91ÖÆƬ³§

00;00;00;27 - 00;00;01;17

Hey everyone!

00;00;01;17 - 00;00;03;21

I'm super excited to be sitting down

with Pulitzer

00;00;03;21 - 00;00;06;21

Prize winning investigative journalist

Gary Rivlin.

00;00;06;24 - 00;00;10;20

He's been reporting on the personalities

and power structures behind Silicon Valley

00;00;10;20 - 00;00;12;08

for nearly 30 years.

00;00;12;08 - 00;00;13;16

What's cool about Gary is

00;00;13;16 - 00;00;16;27

he has the inside scoop on the who

and what of the AI boom.

00;00;17;09 - 00;00;20;02

I want to ask him about the technology,

risks and promises

00;00;20;02 - 00;00;23;02

he sees on the horizon,

as well as the players driving it.

00;00;23;09 - 00;00;26;09

And if we can trust them, let's find out.

00;00;29;27 - 00;00;32;21

The first thing I wanted to ask you,

I realize, is like a very loaded question

00;00;32;21 - 00;00;35;28

for a technology podcast, but maybe

I'll ask it anyway and see where it goes.

00;00;36;22 - 00;00;40;08

Gary,

I read that you have or used to have,

00;00;41;09 - 00;00;45;05

basically a plaque on your desk,

that your North Star is to state

00;00;45;05 - 00;00;49;17

complex social issues in human terms,

impossible to ignore.

00;00;50;02 - 00;00;51;15

Which I love, by the way.

00;00;51;15 - 00;00;56;01

But I wanted to ask you, you know,

as you kind of look from your vantage

00;00;56;01 - 00;01;01;05

point in 2025, you know what

are the complex issues of the day

00;01;01;10 - 00;01;05;00

or, you know, the things that you think

are worth crafting these narratives

00;01;05;00 - 00;01;07;13

around that,

that you want to be a part of,

00;01;07;13 - 00;01;10;13

or at least want somebody

you know, telling stories of.

00;01;10;16 - 00;01;14;03

And by the way, it's still right

there above my desk, a little bit faded.

00;01;14;05 - 00;01;14;22

I should probably,

00;01;16;03 - 00;01;17;08

redo it, but,

00;01;17;08 - 00;01;23;22

so, you know, my whole approach

is to tell stories through the people.

00;01;24;03 - 00;01;25;26

You know, we're talking about AI.

00;01;25;26 - 00;01;27;25

Like, you know, is it a good thing?

00;01;27;25 - 00;01;29;02

Is it a bad thing?

00;01;29;02 - 00;01;33;02

Why did arrive when it did

you know that set of questions?

00;01;33;02 - 00;01;34;14

What's it going to mean?

00;01;34;14 - 00;01;37;14

Or the impact

on, you know, society, jobs, etc.?

00;01;37;21 - 00;01;40;06

You know, rather than,

00;01;40;06 - 00;01;44;02

writing a book exploring each one of those

issues chapter by chapter.

00;01;44;14 - 00;01;48;09

I really kind of try to stay true

to to that motto.

00;01;48;09 - 00;01;49;27

And just like

00;01;49;27 - 00;01;52;24

I think it would be dull for people,

but if I can kind of introduce you

00;01;52;24 - 00;01;55;13

to a set of characters,

you're almost emblematic, right?

00;01;55;13 - 00;01;58;19

I mean, here

is someone trying to cash in on AI.

00;01;58;20 - 00;02;03;06

Here's somebody who had the opposite view

by telling their stories.

00;02;03;10 - 00;02;06;11

You know, I think people could understand

00;02;06;11 - 00;02;09;27

these complex social issues

in a very relatable human way.

00;02;11;09 - 00;02;12;02

Right.

00;02;12;02 - 00;02;14;02

So, so let's let's talk about that

a little bit.

00;02;14;02 - 00;02;18;16

And you've been, you know, the

the questions of AI and the story of AI

00;02;18;16 - 00;02;21;18

obviously is kind of intrinsically

linked to Silicon Valley today.

00;02;21;18 - 00;02;24;21

And you've been following

the Silicon Valley beat, if I can call it

00;02;24;21 - 00;02;27;21

that, for, you know, quite a long time.

00;02;27;23 - 00;02;31;01

As you went back there and looked

at some of these characters, you know,

00;02;31;01 - 00;02;35;10

what's what's changed in the past

20 some odd years there.

00;02;35;10 - 00;02;38;02

And you know what's fundamentally

the same.

00;02;38;02 - 00;02;38;10

Right?

00;02;38;10 - 00;02;43;14

So, you know, I started writing about tech

in the mid 1990s.

00;02;43;14 - 00;02;45;27

In fact, I moved from writing

about politics at tech.

00;02;45;27 - 00;02;47;10

I have an engineering background.

00;02;47;10 - 00;02;53;01

I program earlier in my life and just

boom, the.com, the internet, was with us.

00;02;53;01 - 00;02;56;28

And so,

you know, some things don't change, right?

00;02;56;28 - 00;03;00;18

You know,

just like when Netscape went public,

00;03;00;18 - 00;03;02;18

the internet was starting to spread.

00;03;02;18 - 00;03;07;01

You know, you saw a lot of excitement,

a lot of over enthusiasm

00;03;07;14 - 00;03;11;19

for the ASM, you know, and so, you know,

kind of the the hype, the overhype,

00;03;11;19 - 00;03;15;20

the promise, the over promises, you know,

the internet was going

00;03;15;20 - 00;03;17;07

to bring about world peace.

00;03;17;07 - 00;03;18;14

You know, essentially,

00;03;18;14 - 00;03;22;11

you know, now AI is going to equalize

between the developing world

00;03;22;11 - 00;03;24;25

and the developed world

is going to cure cancer.

00;03;24;25 - 00;03;26;04

Yes. Yeah.

00;03;26;04 - 00;03;28;22

I think AI is an incredible technology.

00;03;28;22 - 00;03;32;26

Obviously, the internet, you know,

has changed society in profound ways.

00;03;32;26 - 00;03;39;04

But know some of the over promise

almost feeds the other side's, skepticism.

00;03;39;04 - 00;03;41;04

And like AI is not good.

00;03;41;04 - 00;03;44;14

It might help some scientists help cure,

00;03;45;12 - 00;03;46;28

cancer, but you know AI, is

00;03;46;28 - 00;03;50;28

it not going to cure cancer,

at least anytime, anytime soon?

00;03;52;06 - 00;03;55;04

You know, one big difference is the money.

00;03;55;04 - 00;03;57;14

So when I first started

writing about tech,

00;03;57;14 - 00;03;59;05

I was always interested

in the venture capitalists

00;03;59;05 - 00;04;02;11

and the startups and that whole ecosystem,

like this idea, like,

00;04;03;13 - 00;04;06;05

you know, our idea for a company

is either going to work,

00;04;06;05 - 00;04;09;18

be worth back then, tens of millions,

hundreds of millions

00;04;09;18 - 00;04;12;22

now, billions, if not trillions,

or it's going to be worth nothing.

00;04;12;26 - 00;04;15;07

And the venture capitalists

who are staking, you know, back

00;04;15;07 - 00;04;18;07

then, millions now, tens of millions,

hundreds of millions, billions.

00;04;18;19 - 00;04;21;04

But, you know, in 90, 1995,

00;04;21;04 - 00;04;24;03

venture capital was under $10 billion

a year.

00;04;24;03 - 00;04;27;25

By 2021, it was over 300 billion a year.

00;04;27;25 - 00;04;32;13

You know, roughly 100 and 3050

billion went into AI startups,

00;04;33;17 - 00;04;34;13

last year.

00;04;34;13 - 00;04;37;12

I mean, a lot of it went into a few,

you know, like,

00;04;37;13 - 00;04;40;25

anthropic, OpenAI x AI, that's Elon Musk.

00;04;41;02 - 00;04;44;01

You know, they raise

collectively tens of billions of dollars,

00;04;44;01 - 00;04;47;01

you know, almost $100

billion just between those three.

00;04;47;24 - 00;04;50;18

But there's still a lot more money

going to AI startups.

00;04;50;18 - 00;04;53;14

So the money has really changed.

00;04;53;14 - 00;04;57;08

I guess the final difference is,

you know, when the internet came out,

00;04;57;09 - 00;05;01;06

like maybe the biggest criticism

was around the attention span.

00;05;01;17 - 00;05;02;09

You know, oh,

00;05;02;09 - 00;05;06;02

there is if you're always online,

you know, this instant gratification.

00;05;06;02 - 00;05;08;22

Well,

what was it going to do for consumerism?

00;05;08;22 - 00;05;11;22

In our in our society,

00;05;12;14 - 00;05;14;12

with AI,

00;05;14;12 - 00;05;18;17

there was much more of,

a worry, much more of a, a backlash.

00;05;18;17 - 00;05;21;25

People didn't,

you know, greet open AI with, excuse me,

00;05;21;25 - 00;05;26;21

I with open arms the way they did

engineered people are fearful of it.

00;05;26;22 - 00;05;27;28

You know, we could talk about that.

00;05;27;28 - 00;05;30;26

I you know, I think it's kind of Hollywood

induced fear.

00;05;30;26 - 00;05;34;21

I don't think the media has done

such a great job, with AI.

00;05;35;15 - 00;05;39;09

So, you know,

I cause a double battle, like, there's

00;05;39;09 - 00;05;42;28

the usual battle of creating a startup

and trying to cash in.

00;05;42;28 - 00;05;45;26

But, you know, the second battle

of trying to convince people

00;05;45;26 - 00;05;49;01

that this is a good thing,

and the laser eyed robots aren't going to,

00;05;50;08 - 00;05;52;23

beat us into submission.

00;05;52;23 - 00;05;55;07

So. So let's talk about those laser

eyed robots for a minute.

00;05;55;07 - 00;05;58;14

I'm curious, Gary,

because, yeah, you've painted,

00;05;59;03 - 00;06;02;17

I think pretty accurately the picture

of how people talk about AI.

00;06;02;17 - 00;06;07;06

Everything from, you know, this, you know,

utopia of abundance, which, by the way,

00;06;07;06 - 00;06;10;22

I get people,

you know, who I talk to, I'd say routinely

00;06;10;22 - 00;06;13;28

who use words like that to describe

where AI is going to take us,

00;06;14;03 - 00;06;17;11

you know, curing cancer,

you know, and, you know, laser robots

00;06;17;11 - 00;06;20;15

who are going to destroy the world and,

you know, the blink of an eye, you know,

00;06;21;18 - 00;06;25;22

from, from your vantage point, like,

how transformational is this?

00;06;25;22 - 00;06;29;11

Like, what's your outlook for what

this technology can and cannot do

00;06;29;18 - 00;06;31;00

in the next few years?

00;06;31;00 - 00;06;33;21

So Reid Hoffman,

who's the main character for my book.

00;06;33;21 - 00;06;34;10

All right.

00;06;34;10 - 00;06;35;13

Here's a set of terms.

00;06;35;13 - 00;06;38;04

There's the Zoomers, the laser robots.

00;06;38;04 - 00;06;39;08

We're fearful of this thing.

00;06;39;08 - 00;06;42;25

Let's bomb the data centers before they

take over to give the extreme of that,

00;06;43;26 - 00;06;48;07

the Zoomers, who say nothing.

00;06;48;07 - 00;06;51;05

There should be no speed bumps,

no regulation.

00;06;51;05 - 00;06;55;19

The only role of government,

is to speed this thing up, in part

00;06;55;19 - 00;06;57;15

because we're in a race

with China, in part

00;06;57;15 - 00;07;02;01

because tech is what's kept

the US, economy going and strong.

00;07;02;01 - 00;07;05;18

And, you know, it's one of our competitive

advantages as a country.

00;07;05;25 - 00;07;08;16

You know, I'm more of

what Reid Hoffman calls a bloomer.

00;07;08;16 - 00;07;10;29

You know, I, I definitely am an optimist.

00;07;10;29 - 00;07;15;02

I think AI is going to do

extraordinary things around science,

00;07;15;09 - 00;07;19;23

medicine, education,

a whole wide range of things.

00;07;19;23 - 00;07;23;16

But I do think we have to be deliberate,

about it, in part

00;07;23;20 - 00;07;26;26

because AI is

where way ahead of the public.

00;07;26;27 - 00;07;29;28

You know, Pew did a study last year,

a poll last year,

00;07;30;05 - 00;07;34;06

and the majority of Americans

are fearful, of of AI,

00;07;34;12 - 00;07;38;22

less than a third

are optimistic or, excited about AI,

00;07;39;24 - 00;07;40;19

being here.

00;07;40;19 - 00;07;42;23

And, you

know, I didn't have the best time.

00;07;42;23 - 00;07;46;06

I mean, you know, innovation happens

when innovation happens, but, you know,

00;07;46;06 - 00;07;50;06

kind of hitting at the end of 2022,

the distrust for big tech,

00;07;50;06 - 00;07;54;11

the distrust for tech generally

is pretty much at an all time high.

00;07;54;19 - 00;07;57;27

And, you know, of course, I like,

you know, ultimately humans

00;07;57;27 - 00;08;03;25

will lose their apex status is the, you

know, the smartest entity, on on Earth.

