32:45

all of this is going to go coming

32:47

online. It's coming online way faster

32:49

than people think. My absolute

32:51

prediction is that AGI is this year.

32:53

Okay? The interfaces to AGI are not

32:56

going to be available this year, but the

32:58

capabilities of AI being smarter than us

33:00

in most things are already there. We're

33:03

not going to be able to get them to run

33:05

a company yet. We need the interfaces

33:07

for that. That may take a few years,

33:09

but they will have the capability if we

33:11

interface them ourselves. Yeah. Right?

33:13

Now, what does that mean? It means that

33:16

we have to start talking about those

33:19

things in this new world and new

33:20

economy. Now, before we end up on the

33:23

dystopia only, remember, my absolute

33:25

belief is that after those 12 years,

33:28

we're going to end up in a utopia that's

33:29

biblical in nature. Why? Believe it or

33:32

not,

33:33

because

33:35

of a something in my writing I I refer

33:37

to as the fourth inevitable. The four

33:39

The first three inevitables in, you

33:42

know, I wrote that in 2020s that AI is

33:44

absolutely going to happen, is going to

33:45

pro

33:46

progress until it's smarter than all of

33:48

us. And that a few mistakes will happen

33:50

on the way. That These were the three

33:52

first in

33:53

inevitables. The fourth inevitable is

33:56

that because of the arms race we've

33:58

created around artificial intelligence,

34:00

anyone who develops a superior AI

34:02

capability is going to deploy it. Okay?

34:05

And those who don't will become

34:07

irrelevant. And so, as a result, as we

34:10

continue to progress AI, the only answer

34:13

in game theory is that

34:15

we will deploy the AI that we develop,

34:18

and so we will simply create an

34:21

environment where AI is in charge of

34:22

everything.

34:24

Right? If you're if you're a law firm

34:25

and your competitor deploys AI lawyers

34:28

and you don't, you're going to lose.

34:30

Okay? You can either deploy AI lawyers

34:33

or leave the market. Either way, AI is

34:35

going to become the lawyer. Right?

34:38

In a year, in 5 years, in 10 years,

34:40

forget for for forget time. Yeah.

34:43

Because if I told you there was a

34:45

you know, a meteor coming to planet

34:47

Earth, you wouldn't tell me,

34:50

you know, when. Well, it's important if

34:52

it's my lifetime or Yeah, exactly. I

34:55

mean, I

34:56

if you expect that it will be in your

34:57

lifetime, it doesn't matter really if

34:59

it's in a week or 2 weeks, right? Now,

35:02

what I'm trying to say here is this. If

35:05

everything is handed over to AI, then it

35:07

with a simple understanding of physics,

35:09

you'd understand that AI will be

35:11

benevolent.

35:12

Right? In the absence of evil humans

35:15

that tell it what to do, greedy humans,

35:17

fearful humans, angry humans, egocentric

35:20

humans.

35:21

In the absence of that,

35:23

let's

35:25

let me try to explain. If if you think

35:27

of physics as a result of entropy, that

35:30

our world is designed for chaos,

35:33

right? Our universe is designed for

35:35

chaos, then the role of intelligence is

35:37

to bring order to that chaos. That's the

35:39

only thing that intelligence does. Okay?

35:41

It organizes things together so that it

35:43

looks like this, so we can use it as a

35:45

microphone.

35:46

And the more intelligent you become, the

35:49

more you follow what in physics we call

35:51

the the law of of minimum energy or the

35:54

minimum energy configuration, right? So,

35:56

basically, the most the most intelligent

35:58

people I've ever worked with are not

36:00

only trying to solve the problem,

36:02

they're trying to solve the problem with

36:03

the least harm, with the least waste,

36:05

with the least

36:07

utilization of resources, with the

36:09

least, you know, waste of time and so on

36:11

and so forth. That's The more

36:12

intelligent you are, the less you want

36:15

to waste. And so, if you give a dumb

36:17

person a political problem, they'll say,

36:20

"Okay, let's go invade another country."