00;08;03;25 - 00;08;06;07

But, you know,

we're asking for a lot of trust, like, if,

00;08;06;07 - 00;08;07;15

you know, the big thing this year.

00;08;07;15 - 00;08;09;00

Of course

I'm sure you've been talking about.

00;08;09;00 - 00;08;11;21

This is personal agents, a personal agent.

00;08;11;21 - 00;08;14;20

They really have to get to know me.

They have to know my private details.

00;08;14;20 - 00;08;17;19

They're going to have

a lot of agency in my life.

00;08;17;19 - 00;08;21;13

And if if these companies get ahead, it

the this, you know,

00;08;21;13 - 00;08;24;27

the Googles and Microsofts of the world

or the startups

00;08;25;15 - 00;08;28;22

that hope people are using their agents,

they get too far ahead of people.

00;08;28;22 - 00;08;31;25

I, I'm not sure people

are going to trust these things.

00;08;31;25 - 00;08;34;16

And, you know, one less, point on that is

00;08;34;16 - 00;08;38;12

there is going to be something bad

that happens because I

00;08;38;13 - 00;08;42;09

just to make one up, like,

you know, $1 trillion is siphoned off

00;08;42;19 - 00;08;48;01

from the global economic system

before a human could even notice it.

00;08;48;01 - 00;08;51;07

And, you know, at that point, of course,

we all know the way these things play out,

00;08;51;13 - 00;08;52;28

you know, people are going to be fearful.

00;08;52;28 - 00;08;57;29

And so I, I'm a bloomer who thinks

there should be some regulation.

00;08;57;29 - 00;09;02;03

But I also think that

AI is going to be central

00;09;02;03 - 00;09;06;07

to our lives

within a decade, maybe 15 years.

00;09;07;26 - 00;09;10;02

It has vast potential.

00;09;10;02 - 00;09;11;01

Right.

00;09;11;01 - 00;09;13;21

So, so with that in mind and,

and thinking about the,

00;09;13;21 - 00;09;16;21

you know, the trust part.

00;09;17;01 - 00;09;19;26

We're in a space now where,

00;09;19;26 - 00;09;22;01

well, let me ask it as a question.

00;09;22;01 - 00;09;24;15

There's in your book

and I'm sure a little bit beyond

00;09;24;15 - 00;09;28;06

your book, there's a handful of top dogs,

I'll call them, who it seems like

00;09;28;06 - 00;09;32;28

have a really outsized influence

over this technology, how it's developed,

00;09;33;04 - 00;09;36;09

how it's played out,

how it collects data, what it does.

00;09;36;23 - 00;09;38;04

Is that fair?

00;09;38;04 - 00;09;39;25

Is there really kind of a,

00;09;39;25 - 00;09;43;18

you know, a small cabal that has,

you know, an oversize amount of control

00;09;43;26 - 00;09;47;15

and, you know, do we trust them

and should we trust them?

00;09;48;02 - 00;09;48;11

Right.

00;09;48;11 - 00;09;52;22

So so you're getting one of my fears

around AI.

00;09;52;22 - 00;09;56;08

It's nothing to do with laser eyed robots

or anything like that.

00;09;56;08 - 00;10;00;12

It's the consolidation in the hands

of the same few tech companies

00;10;00;12 - 00;10;04;00

that have been dominant,

for the last decade or two.

00;10;04;01 - 00;10;04;23

You know, it's funny.

00;10;04;23 - 00;10;09;05

So I started this book right

the end of 2022, started 2023,

00;10;09;14 - 00;10;12;14

and I went in search of the next Google,

the next meta.

00;10;12;17 - 00;10;15;05

And, you know,

I ended up concluding, like,

00;10;15;05 - 00;10;19;11

I fear that the next Google is Google in

AI and the next meta is meta.

00;10;19;22 - 00;10;21;13

You know that. Yeah, big.

00;10;21;13 - 00;10;23;03

This stuff is really expensive.

00;10;23;03 - 00;10;26;03

When I first started,

you know, people were talking about,

00;10;26;14 - 00;10;32;06

you know, millions, tens of millions

to train, fine tune and operate

00;10;32;06 - 00;10;36;16

these chatbots, large language models

where whatever you want to call it.

00;10;36;16 - 00;10;38;23

And the same with text to video, audio to,

00;10;39;25 - 00;10;42;03

a text audio.

00;10;42;03 - 00;10;45;24

By the time I was done

reporting at the end of 2024,

00;10;45;24 - 00;10;50;16

May 2024, it was hundreds of millions,

if not billions.

00;10;50;16 - 00;10;54;28

And Dario Modi for from anthropic

they do clogged the chat bot clogged.

00;10;55;08 - 00;10;57;28

You know he's

estimating that they're going to need $100

00;10;57;28 - 00;11;01;08

billion by 2027 to train these things.

00;11;01;20 - 00;11;04;11

And so who has that kind of money?

00;11;04;11 - 00;11;06;07

You know, Google, Microsoft.

00;11;06;07 - 00;11;09;01

They have 100 billion

or so laying around in cash.

00;11;09;01 - 00;11;13;26

But if you have to raise $100 billion

or even if it's only, you know,

00;11;14;12 - 00;11;17;19

three, five, $10 billion, well,

a large venture capital outfit

00;11;17;19 - 00;11;21;27

in Silicon Valley has $1 billion,

all told, in a fund.

00;11;22;06 - 00;11;23;13

And so we're talking about billions.

00;11;23;13 - 00;11;27;23

And so that's one way this is,

way to to big tech.

00;11;27;23 - 00;11;29;08

And the other is data. Right.

00;11;29;08 - 00;11;32;26

This is really central to the,

the remedies that government

00;11;32;26 - 00;11;36;13

is now talking to remedies

for the Google antitrust trial.

00;11;36;19 - 00;11;42;04

You know, a, a federal judge found that,

you know, Google is a monopolist.

00;11;42;05 - 00;11;44;22

It abused its power.

So now what should we do?

00;11;44;22 - 00;11;49;17

And a lot of the discussion

I think rightfully is around the data.

00;11;49;25 - 00;11;52;25

You know, OpenAI approached

Google and said, hey,

00;11;53;08 - 00;11;57;03

can we lease we

can we kind of buy access to your data?

00;11;57;03 - 00;11;58;14

And they said, no.

00;11;58;14 - 00;12;00;23

And that's a huge advantage.

00;12;00;23 - 00;12;04;29

In fact, anyone reads the book

every time Google steps on stage

00;12;04;29 - 00;12;08;07

in 2023 or 2024,

they fall flat on their face.

00;12;08;14 - 00;12;12;27

Google was so far ahead of every other

big tech company around machine learning.

00;12;13;00 - 00;12;13;25

You know, give them credit.

00;12;13;25 - 00;12;17;27

They were using it in the 2000s

and they went on a hiring binge binge.

00;12;18;07 - 00;12;21;03

You know, in the early 20 tens, long

before anyone else, you know,

00;12;21;03 - 00;12;25;04

they came up with the transformer

paper in 2017 that T and GPT,

00;12;26;17 - 00;12;27;20

and so it's, you know,

00;12;27;20 - 00;12;32;03

foundational research, and yet, you know,

it's a big company that was scared.

00;12;32;03 - 00;12;34;17

So OpenAI puts out ChatGPT.

00;12;34;17 - 00;12;38;06

They try to catch up to a race, they're

already winning, but they're scared.

00;12;38;06 - 00;12;39;13

So they weren't winning it.

00;12;39;13 - 00;12;41;09

And they kept on following their phase.

00;12;41;09 - 00;12;45;13

Despite that, Gemini,

their chat bot, is one of the 2

00;12;45;13 - 00;12;48;10

or 3 most popular chat bots on the planet

because they're a Google.

00;12;48;10 - 00;12;49;29

You know, we have many of us.

00;12;49;29 - 00;12;51;26

Google's the front door to the internet.

00;12;51;26 - 00;12;55;15

They have all this data

and they can, you know, train,

00;12;55;27 - 00;12;59;07

better and cheaper, than others.

00;12;59;07 - 00;13;02;05

So yeah, it's it it's interesting.

00;13;02;05 - 00;13;06;28

And, you know, your position is very

similar to mine, I think, which is,

00;13;08;19 - 00;13;10;26

it really feels like there's,

I don't know if I want to call it

00;13;10;26 - 00;13;15;03

an arms race or a space race,

but there's this race

00;13;15;03 - 00;13;19;05

of some kind between the these,

you know, tech giants to try and just

00;13;20;04 - 00;13;22;06

own the platform of the future,

00;13;22;06 - 00;13;25;17

you know, own the data, use it as a way

00;13;25;17 - 00;13;28;17

to, continue to monetize,

00;13;29;00 - 00;13;31;16

you know, the the public here, I guess

00;13;31;16 - 00;13;34;16

so, you know, with that in mind,

00;13;34;23 - 00;13;38;08

who are going to be the winners

and the losers of,

00;13;38;09 - 00;13;41;09

you know, that this shift to AI

and is it necessary, you know,

00;13;41;09 - 00;13;43;17

when you look out over it,

do you really see it's

00;13;43;17 - 00;13;47;17

winner take all with one or a few,

you know, technology giants

00;13;47;17 - 00;13;51;15

that probably are the OpenAI, the Googles,

the Microsofts, the anthropic?

00;13;51;23 - 00;13;54;23

Or is it going to be more democratic

than that?

00;13;55;12 - 00;13;58;09

Yeah, I mean, we're talking

about predicting the future.

00;13;58;09 - 00;14;02;20

So, you know, take anything coming out of

my mouth right now with a grain of salt.

00;14;03;26 - 00;14;04;15

You know, I

00;14;04;15 - 00;14;08;02

would kind of distinguish

between OpenAI anthropic, they're,

00;14;08;10 - 00;14;11;14

you know, relatively new startups

that have a lot of cash.

00;14;11;14 - 00;14;14;03

They've raised billions, tens

of billions of dollars

00;14;14;03 - 00;14;18;26

in the case of OpenAI, versus

a Google, Microsoft,

00;14;19;20 - 00;14;23;11

MIT, Amazon and Apple or they're, they're

00;14;24;00 - 00;14;27;03

they're not quite doing it yet,

but they're, they're in the race,

00;14;27;29 - 00;14;31;15

to, you know, I, I, I want to be optimist.

00;14;31;15 - 00;14;34;17

I, I guess I'm built

I, I root for startups,

00;14;35;01 - 00;14;36;25

you know, I root for the underdog.

00;14;36;25 - 00;14;37;27

You know, to me.

00;14;37;27 - 00;14;42;09

Well, one thing that was exciting about

AI is the possibility that there would be

00;14;42;19 - 00;14;46;29

kind of a new set, of major, major,

major players.

00;14;47;24 - 00;14;52;18

But I, I even wonder if an OpenAI AI,

despite all the successes.

00;14;52;18 - 00;14;52;25

Right.

00;14;52;25 - 00;14;55;26

It just raised $40 billion earlier

this year.

00;14;55;26 - 00;14;59;12

That's the largest raise

in the history of venture, capital.

00;14;59;12 - 00;15;03;02

It's worth, you know, it's it's

got a paper worth of $300 billion right

00;15;03;02 - 00;15;07;12

now, which, if you put in the fortune

500, would make it top 50.

00;15;07;16 - 00;15;09;17

I think it's top 50 for the globe.

00;15;09;17 - 00;15;13;11

And so, you know, it's it's

a, it's a, you know, it's ten years old,

00;15;13;19 - 00;15;18;17

but it's, it's seems on paper just kind of

a global AGI has been around forever.

00;15;18;19 - 00;15;21;02

But I'm still not sure they

can make it because they,

00;15;22;14 - 00;15;25;12

they're bringing in revenue,

but they're still losing a ton of money.

00;15;25;12 - 00;15;27;07

I mean, give them credit.

00;15;27;07 - 00;15;30;16

They're making virtually nothing

in 2022 and 2024.

00;15;30;24 - 00;15;34;02

They brought in almost $4 billion.

00;15;34;02 - 00;15;36;29

They'll have a much bigger number

that than that this year,

00;15;36;29 - 00;15;38;08

you know, 10 billion or more.

00;15;38;08 - 00;15;42;20

But they're still losing money because

this stuff is so expensive to operate.

00;15;42;21 - 00;15;44;29

You can use ChatGPT here.

00;15;44;29 - 00;15;46;23

There are other products for free.

00;15;46;23 - 00;15;51;01

You can pay a premium,

but you get a very good service for free.

00;15;51;01 - 00;15;51;16

And so,

00;15;51;16 - 00;15;54;25

you know, it's it's like kind of that

consumer thing and you could just go to,

00;15;55;04 - 00;15;58;20

you know,

Claude or EW.com or Gemini or Exa,

00;16;00;11 - 00;16;03;04

what it was

00;16;03;04 - 00;16;06;04

grok you could use, you know, there's mass

00;16;07;05 - 00;16;09;28

and so that that one costs

00;16;09;28 - 00;16;12;04

and so I, I,

00;16;12;04 - 00;16;15;02

I do fear that what's going to happen

00;16;15;02 - 00;16;18;28

is it's inevitable

that one of the giants of Microsoft

00;16;18;28 - 00;16;21;28

or Google are going to buy,

00;16;21;29 - 00;16;24;15

an anthropic or an open AI.

00;16;24;15 - 00;16;28;10

You know, right now,

Microsoft's put around 15 billion or so

00;16;28;27 - 00;16;31;18

into OpenAI, you know, anthropic.

00;16;31;18 - 00;16;35;08

There are two biggest

funders are Amazon and Google.

00;16;35;23 - 00;16;38;10

And so let's take Dario

00;16;38;10 - 00;16;41;14

Amodei at at his word in two years.

00;16;41;14 - 00;16;43;15

Where's he getting that $100 billion?

00;16;43;15 - 00;16;46;20

I mean I suppose

he could raise that money,

00;16;46;20 - 00;16;50;07

but I'm not even sure

they're profitable at that point.

00;16;50;07 - 00;16;53;16

So I do wonder if it's inevitable

that these

00;16;53;16 - 00;16;57;03

popular chat bots get bought,

by one of the tech giants.

00;16;57;14 - 00;17;01;00

Well, I have to imagine the tech giants

are chomping at the bit at the opportunity

00;17;01;00 - 00;17;04;07

to buy, you know, an opening

Iron Anthropic two and add them to their,

00;17;04;15 - 00;17;05;18

you know, to their stack.

00;17;05;18 - 00;17;09;23

Well, and they have been doing that,

you know, kind of this idea of an acquire

00;17;10;05 - 00;17;14;04

or inflection was having to be the company

at the center of my, book.

00;17;14;04 - 00;17;17;04

It was started by Reid Hoffman,

this guy, Moustafa Suleiman,

00;17;17;09 - 00;17;20;03

know, kind of a co-founder of DeepMind,

which, you know, it's kind of the first

00;17;20;03 - 00;17;24;08

great machine learning, startup

that started in 2010, in London.

00;17;24;08 - 00;17;27;28

And Google

bought it for 650 million, dollars.

00;17;28;03 - 00;17;30;26

You know,

they had everything going for them.

00;17;30;26 - 00;17;35;02

You know, it's like Reid Hoffman, best

connected guy, arguably in Silicon Valley.

00;17;35;02 - 00;17;40;16

You know, kind of, the the pedigree

of a moustafa Suleyman, a great CEO.

00;17;40;22 - 00;17;41;29

You know, Bill gates funded it.

00;17;41;29 - 00;17;45;24

Well, I am, you know, kind of had that,

you know, kind of Hollywood thing.

00;17;46;23 - 00;17;49;27

Ashton Kutcher was, was a, investor.

00;17;49;29 - 00;17;51;16

They hoarded all this talent.

00;17;51;16 - 00;17;55;12

Some of the top researchers in the world

raise a billion

00;17;55;12 - 00;17;59;04

and a half dollars, in, like, 15 months.

00;17;59;16 - 00;18;00;24

But it wasn't enough.

00;18;00;24 - 00;18;03;25

And so basically,

everyone at the company jumped,

00;18;04;29 - 00;18;07;29

Microsoft Microsoft paid the investors

00;18;08;00 - 00;18;11;00

$650 million to make everyone happy.

00;18;11;02 - 00;18;15;16

And, you know, essentially this company

inflection, really cutting edge chatbot.

00;18;15;21 - 00;18;18;22

What I loved about it is it's great at IQ,

00;18;18;22 - 00;18;21;22

emotional intelligence, not just IQ.

00;18;21;28 - 00;18;26;12

But now the talent, virtually

the whole company is inside Microsoft.

00;18;26;12 - 00;18;29;27

Google did the same thing with character

AI spent billions of dollars,

00;18;30;27 - 00;18;34;03

essentially to hire the two creators of,

00;18;35;03 - 00;18;35;22

of character.

00;18;35;22 - 00;18;38;17

I really popular another popular chat bot.

00;18;38;17 - 00;18;41;27

These two founders who are Google ites,

got frustrated at Google

00;18;41;27 - 00;18;44;24

because they couldn't do what they wanted

to do, went and did it,

00;18;44;24 - 00;18;45;23

and it started up.

00;18;45;23 - 00;18;49;11

And then Google had to pay billions

of dollars just to bring them back.

00;18;51;04 - 00;18;51;26

Yeah.

00;18;51;26 - 00;18;54;24

So so, you know, when I think about AI

00;18;54;24 - 00;18;57;24

and the impact it's going to have,

there's clearly this kind of,

00;18;58;28 - 00;19;00;11

I don't know

if I'm supposed to call it this,

00;19;00;11 - 00;19;03;01

but I'll call it a boys

club of like the big,

00;19;03;01 - 00;19;05;04

you know, incumbent

or emerging organizations

00;19;05;04 - 00;19;07;27

and the tens of billions of dollars

and the kind of,

00;19;07;27 - 00;19;10;08

you know, political infighting

you see there. And it's.

00;19;10;08 - 00;19;14;19

Yeah, you know, it's funny, Gary,

and I don't know if this is a result of

00;19;14;19 - 00;19;18;07

your background or just kind of a parallel

I picked up on, but I can

00;19;18;10 - 00;19;21;10

I can certainly see through lines

to what's going on here, to

00;19;21;10 - 00;19;24;10

what you probably saw,

you know, covering City Hall,

00;19;24;29 - 00;19;28;03

and those kind of political beats

all those years ago where it's jockeying

00;19;28;03 - 00;19;31;28

for, for power and influence is that

I mean, is that fair?

00;19;31;28 - 00;19;35;18

Is are there is that just coming through,

00;19;35;29 - 00;19;38;28

do you think sorry,

is there merit to that?

00;19;39;11 - 00;19;40;00

Yeah.

00;19;40;00 - 00;19;42;19

There is like, you know, in, in,

00;19;42;19 - 00;19;45;23

political sense

we call backroom deals kind of

00;19;46;16 - 00;19;49;17

going on, you know,

that politics is happening in front of us,

00;19;49;17 - 00;19;53;12

but the real deals are being

cut, behind closed doors and

00;19;54;16 - 00;19;56;08

it is true of Silicon Valley.

00;19;56;08 - 00;19;57;12

It is true of tech.

00;19;57;12 - 00;19;59;22

Like it's one of my big,

you know, Chris is.

00;19;59;22 - 00;20;04;29

But let's just get five smart guys

and they are invariably guys in a room

00;20;04;29 - 00;20;06;06

and we'll figure it out.

00;20;06;06 - 00;20;08;23

And, you know,

I mean, that works for your app.

00;20;08;23 - 00;20;13;02

That has worked very well, for for tech.

00;20;13;02 - 00;20;16;02

But I think I different, you know,

00;20;16;02 - 00;20;20;26

I, I requires

a much broader, group of people.

00;20;20;26 - 00;20;24;10

I mean, some of it is what discipline

are you from?

00;20;24;10 - 00;20;25;28

Like, it's not just computer scientists.

00;20;25;28 - 00;20;26;28

It's not just,

00;20;26;28 - 00;20;31;03

you know, tech folks, you know, I,

I mean, first off, it's it's linguists.

00;20;31;15 - 00;20;32;14

You know, you need linguists.

00;20;32;14 - 00;20;34;18

It's math. It's math.

00;20;34;18 - 00;20;37;26

And so but I, I really want a much broader

group,

00;20;37;26 - 00;20;42;23

given the power and potential of AI

and what AI is.

00;20;43;20 - 00;20;46;11

I want a broader group

in terms of discipline.

00;20;46;11 - 00;20;49;10

I want a broader group

in terms of who we're talking about.

00;20;49;16 - 00;20;52;08

I mean, tech is heavily

skewed male, of course,

00;20;52;08 - 00;20;55;11

and these days it's very white and Asian.

00;20;55;26 - 00;20;57;14

Our world is much broader than that.

00;20;57;14 - 00;20;58;17

It's a global thing.

00;20;58;17 - 00;21;02;22

And so bringing bringing in people

from different backgrounds from,

00;21;02;23 - 00;21;06;19

you know, around the globe,

you know, to figure this out, I again, I

00;21;06;28 - 00;21;07;25

coming back to a point

00;21;07;25 - 00;21;11;04

I made before about the strategy,

I think it's gonna be very dangerous,

00;21;12;04 - 00;21;13;01

for tech,

00;21;13;01 - 00;21;16;07

if this is the same old, same old

where it's, you know,

00;21;16;16 - 00;21;20;13

Sam Altman

negotiating with a Sachin Nadella

00;21;20;13 - 00;21;24;12

at Microsoft with a few other people,

in the in the room.

00;21;24;12 - 00;21;29;24

I really think for kind of trust

and safety, reasons, there needs to be

00;21;29;24 - 00;21;33;05

a much broader group than the same old,

same old that Silicon Valley does.

00;21;34;24 - 00;21;35;22

It makes sense.

00;21;35;22 - 00;21;38;21

And as I think about

like what that impact could be

00;21;38;21 - 00;21;42;21

and what this technology could do, I mean,

we talked about why, you know, the whole

00;21;42;21 - 00;21;46;02

the hype is kind of, you know, inflated

potentially on both sides.

00;21;46;23 - 00;21;49;23

What impact do you see this having,

00;21;50;29 - 00;21;53;25

I guess across, across the population.

00;21;53;25 - 00;21;57;22

And is this a leap forward

in some ways for,

00;21;57;23 - 00;22;01;28

you know, the poor or the traditionally

marginalized or underserved or is it

00;22;02;12 - 00;22;06;28

do you foresee it just going to end up

being like a rich, get richer, platform?

00;22;07;21 - 00;22;08;00

Yeah.

00;22;08;00 - 00;22;10;15

So again, I'm about to predict the future.

00;22;10;15 - 00;22;12;03

So take anything I say.

00;22;12;03 - 00;22;14;04

That's what we do here, Gary.

00;22;14;04 - 00;22;16;02

But a great grain of salt.

00;22;16;02 - 00;22;17;06

But, you know, I.

00;22;20;16 - 00;22;23;01

So I do

00;22;23;01 - 00;22;26;01

think I can be

00;22;27;03 - 00;22;29;29

a great equalizer.

00;22;29;29 - 00;22;30;29

You know, kind of.

00;22;30;29 - 00;22;32;24

I give a couple examples.

00;22;32;24 - 00;22;38;10

So in the sense of making art, you know,

if you want to make your, your movie

00;22;38;10 - 00;22;41;22

maker, you want to make a movie,

you need a lot of money to do that.

00;22;41;22 - 00;22;43;05

You have to hire people and stuff.

00;22;43;05 - 00;22;46;16

But with AI, you can make a movie.

00;22;46;23 - 00;22;50;28

You know, it's it's kind of the great

equalizer in in terms of the arts.

00;22;50;28 - 00;22;54;06

It gives you these superpowers,

to do things.

00;22;54;06 - 00;22;56;02

We could talk about the ramifications

that, you know,

00;22;56;02 - 00;22;59;29

all those actors and production

people, etc., who aren't getting the job.

00;22;59;29 - 00;23;01;23

And, you know, to add a caveat, it's

not just like,

00;23;01;23 - 00;23;03;24

make me a Martin Scorsese movie.

00;23;03;24 - 00;23;06;11

You know, the human is still the driver.

00;23;06;11 - 00;23;09;17

The human has to bring the the storylines

and the plots and,

00;23;10;04 - 00;23;12;09

you know, plot twists and all that.

00;23;13;14 - 00;23;15;17

And, you know, another example is, again,

00;23;15;17 - 00;23;19;06

it could be, an equalizer that,

00;23;21;02 - 00;23;23;24

anyone can code, right?

00;23;23;24 - 00;23;26;29

I mean, I don't think we're quite

at that point yet, but we're not that far

00;23;26;29 - 00;23;31;17

away for a point where you don't have

to have a computer science degree.

00;23;31;17 - 00;23;34;17

You don't have to have much experience

of all, if any at all.

00;23;34;18 - 00;23;35;10

And you could code

00;23;35;10 - 00;23;39;05

and that that that will really helps

as so in, in the developing world

00;23;39;14 - 00;23;41;22

could see a problem

that's not a first world problem

00;23;41;22 - 00;23;46;00

but a problem in their world

and do code, you know, and,

00;23;46;01 - 00;23;50;27

and have this magic power

of coding to come up with a solution,

00;23;52;00 - 00;23;54;00

tailored to their world.

00;23;54;00 - 00;23;56;11

You know,

just use the example of education.

00;23;56;11 - 00;24;01;02

I think a has amazing potential,

for for education.

00;24;01;09 - 00;24;04;09

They did a study, in Nigeria,

00;24;04;13 - 00;24;08;11

where they had teachers use AI tutors.

00;24;08;11 - 00;24;11;13

And of course, an AI tutor

could figure out exactly what

00;24;11;13 - 00;24;15;19

a student doesn't understand

and then feed them, examples,

00;24;16;12 - 00;24;20;11

information that helps them,

learn that aspect and stuff.