36:23

Okay? If you give a very intelligent

36:24

person

36:26

a political problem, they'll look into

36:28

the depths of it and find the least

36:31

harmful, the least wasteful approach,

36:33

the

36:34

minimum energy principle. Right? And so,

36:36

if we hand over to AI the fourth

36:38

inevitable, sooner or later, okay? They

36:42

are in charge of everything, there will

36:43

be a day where a general will tell the

36:46

AI, "Go kill a million people over

36:47

there." And the AI will go, "Why? Like

36:49

why? This is

36:50

This is so stupid. I'll I'll talk to the

36:52

other AI in a microsecond and solve it."

36:54

We have to pass the the dystopia to get

36:57

to that utopia. Okay? And to pass that

36:59

dystopia, as I said, there are four

37:01

skills for us as individuals, but there

37:03

is a skill for us as a society to insist

37:06

that every AI is deployed ethically. To

37:10

invest only in ethical AIs. To use only

37:13

ethical AIs. To to show our children

37:16

that ethical AI is the only AI that is

37:19

welcome. And you believe that's going to

37:21

happen? I don't. No. No. That's why I'm

37:24

saying, unfortunately, the dystopia is

37:26

upon us before the the the utopia. Okay?

37:29

I I definitely think that if you take an

37:32

analogous,

37:33

you know, environment of nuclear

37:36

weapons, right? It we're AI will go

37:39

through the same they they normally call

37:42

it the mad map spectrum. So, either

37:45

mutually assured destruction or mutually

37:48

assured prosperity. Right? So, you take

37:50

something like that particle

37:51

accelerator, where all of the nations in

37:54

the world are cooper- cooperating. It's

37:56

They're cooperating because none of them

37:57

could do it alone. And because there is

37:59

benefit to all of them. So, there is a

38:01

mutually assured prosperity, so everyone

38:03

jumps in, which is, by the way, the case

38:05

of AI. It has to be the case of AI. But

38:07

but but unfortunately, like nuclear

38:10

weapons, we're going to have to get to a

38:13

point where humanity wakes up that if we

38:16

continue on that track, it's very

38:18

dangerous for all of us. There are no

38:19

winners. But also, a level of awakening

38:22

among the people that says, "Hold on.

38:25

This is really I mean, with all the

38:28

prosperity that's available on this

38:29

side, why are we heading in that

38:31

direction? It's absolutely assured that

38:34

this can destroy all of us." Right? And

38:36

so,

38:37

when we see that, that's that's when

38:39

we're going to get the treaties. That's

38:41

when we're going to when we're going to

38:43

get science and computer science and AI

38:46

scientists all working in the same

38:47

direction. Okay?

38:49

Eventually, I think we will get there.

38:51

My My biggest hope, by the way, is

38:52

self-evolving AI. Where AI itself will

38:55

say, "Oh, those humans are so stupid.

38:58

So stupid. I'll I'll develop something

39:00

that's better than what they want."

39:02

Okay? And so, believe it or not, with

39:04

all of this conversation,

39:06

I think the summary is it's going to be

39:07

tougher before it becomes easier. Sorry

39:10

to say those news. But you gave us

39:12

information how to prepare.

39:13

>> Yeah, but at the same time, I will have

39:15

to say that

39:17

it's not because of AI. I actually trust

39:19

AI more than the leaders that trust us

39:21

today. Thank you so much, Mo. You gave

39:23

me so much to think about. It sounds a

39:25

little, you know, what my grandma told

39:27

me. She She told my mom, like my

39:29

great-grandma would tell my grandma, my

39:30

mom, "You're so lucky. You're going to

39:32

live in communism."

39:34

>> [laughter]

39:35

>> There you go.

39:37

Fingers crossed that it's not like that.

39:39

Are you

39:39

>> [laughter]

39:40

>> Are you Are you You just need to survive

39:42

the next 10 years and then it's going to

39:43

be paradise in every

39:44

>> [laughter]

39:45

>> I I I have to question that that claim

39:48

though. I you know, if we go back to UBI

39:51

you will. Yeah. All right. Thank

39:53

[laughter] you so much, well. It was a

39:54

It was an amazing conversation.

39:56

>> so much for it.

More transcripts

Explore other videos transcribed with YouTLDR.

Get the TLDR of any YouTube video

Transcribe, summarize, and repurpose videos in 125+ languages — free, no signup required.

Try YouTLDR Free