00;24;20;11 - 00;24;23;14

And what they found

is just in a six week period,

00;24;23;22 - 00;24;26;20

they had like two years of advancement

in the students.

00;24;26;20 - 00;24;27;17

They worked with.

00;24;27;17 - 00;24;31;13

And I really do think, you know, this,

this idea that, you know, a tutor,

00;24;31;24 - 00;24;35;06

you know, $100 an hour or $200 an hour,

I, I don't know what a tutor,

00;24;35;13 - 00;24;36;08

but it's a lot of money.

00;24;36;08 - 00;24;40;08

It's beyond the means for,

you know, most people on the planet.

00;24;40;15 - 00;24;43;19

But what of this tutor was in your pocket?

00;24;44;23 - 00;24;45;07

And it was,

00;24;45;07 - 00;24;48;10

if not free, virtually

free or health care.

00;24;48;10 - 00;24;52;15

This is a great stat

that basically a billion or so people

00;24;52;24 - 00;24;56;01

have access to ready health care,

00;24;56;15 - 00;24;59;20

but there's 5 billion

or so who have a smartphone.

00;25;00;03 - 00;25;01;29

And should you go to a doctor?

00;25;01;29 - 00;25;04;16

Is a doctor better

or a nurse practitioner or whatever?

00;25;04;16 - 00;25;06;18

A health care

professional better than a phone?

00;25;06;18 - 00;25;08;10

Yes yes yes yes yes.

00;25;08;10 - 00;25;11;15

But what about all those people

who have zero access

00;25;11;25 - 00;25;15;03

and you know, you could take a picture

and my child has this rash,

00;25;15;18 - 00;25;18;18

you know, should I be worried and

00;25;19;00 - 00;25;21;20

move heaven and earth

to get to a hospital?

00;25;21;20 - 00;25;22;21

Or is it nothing?

00;25;22;21 - 00;25;25;01

And so that could be a life

and death thing.

00;25;25;01 - 00;25;27;09

So, you know, I, I do think,

00;25;27;09 - 00;25;31;03

you know, kind of a

this idea of a eye doctor in your pocket,

00;25;31;12 - 00;25;35;24

you know, it's certainly better

than not having any access to health care.

00;25;35;27 - 00;25;38;26

So that's another way I see,

00;25;38;26 - 00;25;43;19

you know, education or medicine

that I could be an equalizer.

00;25;43;19 - 00;25;47;19

But the flip of that is that,

you know, winner takes all,

00;25;47;19 - 00;25;49;16

if not winner takes most.

00;25;49;16 - 00;25;52;18

And, you know, I, I am nervous that

00;25;53;17 - 00;25;55;15

well, as we've been talking

00;25;55;15 - 00;25;58;20

power over AI

in the hands of just a few people

00;25;58;27 - 00;26;03;15

will mean that that small group

in Silicon Valley or technology

00;26;03;15 - 00;26;08;26

generally, is going to grow super wealthy

and only a little bit will spill over to.

00;26;08;26 - 00;26;10;24

To the rest of us.

00;26;10;24 - 00;26;13;02

Yeah. It's.

00;26;13;02 - 00;26;13;16

Yeah.

00;26;13;16 - 00;26;15;28

Prediction

there is it is a tough game and it's,

00;26;15;28 - 00;26;18;28

there's

like so many moving pieces to this.

00;26;19;05 - 00;26;20;25

And I don't know

if this is something you heard about

00;26;20;25 - 00;26;22;02

or you're concerned about as well.

00;26;22;02 - 00;26;25;06

But when I think about that,

you know, that AI doctor, which

00;26;25;29 - 00;26;27;26

you know, whether whether we choose

to acknowledge it or not,

00;26;27;26 - 00;26;29;07

I feel like that's already

starting to happen.

00;26;29;07 - 00;26;31;28

Right. People are talking to ChatGPT

about their symptoms.

00;26;31;28 - 00;26;36;08

And and frankly, I think for the most

part, that's probably a good thing. But

00;26;37;15 - 00;26;40;03

what I don't know if is happening right

now, but I have to

00;26;40;03 - 00;26;45;13

imagine we're on the cusp of is ChatGPT

or whatever, generative AI saying,

00;26;45;17 - 00;26;49;04

oh, and you should take an Advil

for your symptoms.

00;26;49;22 - 00;26;54;20

And there's quietly, you know, money

changing hands between, you know, the tech

00;26;54;20 - 00;26;57;26

giant and Advil so that you take an Advil

and not a Tylenol.

00;26;57;26 - 00;27;02;26

And suddenly you know the difference to me

between the web

00;27;02;27 - 00;27;03;26

and some of these generative

00;27;03;26 - 00;27;07;09

AI things is just and even even Google,

frankly, is it?

00;27;07;18 - 00;27;11;23

It gives you an answer that it acts like

is definitive, right?

00;27;11;26 - 00;27;14;22

Yeah. And what kind of influence does

that have over people?

00;27;14;22 - 00;27;18;06

Is that something that came up at all or

something that you have a perspective on?

00;27;18;11 - 00;27;19;05

Yeah, yeah.

00;27;19;05 - 00;27;23;26

And these these chat bots speak

with an authority, I joke.

00;27;23;26 - 00;27;28;07

Well they were, they were trained

on a disproportionately male data set.

00;27;28;18 - 00;27;29;23

And so they're no at all.

00;27;29;23 - 00;27;30;08

In fact,

00;27;30;08 - 00;27;31;16

they've had to be teach to say

00;27;31;16 - 00;27;34;16

like it's okay to say,

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

00;27;34;21 - 00;27;37;19

but what you're getting at

is one of the things,

00;27;37;19 - 00;27;40;22

again, before we were talking about

like what the fear out

00;27;40;22 - 00;27;45;11

there is with laser eyed robots,

there are very real fears that are,

00;27;46;27 - 00;27;48;27

that we could see right now.

00;27;48;27 - 00;27;53;21

And one of them is the power of

AI to manipulate or the power of AI

00;27;53;27 - 00;27;57;28

to make suggestions that you think, well,

this is just based on its knowledge

00;27;57;28 - 00;28;01;23

reading, you know, what

it's been trained on and you don't know,

00;28;01;23 - 00;28;05;01

like, oh, no, the maker of Advil has paid,

00;28;05;24 - 00;28;08;19

for this, by the way, right now,

00;28;08;19 - 00;28;13;11

I don't think there's any chat bots

that are giving recommendations.

00;28;14;07 - 00;28;17;15

You, you know, I,

I think people are using and probably

00;28;17;15 - 00;28;22;04

sometimes abusing as if, like people

who do have access to medicine,

00;28;22;08 - 00;28;26;12

like, oh, I yeah, I'm

going to learn about this here.

00;28;26;12 - 00;28;27;22

And like, you know, sometimes that works.

00;28;27;22 - 00;28;31;06

Sometimes it does it, you know,

there's there's stories where, you know,

00;28;31;06 - 00;28;35;22

someone has struggled for months, years,

to figure out what's wrong with them. They

00;28;36;22 - 00;28;39;22

give it

their symptoms to ChatGPT or whatever,

00;28;40;00 - 00;28;43;00

and they figure out the solution.

00;28;43;00 - 00;28;44;25

I mean, you know, I can be incredible.

00;28;44;25 - 00;28;49;27

There's a study where,

they are training a chat bot

00;28;50;12 - 00;28;53;19

to have early detection

00;28;54;06 - 00;28;58;14

of diabetes type two, type two diabetes

just by hearing your voice.

00;28;58;24 - 00;29;00;17

And so, you know,

that's the positive where

00;29;00;17 - 00;29;02;21

there's amazing things that can happen.

00;29;02;21 - 00;29;06;17

You know, they're really accurate

at reading mammograms and,

00;29;06;17 - 00;29;11;01

you know, X-rays and all that

far more accurate than, human doctors.

00;29;11;01 - 00;29;14;01

Me, I want doctors to use,

00;29;14;29 - 00;29;17;29

I as a backup, as a check.

00;29;18;08 - 00;29;21;14

I don't want I reading, my, you know,

00;29;21;23 - 00;29;25;17

I another thing that worries me

is autonomous AI,

00;29;25;26 - 00;29;28;26

you know, for the foreseeable future,

if not forever.

00;29;28;26 - 00;29;31;10

We need humans in the loop. These things,

00;29;32;19 - 00;29;33;19

they know a lot

00;29;33;19 - 00;29;36;19

about everything,

but they don't understand a thing.

00;29;36;25 - 00;29;38;10

They have no common sense.

00;29;38;10 - 00;29;41;10

They have no, you know, the it's

00;29;41;11 - 00;29;44;13

a parrot is repeating words,

but it doesn't really understand.

00;29;44;24 - 00;29;46;08

You know, there's an expression

00;29;46;08 - 00;29;50;17

I learned doing the research here

that it tends to know everything.

00;29;50;19 - 00;29;52;18

These chatbots, these large language

models, tend

00;29;52;18 - 00;29;56;14

to know everything

that humans over 20 years old know.

00;29;56;18 - 00;30;00;22

But they don't know a lot that a kid

under a five year old, five years old,

00;30;02;11 - 00;30;04;03

would know, you know, just kind of

00;30;04;03 - 00;30;07;19

just the common sense stuff,

like just stupid stuff, like,

00;30;07;29 - 00;30;10;26

you chew this, but you don't chew this,

you swallow, you know, that.

00;30;10;26 - 00;30;14;03

Kind of like

they have to be taught every right thing.

00;30;14;11 - 00;30;18;09

And so, you know, it's I, I do fear

00;30;19;00 - 00;30;23;02

that we're going to invest too much power

prematurely in these things.

00;30;23;02 - 00;30;26;02

I really don't want autonomous

AI anytime soon.

00;30;26;10 - 00;30;28;28

That's that that that's fair. What?

00;30;30;03 - 00;30;33;03

When you look at

what we have or where we're going with AI,

00;30;33;07 - 00;30;36;27

is there anything that came across to you

as being underreported right now,

00;30;36;29 - 00;30;39;14

either in capabilities or in risks

that, you know,

00;30;39;14 - 00;30;42;11

you think we need to have a little bit

more of a light on?

00;30;42;11 - 00;30;42;17

Yeah.

00;30;42;17 - 00;30;45;17

So, I mean, broadly speaking,

00;30;46;13 - 00;30;49;28

I don't think the media I mean,

I used to work for the New York Times,

00;30;49;28 - 00;30;53;00

so I am big media,

I guess you can say, at least in my past,

00;30;53;10 - 00;30;57;07

but you know, that the New York Times,

Washington Post of the world,

00;30;57;19 - 00;31;00;22

I don't think they've done a great job,

covering

00;31;01;27 - 00;31;06;17

AI, generative AI in particular,

since the release of ChatGPT.

00;31;06;24 - 00;31;09;16

And that was November of 2022.

00;31;09;16 - 00;31;13;04

You know, there was a little bit about,

you know, its potential,

00;31;13;14 - 00;31;16;28

but right away

it was it's all the negatives.

00;31;16;28 - 00;31;20;08

It's going to destroy education,

you know, that it's going to write

00;31;20;08 - 00;31;23;13

essays for kids and, you know,

kids are going to use it for cheating.

00;31;23;14 - 00;31;24;14

Yeah, that's a problem.

00;31;24;14 - 00;31;28;21

But you know, what about in fact,

I have an eighth grader and,

00;31;28;21 - 00;31;32;01

you know, he writes a composition

and he gives it to ChatGPT

00;31;32;24 - 00;31;35;04

and it gives him feedback.

00;31;35;04 - 00;31;38;19

You know, the teacher's busy, but like,

you know, how can I improve this?

00;31;38;19 - 00;31;43;00

Well, you can make a better topic sentence

and it's just repetitive kind of thing

00;31;43;06 - 00;31;44;13

and, you know, gets good feedback.

00;31;44;13 - 00;31;47;01

And so it's an amazing tool for a teacher.

00;31;47;01 - 00;31;49;15

You could devise lesson plans.

00;31;49;15 - 00;31;54;01

You can, you know, use it to help

with grading and and, and, and all.

00;31;54;25 - 00;31;58;19

And, you know, I think that

I think that has been largely missing

00;31;58;27 - 00;31;59;14

from the coverage.

00;31;59;14 - 00;32;03;22

Occasionally you see an article about, oh,

we could do this around health care.

00;32;04;14 - 00;32;09;03

And, and also,

I really don't think the broader public

00;32;09;03 - 00;32;13;09

has a sense

of what the true promise, of this is.

00;32;13;15 - 00;32;15;19

And the flip side of that is,

00;32;16;23 - 00;32;19;09

I think people in

00;32;19;09 - 00;32;24;04

part because of the media coverage,

are scared of the wrong things.

00;32;24;13 - 00;32;29;03

I mentioned a, autonomous

I bet AI the use of AI in warfare,

00;32;29;24 - 00;32;33;04

the use of AI for surveillance,

you know, a a

00;32;33;26 - 00;32;36;26

AI that can come up with,

00;32;38;04 - 00;32;39;25

a new vaccine

00;32;39;25 - 00;32;43;25

or a new treatment

or tailored treatments for this specific,

00;32;44;12 - 00;32;47;12

cancer or whatever someone is suffering

00;32;47;18 - 00;32;52;12

through can also create a deadly pathogen

that could kill a lot of people,

00;32;52;21 - 00;32;54;09

you know, kind of the potential for

00;32;54;09 - 00;32;58;16

AI to write your wedding toast

or for your friend's 50th birthday,

00;32;58;29 - 00;33;02;09

you know, is a powerful tool

in the hands of scammers.

00;33;02;24 - 00;33;07;28

And so I wish that's

where the coverage was.

00;33;07;28 - 00;33;11;08

You know, around

here are the potential pitfalls

00;33;11;14 - 00;33;14;15

and this is the stuff we need to navigate.

00;33;14;16 - 00;33;20;06

I mean, every tech that

every technology ever is both pro and con.

00;33;20;06 - 00;33;21;06

It goes both ways.

00;33;21;06 - 00;33;24;07

You know, the cars

extraordinary change our society.

00;33;24;20 - 00;33;25;13

But, you know,

00;33;26;22 - 00;33;27;28

global warming

00;33;27;28 - 00;33;32;11

35,040 thousand people in the US

die in car accidents, you know,

00;33;32;11 - 00;33;36;13

so every technology cuts both ways in the

same will be true with AI.

00;33;36;22 - 00;33;40;28

And I think I,

I wish we being smarter, more deliberate

00;33;41;04 - 00;33;45;26

about making sure I was a net positive

rather than net negative right now.

00;33;45;26 - 00;33;49;09

I don't know, like

I'll give you an example of regulation.

00;33;49;13 - 00;33;54;01

So the Biden administration imposed,

I thought light regulation like, okay,

00;33;54;01 - 00;33;56;01

if you're working on one of these

powerful models

00;33;56;01 - 00;33;58;14

before you release it,

we're gonna require you to red team it

00;33;58;14 - 00;34;01;28

to hire an outside group to,

you know, test for vulnerabilities

00;34;01;28 - 00;34;03;25

and then share the results

with the US government.

00;34;03;25 - 00;34;08;05

Didn't seem a very heavy lift,

but as soon as Trump took over, he,

00;34;09;07 - 00;34;11;16

counteracted that executive order.

00;34;11;16 - 00;34;16;16

And JD Vance in January,

I mean, February was in Paris

00;34;16;16 - 00;34;19;20

for the third of these worldwide

AI summits.

00;34;20;00 - 00;34;23;00

The first two were totally focused

on trust and safety.

00;34;23;06 - 00;34;25;06

This one was not. And there was J.D.

00;34;25;06 - 00;34;28;08

Vance, who stood up there and said,

stop with the hand-wringing.

00;34;28;17 - 00;34;32;15

We need to win this right, this race

because it China, China, China thing.

00;34;32;15 - 00;34;35;09

And you know that

that that that worries me.

00;34;35;09 - 00;34;37;11

Government does have a role.

00;34;37;11 - 00;34;40;07

Again,

I don't want them micromanaging this.

00;34;40;07 - 00;34;43;07

That's

not the way we have great innovation.

00;34;43;13 - 00;34;46;22

But I do think they need

I do think we need guardrails.

00;34;47;00 - 00;34;50;19

And I do think

we need kind of this mindset.

00;34;50;19 - 00;34;54;15

So we're really looking at the stuff

around the corner

00;34;54;15 - 00;34;56;23

that's negative and this stuff

around the corner on this positive.

00;34;56;23 - 00;34;59;03

So we can, you know, work it out.

00;34;59;03 - 00;35;00;20

So it's more positive than negative

00;35;01;25 - 00;35;02;06

right.

00;35;02;06 - 00;35;06;08

So so yeah you know I'm thinking about

that kind of duma versus zoomer mentality.

00;35;06;08 - 00;35;12;08

And it feels like, you know as we

look at the regulatory bodies broadly,

00;35;12;14 - 00;35;16;05

we've gone from Duma to Zuma

and just started like ripping off,

00;35;16;06 - 00;35;17;28

you know, red tape

if you can call it that.

00;35;17;28 - 00;35;22;08

But, it sounds like your perspective,

Gary, is we do need

00;35;22;26 - 00;35;28;17

the the focus on some sort of responsible

AI guardrails are important

00;35;28;17 - 00;35;31;26

and that that now is the right time

to get some of that right.

00;35;31;27 - 00;35;33;03

Is that fair? Right.

00;35;33;03 - 00;35;37;06

So I you know, it's funny,

I was so if you if we had spoken in 2023

00;35;37;13 - 00;35;39;24

I would have been really optimistic

about this.

00;35;39;24 - 00;35;40;19

Right?

00;35;40;19 - 00;35;42;08

I think most of us can agree

00;35;42;08 - 00;35;45;21

that the government blew it

as far as social media, right.

00;35;45;21 - 00;35;49;11

It was just like it took 20 years before

their first hearing about social media.

00;35;49;11 - 00;35;52;11

And by that time, social media had,

you know, caused the problems.

00;35;53;05 - 00;35;55;17

It it caused.

00;35;55;17 - 00;35;58;08

But I like, you know, remember Sam Altman

00;35;58;08 - 00;36;01;10

in the spring of 2023, you know, stood

00;36;01;10 - 00;36;05;18

before the Senate and said,

I have fears around this, too.

00;36;05;19 - 00;36;07;09

We need to be regulated.

00;36;07;09 - 00;36;10;16

He was talking about, AI body like,

you know, a nuclear

00;36;10;16 - 00;36;13;16

regulatory agency like thing for AI.

00;36;13;16 - 00;36;16;16

People have talked about

a global version, of that.

00;36;16;16 - 00;36;19;04

And, you know,

the senators were talking about it.

00;36;19;04 - 00;36;21;23

You know, Chuck Schumer,

then the majority leader,

00;36;21;23 - 00;36;25;18

he had a series of talks in 2023

where they had leaders in

00;36;25;18 - 00;36;29;24

AI come in and talk to the senators about,

here's the way it works.

00;36;29;24 - 00;36;30;22

Here's our fears.

00;36;30;22 - 00;36;33;06

Here's where we think the potential is.

00;36;33;06 - 00;36;36;02

But then you know two things.

00;36;36;02 - 00;36;40;16

One, the tech companies themselves,

because of the arms

00;36;40;16 - 00;36;44;16

race, as you called it before, you know,

they really dropped the trust

00;36;44;18 - 00;36;48;11

and safety, aspect,

or at least that took a backseat,

00;36;49;14 - 00;36;51;06

to winning this race.

00;36;51;06 - 00;36;54;15

And, you know,

kind of our, our model out to your,

00;36;55;17 - 00;36;56;08

model.

00;36;56;08 - 00;36;59;08

And, you know, the other thing

just kind of the wider politics

00;36;59;08 - 00;37;03;25

meant there really hasn't

been much room, lately for kind of debates

00;37;03;25 - 00;37;08;01

about AI when we're talking

about a whole different set of issues.

00;37;08;12 - 00;37;09;17

And, and politics.

00;37;09;17 - 00;37;14;09

So, you know, now I am worried,

you know, where I was at this school,

00;37;14;12 - 00;37;18;27

that I thought it was a good thing

that OpenAI put out ChatGPT.

00;37;19;25 - 00;37;21;13

When they did,

it wasn't that powerful yet.

00;37;21;13 - 00;37;25;12

It really couldn't

do much harm, at that point.

00;37;25;18 - 00;37;29;07

But it sparked the discussion,

you know, long

00;37;29;07 - 00;37;32;26

before this thing was super powerful,

you know, so, you know, it sort of.

00;37;33;15 - 00;37;35;11

We figured out how to make,

00;37;36;10 - 00;37;40;00

jet engines, jet airplanes safe

or relatively safe.

00;37;40;00 - 00;37;42;12

There's very few accidents.

00;37;42;12 - 00;37;44;23

Because, you know,

it started off with prop planes

00;37;44;23 - 00;37;47;05

and we came up with rules,

and then they got better and stronger.

00;37;47;05 - 00;37;50;03

You know, cars. We've, you know, cars.

00;37;50;03 - 00;37;54;13

We've, anti-lock brakes and,

you know, kind of all those kind of

00;37;55;01 - 00;37;57;22

although airbags,

all those kind of safety things,

00;37;57;22 - 00;37;59;25

you know, it's it's good

that we're kind of struggling

00;37;59;25 - 00;38;02;11

with the technology

before it's super powerful.

00;38;02;11 - 00;38;05;28

But I think in 2023 we were really jumping

on that opportunity.

00;38;06;17 - 00;38;09;05

But now it just I don't hear it at all.

00;38;09;05 - 00;38;12;05

It doesn't seem an issue.

00;38;12;06 - 00;38;13;10

Yeah. Not it it's fair.

00;38;13;10 - 00;38;16;09

It feels like it has been like,

00;38;16;10 - 00;38;20;14

like almost like a pendulum swing

in the complete opposite direction.

00;38;20;14 - 00;38;23;10

And of course, pendulums

never swing to the middle, where they,

00;38;23;10 - 00;38;25;16

you know, more or less.

Should they swing to the other side.

00;38;25;16 - 00;38;26;07

Yeah.

00;38;26;07 - 00;38;27;11

Yeah, yeah.

00;38;27;11 - 00;38;29;29

No, that's that's well well said.

00;38;29;29 - 00;38;33;04

Moving away from, from, you know,

the government and the regulators here,

00;38;33;10 - 00;38;35;26

if we think about the role

that that organizations play

00;38;35;26 - 00;38;38;02

and you know,

call it call it companies here,

00;38;39;07 - 00;38;40;29

what what would be

00;38;40;29 - 00;38;46;09

your best advice for, you know,

if you were brought in to a boardroom of,

00;38;46;09 - 00;38;49;18

you know, a non,

you know, a non, you know,

00;38;49;18 - 00;38;52;24

whatever you want to call them magnificent

seven tech company.

00;38;52;24 - 00;38;56;06

But just, you know, a company

that sits in the middle of the economy

00;38;56;12 - 00;38;59;12

is looking at

AI, is looking at kind of future tech

00;38;59;20 - 00;39;02;22

wants to know if it can help them

or hurt them, what they need to know.

00;39;02;23 - 00;39;05;23

What's kind of your best advice for what

they should be thinking about. So.

00;39;05;27 - 00;39;09;01

So my my second interview

ever as a tech reporter

00;39;09;09 - 00;39;12;29

in the mid 1990s with this guy

Paul Saffo, is a prognosticator,

00;39;13;06 - 00;39;17;17

a futurist is kind of the more,

more common word for it.

00;39;17;22 - 00;39;18;17

And he said

00;39;19;25 - 00;39;22;22

it wasn't his idea,

but he's the one who shared it with me.

00;39;22;22 - 00;39;26;09

We tend to overestimate the short

term impact,

00;39;27;01 - 00;39;30;20

of a new technology,

an underestimate the long term aspect.

00;39;30;20 - 00;39;32;27

And, you know, the internet was a perfect

example of that.

00;39;32;27 - 00;39;37;03

Like all these startups

thought we're going to get rich overnight.

00;39;37;12 - 00;39;41;11

And, you know,

every company had to embrace the internet

00;39;41;11 - 00;39;42;07

or they were going to die.

00;39;42;07 - 00;39;45;10

Well, you know, most of those companies,

they thought they were going to get rich

00;39;45;10 - 00;39;47;04

overnight, went out of business.

00;39;47;04 - 00;39;52;06

But slowly but surely, the internet

seeped into more or less everything.

00;39;52;17 - 00;39;55;26

And so, you know, probably

took a, I'm going to say, 15 years,

00;39;56;14 - 00;39;58;04

before it was central to everything.

00;39;58;04 - 00;40;00;23

And I think the same thing

is going to happen with AI.

00;40;00;23 - 00;40;02;29

So back to the boardroom.

00;40;02;29 - 00;40;04;06

You know, I be saying

00;40;05;18 - 00;40;07;04

this stuff is real.

00;40;07;04 - 00;40;08;14

It's coming.

00;40;08;14 - 00;40;12;20

Obviously there are tools that individual

employees can use, large language models,

00;40;12;24 - 00;40;17;07

you know, marketing team, you know,

kind of its graphics ability, and all.

00;40;17;07 - 00;40;18;14

And you should be using that.

00;40;18;14 - 00;40;22;10

In fact, I would encourage

everyone in, in an organization to use it

00;40;22;18 - 00;40;26;06

because our construct is a little bit

has been a little bit

00;40;26;06 - 00;40;32;03

wrong, is in the short medium term, it's

not going to be AI who replaces humans.

00;40;32;03 - 00;40;33;18

There'll be some exceptions to that,

00;40;33;18 - 00;40;37;23

but it's going to be humans who use AI,

who get the best, humans who don't.

00;40;37;29 - 00;40;40;22

This is a powerful tool.

I use it as an editor.

00;40;40;22 - 00;40;43;15

I use as a research assistant.

I'm I'm faster.

00;40;43;15 - 00;40;44;27

I think I'm better.

00;40;44;27 - 00;40;45;11

Because of it.

00;40;45;11 - 00;40;47;20

So I've been encouraging

their individual employees.

00;40;47;20 - 00;40;52;14

But as far as their organization

changing as processes and stuff,

00;40;52;24 - 00;40;55;26

I would encourage them to experiment,

to be thinking about it,

00;40;56;07 - 00;41;00;14

to maybe put a pilot project here,

but it's not there yet.

00;41;00;14 - 00;41;03;14

I mean, this is 2025

is supposed to be the year of the agents.

00;41;03;29 - 00;41;06;28

And just like, well, first off,

they don't have much of a memory.

00;41;06;28 - 00;41;11;00

So there's a personal agent

that's going to be as if we're all wealthy

00;41;11;00 - 00;41;14;00

people with a personal assistant

than those all our preferences

00;41;14;08 - 00;41;18;07

and stuff, like it's having a hard time

remembering us from session to session.

00;41;18;07 - 00;41;22;13

So they're not a good personal assistant

in that, sense.

00;41;22;13 - 00;41;27;20

And, you know, there's interesting things

I see companies doing, with with,

00;41;28;21 - 00;41;29;15

these, these

00;41;29;15 - 00;41;32;27

AI agents, but it's a work in progress.

00;41;32;27 - 00;41;34;26

So not quite there yet.

00;41;34;26 - 00;41;38;22

So, you know, bottom line,

like anticipate it,

00;41;39;03 - 00;41;42;03

play with it,

figure out how it might help.

00;41;42;06 - 00;41;46;15

But you know right now

like there's also the exponential factor.

00;41;46;15 - 00;41;49;06

Like these things have been getting like

00;41;49;06 - 00;41;52;19

ten times

better more or less every year and stuff.

00;41;52;19 - 00;41;56;04

So what we have today, we could presume

that, like what we have in

00;41;56;04 - 00;41;59;04

ten years

will be more powerful, different.

00;41;59;07 - 00;42;01;09

And so and at the same time,

00;42;01;09 - 00;42;03;22

there's this whole thrust ever since deep

think like,

00;42;03;22 - 00;42;05;27

wait, maybe we're going to have smaller

models and stuff.

00;42;05;27 - 00;42;10;21

So, you know, this stuff is moving so fast

and there's still like big changes.

00;42;10;28 - 00;42;14;29

You know, I think a corporations job,

a business leader

00;42;15;02 - 00;42;16;17

or a for that matter, an educator.

00;42;16;17 - 00;42;19;12

Anyone running a big organization

is to be paying attention.

00;42;19;12 - 00;42;20;10

This stuff.

00;42;20;10 - 00;42;23;24

Find ways you can use it to help

individual employees.

00;42;24;14 - 00;42;27;18

But I'm not sure I would turn

the company upside down yet.

00;42;28;27 - 00;42;30;08

I don't, and

00;42;30;08 - 00;42;34;19

I like I like that

and it's, you know, it's clear advice.

00;42;34;19 - 00;42;38;10

And if I can like if I can hone

in on a part of it, going back

00;42;38;10 - 00;42;41;17

to what you said about the start ups and,

you know, kind of the, the,

00;42;42;23 - 00;42;47;04

you know, the web, the web bubble

that we had, you know, 25 years ago.

00;42;47;11 - 00;42;50;27

It sounds like you've got a, you know,

kind of a healthy skepticism of like

00;42;52;01 - 00;42;55;11

companies that are thinking, oh, well, I'm

just going to uproot everything

00;42;55;11 - 00;42;55;24

I'm doing.

00;42;55;24 - 00;42;58;27

And we're an AI company

now, that's not that.

00;42;58;27 - 00;43;00;22

That's not the advice you're giving.

00;43;00;22 - 00;43;03;19

Right. So, you know,

00;43;03;19 - 00;43;06;15

I'm going to contradict myself,

but then I want to defend it.

00;43;06;15 - 00;43;10;05

I think AI is under hyped and overhyped,

00;43;10;25 - 00;43;13;15

and I think it's overhyped

00;43;13;15 - 00;43;17;19

because these startups have taken hundreds

of millions,

00;43;17;19 - 00;43;21;15

billions of dollars,

and they have to overhype it.

00;43;21;19 - 00;43;23;19

They have to tell you

that artificial general

00;43;23;19 - 00;43;27;07

intelligence is just around the corner

to justify the huge amount of money

00;43;27;17 - 00;43;31;04

they have Microsoft, Google,

you know, Amazon, etc.

00;43;31;07 - 00;43;35;02

they're putting Facebook meta,

you know, they're putting, you know, tens

00;43;35;02 - 00;43;39;21

of millions, tens of billions of dollars,

you know, into AI.

00;43;39;22 - 00;43;44;00

So of course, they want to see,

they want to recoup their investment.

00;43;44;05 - 00;43;47;08

So they're overhyping this like, hey,

you know, use copilot

00;43;47;08 - 00;43;50;09

is going to change your life

tomorrow, you know, kind of thing.

00;43;51;09 - 00;43;52;02

And so I think

00;43;52;02 - 00;43;55;02

the individual companies, startups

and big tech alike,

00;43;55;27 - 00;43;59;21

the venture capitalist in front of them,

you know, they're overhyping it.

00;44;00;07 - 00;44;03;01

But I do think it's under hype

in that sense.

00;44;03;01 - 00;44;07;08

I said before that we tend

to underestimate the long term impact.

00;44;07;18 - 00;44;09;13

You know, I,

00;44;09;13 - 00;44;12;11

I think everything

00;44;12;11 - 00;44;16;16

business, education and health care

00;44;17;22 - 00;44;19;28

consumers in their personal relationships,

00;44;19;28 - 00;44;24;22

I think everything is going

to be profoundly impacted, by AI.

00;44;24;26 - 00;44;28;05

And again, I don't know if that's

five years from now, ten years from now,

00;44;28;05 - 00;44;33;01

15 years from now, but at some point

in the next decade or so, you know,

00;44;33;01 - 00;44;36;07

the world will look very different,

the same way the world looked different,

00;44;37;01 - 00;44;41;04

after the internet, the same way

the world, look different after mobile.

00;44;42;00 - 00;44;45;23

You know, I think we're going to see

the same impact, if not a greater impact.

00;44;45;23 - 00;44;47;11

I mean, there's this sense that,

00;44;47;11 - 00;44;52;10

you know, kind of the semiconductor

plus personal computer,

00;44;52;20 - 00;44;58;02

plus the internet, plus the cloud,

plus the phone, you know, all of that

00;44;59;04 - 00;45;02;22

is undergirding

all of that is going to be used in AI.

00;45;03;03 - 00;45;05;26

And so it's almost a culmination,

00;45;05;26 - 00;45;08;28

you know, kind of the computer revolution

or whatever you want to call it,

00;45;08;28 - 00;45;14;03

since, you know, over the last 50,

70 years, AI is the combination.

00;45;14;03 - 00;45;18;15

In fact, you know, I had been, Bill gates

as a fun quote about this.

00;45;18;15 - 00;45;19;01

And, you know, that's

00;45;19;01 - 00;45;23;06

been the Holy grail for him

since he first touched a computer, right?

00;45;23;06 - 00;45;29;06

Like, again, the power of Hollywood,

that power of science fiction and media,

00;45;29;06 - 00;45;33;18

this idea that, you know, like Star Trek,

we could talk to the computer.

00;45;33;18 - 00;45;36;05

For me, it was like,

what was it called, Lost in Space?

00;45;36;05 - 00;45;40;02

That we'd all have a robot helper

who had all the information in the world

00;45;40;02 - 00;45;42;12

that was super powerful

and could protect us.

00;45;42;12 - 00;45;45;21

And I do think that's been the dream.

00;45;46;06 - 00;45;50;15

And I do think, like,

we're getting pretty close to that dream.

00;45;50;26 - 00;45;53;12

Yeah, well, I think that's well said.

00;45;53;12 - 00;45;53;29

And it is,

00;45;55;13 - 00;45;57;26

you know, it

makes me reflect on the fact that this is,

00;45;57;26 - 00;46;00;16

this is the continuation

or the extension of a thread

00;46;00;16 - 00;46;02;08

that's been running for a long time.

Right?

00;46;02;08 - 00;46;04;14

People say I and sometimes

00;46;04;14 - 00;46;06;29

sometimes we talk about AI abstractly

in a way

00;46;06;29 - 00;46;08;18

where it's like,

okay, what are you actually talking about

00;46;08;18 - 00;46;12;26

when you say AI because, you know, beyond

like the narrow scope of gen

00;46;12;28 - 00;46;17;21

AI and maybe even with it, you know,

I it's it's an algorithm, right?

00;46;17;21 - 00;46;21;25

Like it's just it's taking input,

transforming into outputs

00;46;21;25 - 00;46;25;01

and in some way, you know, creating

something new or doing something new.

00;46;25;01 - 00;46;27;14

And it's like, well, that's that's

what we've been doing all along, right?

00;46;27;14 - 00;46;29;19

Like that's

what computers were meant to do.

00;46;29;19 - 00;46;34;20

We've just gotten to a point at it

now where like something has tipped over

00;46;34;26 - 00;46;36;14

and what it's capable of doing,

00;46;36;14 - 00;46;39;21

and whether it's a genetic

or whether it's generative, you know,

00;46;39;27 - 00;46;42;29

but it's still the same,

the same thread, if.

00;46;43;06 - 00;46;44;18

You ask me for surprises.

00;46;44;18 - 00;46;50;23

Another surprise is AI has been with us

for a long, long, long time.

00;46;50;24 - 00;46;57;05

And the term artificial intelligence

was first coined in the 1950s.

00;46;57;15 - 00;47;01;08

And you know that there have been people

who saw like, hey, you know,

00;47;02;14 - 00;47;04;24

useful AI is just around the corner.

00;47;04;24 - 00;47;09;01

It's, you know, it's it's been a decade,

a decade away for like 70 years.

00;47;09;01 - 00;47;11;05

And we're finally around that corner.

00;47;11;05 - 00;47;14;05

We're now finally

where, you know, people can use it.

00;47;14;11 - 00;47;16;16

But the truth is

00;47;16;16 - 00;47;19;27

Google has been using AI, in search.

00;47;19;27 - 00;47;21;08

You know, it's had to help.

00;47;21;08 - 00;47;24;17

Translated, garbled search,

you know, spelling mistakes and such.

00;47;24;17 - 00;47;29;00

They've been using AI since the 2000

to more efficiently deliver up ads.

00;47;29;11 - 00;47;32;18

You know, in the 20 tens,

I'll stick with Google, Google Translate.

00;47;32;18 - 00;47;33;22

That's AI.

00;47;33;22 - 00;47;37;21

And, you know, it's I think 2015

is when Google Translate came out.

00;47;37;21 - 00;47;38;16

So, you know,

00;47;38;16 - 00;47;43;00

billions of people have been using

AI without knowing they're using AI.

00;47;43;00 - 00;47;44;11

You know, suggestion engines.

00;47;44;11 - 00;47;49;05

You go to Netflix, you, you know, you

you get a, recommendation

00;47;49;05 - 00;47;52;24

of, movies

you might like, you know, that's AI.

00;47;52;25 - 00;47;55;10

So we've been using AI for a long time.

00;47;55;10 - 00;47;58;23

I think the difference with generative AI,

the difference with ChatGPT is

00;47;58;23 - 00;47;59;28

we could talk to it.

00;47;59;28 - 00;48;03;09

It wasn't this thing behind the scenes,

that you had

00;48;03;09 - 00;48;06;28

to kind of know how it works and to know

that was artificial intelligence.

00;48;07;10 - 00;48;10;00

Now, you could speak with it

and, you know, it's funny, I,

00;48;10;00 - 00;48;13;00

I kind of was a skeptic

in the second half of the

00;48;13;01 - 00;48;16;09

90s where, okay,

this stuff is really interesting.

00;48;16;09 - 00;48;20;22

I was, you know, I used email, I love,

you know, browsing things on the internet.

00;48;20;22 - 00;48;25;23

But, you know, these dot coms seem

kind of preposterous, to me.

00;48;25;23 - 00;48;27;18

And some of the claims they're being made.

00;48;27;18 - 00;48;30;17

So, you know, the end of 2022,

I started using ChatGPT.

00;48;31;13 - 00;48;32;26

I'm a I'm a journalist.

00;48;32;26 - 00;48;35;16

I was, you know, a reporter

who covered this stuff.

00;48;35;16 - 00;48;41;03

I'm used to tech companies overstating the

potential of something. But

00;48;42;02 - 00;48;44;16

first off, OpenAI, there was no press

release.

00;48;44;16 - 00;48;49;27

There's no press event that's released

this thing in the on November 30th, 2022

00;48;50;01 - 00;48;54;00

with a research report and, you know,

kind of word spread there rather quickly

00;48;54;07 - 00;48;54;21

and stuff.

00;48;54;21 - 00;48;58;06

But, you know, the first time I used this,

like I was ready to be unimpressed.

00;48;58;15 - 00;49;01;07

And, you know, it was magical.

It was cool.

00;49;01;07 - 00;49;02;22

It was fun.

00;49;02;22 - 00;49;04;05

It was like extraordinary.

00;49;04;05 - 00;49;07;25

Like, first thing I did, like,

write me a 5000 word book proposal,

00;49;08;25 - 00;49;12;25

justifying a book about AI

at the end of 2022 and stuff.

00;49;12;25 - 00;49;16;21

And, you know, two things it's far better

read than I am has a far better memory,

00;49;18;10 - 00;49;19;13

that that I do.

00;49;19;13 - 00;49;21;20

But the other thing, it's it was flat.

00;49;21;20 - 00;49;24;20

I mean, I couldn't like, hit,

you know, print and turn it in,

00;49;24;24 - 00;49;26;02

you know, kind of thing.

00;49;26;02 - 00;49;31;03

But it was like, it took like, what would

what ended up taking me a week, you know,

00;49;31;03 - 00;49;35;09

it produced a serviceable proposal

within a minute.

00;49;35;27 - 00;49;40;08

And, you know, just like its ability

to write poems, one of my favorite

00;49;40;08 - 00;49;44;25

early uses was in this kind of underscores

at its original is generative AI.

00;49;44;25 - 00;49;47;17

It's creating something original.

00;49;47;17 - 00;49;49;29

Explain Marx's Marx's

00;49;49;29 - 00;49;53;12

economic theory,

in the form of a Taylor Swift song.

00;49;53;25 - 00;49;55;06

And it was really good.

00;49;55;06 - 00;49;57;23

It was really clever at doing it.

00;49;57;23 - 00;49;59;05

Obviously, that was original.

00;49;59;05 - 00;50;02;19

I don't think they went out and found

that somebody had done the,

00;50;03;09 - 00;50;04;25

before and produce it.

00;50;04;25 - 00;50;07;11

So, you know, I, I kind of did start with,

00;50;08;20 - 00;50;10;24

this stuff being

00;50;10;24 - 00;50;14;01

magical, you know,

I could be talking to you right now.

00;50;14;01 - 00;50;18;10

I don't understand Italian, but

we could translate my words with my voice.

00;50;18;14 - 00;50;21;26

And I could do this interview in Italian

or Chinese or, you know, just just,

00;50;22;00 - 00;50;24;28

you know, any of a dozen, two

00;50;24;28 - 00;50;28;08

dozen languages, probably by this point,

100, plus languages.

00;50;28;16 - 00;50;32;05

And so, yeah, I do think it's

an extraordinary technology.

00;50;32;05 - 00;50;34;02

That was kind of fun for me

because, again,

00;50;34;02 - 00;50;36;17

you know, as a journalist,

you kind of have ideas in your head.

00;50;36;17 - 00;50;37;27

I think is a good journalist.

00;50;37;27 - 00;50;42;03

You should be willing to say,

oh, what I had in my head was wrong.

00;50;42;06 - 00;50;42;26

And that just like,

00;50;42;26 - 00;50;46;15

pursue examples that show like, oh,

this thing is, you know, lousy.

00;50;46;20 - 00;50;49;26

So it was fun for me to like,

kind of have my world turned upside down.

00;50;50;13 - 00;50;54;21

From the start, as I was going out

and introducing myself to founders,

00;50;54;21 - 00;50;58;12

venture capitalists, you know,

those trying to cash in on this I moment.

00;50;59;20 - 00;51;00;06

Yeah.

00;51;00;06 - 00;51;03;06

No, it's it's it's super, super neat.

00;51;03;11 - 00;51;05;10

And yeah, I it's fair.

00;51;05;10 - 00;51;08;10

I think it's,

it came as a surprise to all of us.

00;51;08;17 - 00;51;13;09

I do, Gary want to slightly switch gears

and talk about back to the human

00;51;13;09 - 00;51;17;23

in the loop and, you know, 2025

and maybe we can talk about AI

00;51;17;23 - 00;51;18;27

or maybe not, but

00;51;18;27 - 00;51;22;18

what I, what I wanted to ask you about is,

I think about six years ago,

00;51;22;18 - 00;51;27;08

you wrote a book series called,

Masters of Work on a few different.

00;51;27;08 - 00;51;31;14

You know, what you considered really

interesting, important roles at the time.

00;51;31;14 - 00;51;34;10

You did venture capitalists,

you did cybersecurity.

00;51;34;10 - 00;51;36;09

You did sports agent. Right.

00;51;36;09 - 00;51;39;13

If if you were to do that again

in 2025 or,

00;51;39;13 - 00;51;42;14

you know, 2026,

has that changed for you or there?

00;51;42;14 - 00;51;44;21

Are there new roles that you found

in kind of your research

00;51;44;21 - 00;51;46;13

and what you've been pressing up against

00;51;46;13 - 00;51;49;13

that you think are going to be really,

really interesting in the next few years.

00;51;49;16 - 00;51;51;00

So that was part of a bigger series.

00;51;51;00 - 00;51;54;00

And those are the three I

there were like 20 or so,

00;51;54;19 - 00;51;58;01

professions,

you know, I'll answer it differently.

00;51;58;01 - 00;52;01;17

Like I think something

that's so interesting is,

00;52;01;29 - 00;52;06;06

you know, what's going to

be the impact of AI on jobs.

00;52;07;13 - 00;52;10;05

Already we're seeing AI,

00;52;10;05 - 00;52;12;10

being very productive,

00;52;12;10 - 00;52;17;10

in customer service, centers,

you know, this idea that, you know,

00;52;17;11 - 00;52;21;18

the I could do a perfectly good job

of answering basic questions, I hadn't.

00;52;21;21 - 00;52;25;07

I need to reset my,

you know, password kind of, thing.

00;52;25;07 - 00;52;28;22

So, you know, a lot there's millions

of people who work in customer service.

00;52;29;02 - 00;52;32;02

It's going to take a big, chunk

chunk out of that.

00;52;32;17 - 00;52;35;17

One of the more,

there's 8 to 10 million people,

00;52;36;04 - 00;52;39;12

who make a living in this country,

as drivers, long haul

00;52;39;12 - 00;52;42;11

drivers, Uber drivers, taxi drivers,

local delivery.

00;52;42;16 - 00;52;45;06

You know,

how far away is autonomous vehicles?

00;52;45;06 - 00;52;47;11

I mean,

I was just out in San Francisco and,

00;52;47;11 - 00;52;50;07

you know, there's,

you know, robo taxis everywhere.

00;52;50;07 - 00;52;52;14

And it's, you know, it's working there

in a few different cities.

00;52;52;14 - 00;52;56;03

So I can't predict it for two,

five, ten years out.

00;52;56;10 - 00;52;59;10

But at some point in the foreseeable

future,

00;52;59;11 - 00;53;04;10

you know, all those jobs, as drivers

or most of those,

00;53;04;10 - 00;53;07;29

most of the delivery people that carry it,

unless we have a robot in the car,

00;53;07;29 - 00;53;10;06

you know,

bringing in the the stuff they delivering.

00;53;11;07 - 00;53;12;20

And so, you know,

00;53;12;20 - 00;53;15;20

there be job categories

that are decimated.

00;53;16;06 - 00;53;21;01

But I but there will be these

new categories that we can understand.

00;53;21;06 - 00;53;24;06

One of the obvious ones, right, right now

is prompt engineer.

00;53;24;10 - 00;53;29;02

You know, it's not is you have to

something I think people don't use

00;53;29;02 - 00;53;33;24

I don't understand

is that you're still the creative.

00;53;34;02 - 00;53;39;13

You're still the one who's needs

to find the use in this thing.

00;53;39;23 - 00;53;40;10

You know it.

00;53;40;10 - 00;53;42;22

Use can like,

tell me something interesting.

00;53;42;22 - 00;53;47;07

You know, you you know, I would I use it,

I use my knowledge

00;53;47;07 - 00;53;51;12

and you know, I've done a little bit

research and I can guide it and stop.

00;53;51;19 - 00;53;55;20

And so prompt engineer is going to be

or is a job.

00;53;55;20 - 00;53;58;18

But there's all these job categories

that we can't imagine

00;53;58;18 - 00;54;00;05

I might make the example I use.

00;54;00;05 - 00;54;03;05

It's, you know,

Gutenberg comes up with the printing press

00;54;03;12 - 00;54;04;29

and, you know,

we could worry about like, oh,

00;54;04;29 - 00;54;08;06

where are all these monks going to do,

how they can earn a living kind of thing.

00;54;09;03 - 00;54;11;25

It gave rise to the publishing industry.

00;54;11;25 - 00;54;16;10

It gave rise to newspapers and magazines,

like there's millions of people, you

00;54;16;10 - 00;54;21;03

know, who have jobs based on essentially

this technology in the printing press.

00;54;21;12 - 00;54;24;14

And so I can't help but think that

00;54;24;28 - 00;54;28;15

AI is going to create job categories

that our brains

00;54;28;15 - 00;54;32;29

can't quite understand,

because we need to see how it plays out.

00;54;32;29 - 00;54;35;24

I do fear that AI is going to be,

00;54;36;26 - 00;54;38;23

is going to lead to

00;54;38;23 - 00;54;42;13

the destruction of more jobs

than the creation of new jobs.

00;54;42;17 - 00;54;43;28

But I could be wrong.

00;54;43;28 - 00;54;46;29

I think I said the same thing

in the mid 1990s.

00;54;47;11 - 00;54;49;01

And the internet. Right?

00;54;49;01 - 00;54;52;00

I mean, you need less people

at the Amazon bookstore

00;54;52;05 - 00;54;54;24

than you would at the local, bookstore.

00;54;54;24 - 00;54;58;11

So in my head like, oh my goodness,

the internet is going to destroy

00;54;58;24 - 00;54;59;22

all these jobs.

00;54;59;22 - 00;55;03;02

But, you know,

we have 4% unemployment right now.

00;55;03;02 - 00;55;05;21

So somehow we've we figured that out.

00;55;05;21 - 00;55;08;07

But, you know, I, I am scared of that.

00;55;08;07 - 00;55;11;07

And it's actually happens to be something

I'm writing about a research

00;55;11;09 - 00;55;14;15

now because I,

I don't know the answer in my head.

00;55;14;15 - 00;55;18;14

I again I fear, is to be largely,

a big net negative.

00;55;19;07 - 00;55;22;07

But I've had that fear

before that have been wrong.

00;55;22;20 - 00;55;24;22

Well, so can you tell me a little bit more

about that?

00;55;24;22 - 00;55;27;17

What? Your, your next project

and what you're looking into?

00;55;28;20 - 00;55;29;13

I mean,

00;55;29;13 - 00;55;32;04

this is just an this is just said article

00;55;32;04 - 00;55;35;04

I'm working on for the New York Times.

00;55;35;15 - 00;55;38;15

But, you know, I,

you know, if you look at my career,

00;55;39;20 - 00;55;43;13

I've written about tech

or I've written about,

00;55;44;11 - 00;55;48;17

politics, urban issues,

and of course, what's been happening

00;55;48;17 - 00;55;52;20

the last year or so is,

you know, tech is eating politics.

00;55;52;20 - 00;55;54;05

Tech is at the center of politics.

00;55;54;05 - 00;55;57;21

I would argue that tech is the largest

00;55;58;05 - 00;56;01;01

special interest corporate interests,

but whatever you want to call it,

00;56;01;01 - 00;56;05;01

like in the history of politics,

like in the old days, you know,

00;56;06;14 - 00;56;09;14

the rich moguls would write big checks,

00;56;09;28 - 00;56;14;21

and you know, kind of have their

politicians, they treated these puppets.

00;56;14;28 - 00;56;15;26

But this is different.

00;56;15;26 - 00;56;19;20

I'm first off, you know,

look at Elon Musk, 250 million or so

00;56;19;27 - 00;56;23;27

crypto

put like $200 million into the 2024,

00;56;25;00 - 00;56;25;22

election.

00;56;25;22 - 00;56;29;26

They're running much, much,

much, much, much bigger checks.

00;56;29;26 - 00;56;32;21

And so somehow I wonder if there's,

00;56;32;21 - 00;56;33;23

a book, a book there.

00;56;33;23 - 00;56;35;15

The billionaire Boys club, I did.

00;56;35;15 - 00;56;37;09

Gary, want to talk to you

about something else outside

00;56;37;09 - 00;56;41;04

of, tech and politics

that you kind of touched on earlier,

00;56;41;20 - 00;56;45;05

but both in your career and with, like

the Gutenberg press example, which is,

00;56;45;24 - 00;56;49;26

journalism

and how journalism fits into all of this

00;56;49;26 - 00;56;54;16

and what I means for journalism

and journalism.

00;56;54;16 - 00;56;56;21

I mean,

I don't think I'm being controversial

00;56;56;21 - 00;56;59;21

by saying it's

in a pretty precarious place right now.

00;57;00;10 - 00;57;03;10

And has seen,

00;57;03;18 - 00;57;06;18

in my eyes, like a rapid and depressing,

00;57;06;20 - 00;57;10;06

you know, decline in the past

decade or more.

00;57;11;06 - 00;57;13;03

What does all of this mean for journalism?

00;57;13;03 - 00;57;16;04

And if you had, like,

well, let me maybe just leave it there.

00;57;16;14 - 00;57;20;09

Yeah, I so is that this is been going on

00;57;20;19 - 00;57;23;26

long before I grew front and center.

00;57;23;26 - 00;57;25;08

In fact, you can make the argument

00;57;25;08 - 00;57;28;28

that it was the internet,

you know, the internet.

00;57;29;03 - 00;57;31;28

Craigslist wiped away

classified advertising.

00;57;31;28 - 00;57;35;29

That was a big piece, of of, the revenue

00;57;36;07 - 00;57;39;28

for any, newspaper,

you know, a whole set of things.

00;57;40;01 - 00;57;43;08

You know, I, I fear accelerates that.

00;57;43;08 - 00;57;46;29

I mean, again,

I think I could be a, a good tool

00;57;47;08 - 00;57;50;05

for a journalist

to help him or her do her job.

00;57;50;05 - 00;57;52;20

Job. But, you know,

00;57;52;20 - 00;57;56;06

I, I do fear, in fact,

we're already seeing it.

00;57;56;17 - 00;57;59;29

That AI is good enough right now

00;58;00;13 - 00;58;03;18

where it could write a basic news story.

00;58;03;26 - 00;58;04;10

You know,

00;58;04;10 - 00;58;08;00

kind of that one of the coins of the realm

these days is like the heartache.

00;58;08;09 - 00;58;12;02

Well, I could just digest

what's being said about,

00;58;12;11 - 00;58;16;05

you know, this, this event or that,

and you train it.

00;58;16;05 - 00;58;18;21

It could come up with a hot

take much faster.

00;58;18;21 - 00;58;23;11

And many, many, many, many more than any

human, Sports Illustrated.

00;58;23;12 - 00;58;24;23

They got in trouble for it.

00;58;24;23 - 00;58;28;08

But they created

these bots, even gave them

00;58;28;08 - 00;58;31;20

names and backstories,

as if they were real humans.

00;58;31;20 - 00;58;37;00

They were deceptive, about it,

but they're using them to write stories.

00;58;37;07 - 00;58;38;18

CNet the same thing.

00;58;38;18 - 00;58;43;00

They they did it, but they, they stopped

because they found that there, you know,

00;58;43;02 - 00;58;47;06

half the stories or more, had errors,

some of them serious errors.

00;58;47;06 - 00;58;52;21

So again, I come back to this idea like,

okay, we could use AI bots.

00;58;52;21 - 00;58;56;24

I think it's dangerous

to create a rough draft of a story,

00;58;57;07 - 00;59;00;26

but there has to be humans in the loop

who are skeptical,

00;59;01;01 - 00;59;02;26

who are like, wait,

what's the source for this?

00;59;02;26 - 00;59;06;16

When I use, you know, a chat

bot to do research,

00;59;06;23 - 00;59;09;20

you know, I, I use one of them

like perplexity as an example

00;59;09;20 - 00;59;14;10

that footnotes because I, you know,

I'm not going to trust anything it says.

00;59;14;12 - 00;59;17;13

Hey, most of what it says is right,

but most of what it says is

00;59;17;13 - 00;59;18;12

right is not good enough.

00;59;19;11 - 00;59;21;07

So I always want to go, oh, okay.

00;59;21;07 - 00;59;26;07

That quote here's a CNBC article

that had that quote, in it.

00;59;26;14 - 00;59;30;11

And so, you know, it can be a tool, but,

you know, I'm scared that,

00;59;30;17 - 00;59;33;17

you know, with all the pressure on,

00;59;34;06 - 00;59;37;08

news outlets to cut costs

because we're getting,

00;59;37;08 - 00;59;41;20

you know, they're getting less revenue

from subscriptions, less revenue

00;59;41;20 - 00;59;44;20

from classified ads,

less revenue for advertising.

00;59;45;21 - 00;59;50;14

That their solution is going to be,

you know, kind of reporter bots and all.

00;59;50;14 - 00;59;52;09

But, you know,

00;59;52;09 - 00;59;54;21

who's going to go to the meeting,

00;59;54;21 - 00;59;57;05

the city council meeting

and get the quotes.

00;59;57;05 - 01;00;01;09

I mean, I still have to think

that needs to be journalists

01;00;01;09 - 01;00;04;24

out there doing this shoe

leather reporting, and all.

01;00;04;24 - 01;00;09;15

But, you know, big companies are always

looking for ways to, to save money.

01;00;09;17 - 01;00;14;23

You know, maybe we'll have reporters, but

the editors and copy editors, they'll be.

01;00;15;01 - 01;00;19;07

But so I, I think there'll be a profound

in, impact of AI.

01;00;19;07 - 01;00;20;22

In fact, one more data point.

01;00;20;22 - 01;00;25;27

And again, this is an I this is before I,

I, you know, newspapers and magazines

01;00;25;27 - 01;00;31;08

in the US spreading like $50 billion

in advertising collectively in 2000.

01;00;31;16 - 01;00;34;00

But Google and Facebook now matter.

01;00;34;00 - 01;00;34;11

You know,

01;00;34;11 - 01;00;35;05

they have even

01;00;35;05 - 01;00;38;11

huge portions of that to the point

where it's now closer to 10 billion

01;00;38;11 - 01;00;39;17

rather than 50 billion.

01;00;39;17 - 01;00;42;14

That's a lot of money

that they're no longer getting.

01;00;42;14 - 01;00;43;29

And so how do you make it work?

01;00;43;29 - 01;00;47;14

How do you, as a newspaper, magazine

or publication, make it work?

01;00;47;25 - 01;00;50;18

You know, AI is a way you can save money,

01;00;52;11 - 01;00;54;21

right? And

01;00;54;21 - 01;00;55;20

it makes sense.

01;00;55;20 - 01;00;58;02

And I really like,

you know, your take on it

01;00;58;02 - 01;00;59;01

and what you covered

01;00;59;01 - 01;01;01;09

that there's one like specific piece

that I want to dig

01;01;01;09 - 01;01;04;27

in a little bit deeper, which is

specifically investigative journalism,

01;01;05;04 - 01;01;08;07

which is something that

that you've focused on in your career.

01;01;08;15 - 01;01;13;19

And to me is, I don't know,

maybe the piece that's least

01;01;13;19 - 01;01;17;22

replicable by AI, either

because they can't do it or because,

01;01;17;22 - 01;01;20;24

you know, the news outlets of the future

don't necessarily want them to do it.

01;01;21;06 - 01;01;24;28

Is that like, what is the future

of investigative journalism

01;01;24;28 - 01;01;27;03

and is it in trouble in your eyes?

01;01;27;03 - 01;01;27;24

Well, investigative

01;01;27;24 - 01;01;31;17

journalism is in trouble generally

because it's very expensive, right?

01;01;31;17 - 01;01;33;02

Yeah.

01;01;33;02 - 01;01;37;09

I could write two news stories a day

if I'm just go to a city

01;01;37;09 - 01;01;40;25

council meeting or I go press release,

make a couple of calls and crank it out.

01;01;40;25 - 01;01;43;29

You know, I've,

I've spent months on investigate of,

01;01;45;00 - 01;01;45;29

pieces.

01;01;45;29 - 01;01;48;29

So investigative reporting is expensive.

01;01;49;01 - 01;01;51;11

Generally,

there's been kind of a nonprofit model,

01;01;51;11 - 01;01;54;11

you know, model where like,

oh, ProPublica.

01;01;54;16 - 01;01;58;06

You know, we'll have an outlet

that does investigative journalism.

01;01;58;06 - 01;01;59;27

That's their thing. And they're, you know,

01;01;59;27 - 01;02;03;09

funded by foundations and such and,

you know, use users.

01;02;04;07 - 01;02;06;14

Who appreciate their, their work.

01;02;06;14 - 01;02;10;21

You know, I, I so I don't think

that's reversing anytime soon.

01;02;10;21 - 01;02;14;13

In fact,

I fear that it's going to be, you know,

01;02;14;13 - 01;02;19;06

less and less what media outlets are able

to, invest in.

01;02;19;06 - 01;02;22;15

But, you know, to kind of express

an optimistic note for a moment like,

01;02;22;28 - 01;02;25;15

hey, I could be an amazing tool.

01;02;25;15 - 01;02;29;11

One one thing that I so good at

is going through vast

01;02;29;27 - 01;02;33;09

sets of data

and finding connections that no human,

01;02;33;21 - 01;02;37;25

you know, could, could, could see and so,

you know, the SEC,

01;02;37;25 - 01;02;41;23

you know, there's a gazillion filings

from publicly traded companies.

01;02;42;00 - 01;02;45;20

You could train

AI to look for patterns like, you know,

01;02;45;28 - 01;02;50;20

look for times where there was big news

announced by a company

01;02;50;20 - 01;02;53;25

who sold stock inside the company

ahead of that, you know,

01;02;53;25 - 01;02;56;25

just to use a quick example

off the top of my head.

01;02;56;27 - 01;03;00;08

But so, you know,

I could be an extraordinary tool again,

01;03;01;00 - 01;03;04;08

as long as there's a writer

or a set of writers

01;03;04;14 - 01;03;07;07

who are using that tool,

because it's the human instinct

01;03;07;07 - 01;03;08;27

that's going to understand, like

01;03;09;27 - 01;03;12;27

where the story is, you know, these

these things that,

01;03;13;02 - 01;03;16;15

you know,

we're at GPT 4.5, you know, right now.

01;03;16;18 - 01;03;20;18

So I don't know what GPT 8 or 9 or 10

or 12 is going to be like.

01;03;20;18 - 01;03;24;22

But these things cannot write

a good investigative story.

01;03;24;24 - 01;03;27;07

These things could not write

a good magazine story.

01;03;27;07 - 01;03;29;04

These things can't write a book.

01;03;29;04 - 01;03;31;10

And they seem a long, long,

long way that you flat.

01;03;31;10 - 01;03;31;27

They're boring.

01;03;31;27 - 01;03;33;15

They're, you know,

01;03;33;15 - 01;03;38;13

as you point out before, they're just

mathematical models looking for patterns.

01;03;40;00 - 01;03;42;22

And so it's derivative by definition.

01;03;42;22 - 01;03;45;26

What makes a book good is its original.

01;03;45;26 - 01;03;49;06

I would argue that the human toil

and sweat

01;03;49;18 - 01;03;52;17

is what makes it good.

01;03;52;17 - 01;03;56;20

It's the sense

the approach, of of the writer.

01;03;56;27 - 01;03;58;23

But, you know, I, I don't know.

01;03;58;23 - 01;04;01;24

I mean, I do wonder

if these things are getting them

01;04;02;07 - 01;04;05;07

ten times better, more or less every year.

01;04;05;09 - 01;04;08;21

Will there ever be a point

where could write a beautiful novel,

01;04;08;25 - 01;04;12;25

will ever get a point where it could,

like, write a really engaging, page

01;04;12;25 - 01;04;17;27

turning work of nonfiction that's told

through characters like replacing me,

01;04;18;08 - 01;04;19;21

I, I don't I,

01;04;19;21 - 01;04;20;16

you know, I, I joke,

01;04;20;16 - 01;04;24;06

I'm writing as fast as I can because

I don't know if that's going to happen,

01;04;24;06 - 01;04;26;03

but I definitely think it's possible

it's going to happen.

01;04;27;07 - 01;04;27;11

Yeah.

01;04;27;11 - 01;04;29;24

What do you think? No, it's it's

what do you think about it?

01;04;29;24 - 01;04;31;15

I mean,

01;04;31;15 - 01;04;33;09

I first of all, I was going to say

01;04;33;09 - 01;04;37;17

I really appreciate the note of optimism

there about what it can do.

01;04;37;17 - 01;04;39;07

And, you know, I was I was just,

01;04;39;07 - 01;04;42;28

you know, as the wheels were turning,

as you were talking, I don't know, I think

01;04;44;00 - 01;04;46;15

it wouldn't shock me if at some point

01;04;46;15 - 01;04;51;00

I could get to writing articles,

if not books, maybe books

01;04;51;00 - 01;04;53;12

that the thing that you've probably seen

this too, in your research,

01;04;53;12 - 01;04;58;07

one of the limitations right now

is that the longer you let generative

01;04;58;07 - 01;04;59;12

AI generate,

01;04;59;12 - 01;05;03;04

the more likely it is to go off the rails

and go in some wacky direction.

01;05;03;11 - 01;05;06;11

And the analogy, by the way, that

I kind of thought of for this is like,

01;05;06;28 - 01;05;10;00

I feel like that's true

of like the human body, right?

01;05;10;00 - 01;05;11;20

Like once a human,

01;05;11;20 - 01;05;13;04

you know, turns 100, like things

01;05;13;04 - 01;05;15;19

start going wrong with the body

because it's been running for too long

01;05;15;19 - 01;05;17;12

and all those little things

start to add up.

01;05;17;12 - 01;05;19;07

Generative AI, I guess, is the same way.

01;05;20;10 - 01;05;21;14

So, so long

01;05;21;14 - 01;05;24;24

form I think

is probably safe for a little bit, but

01;05;24;24 - 01;05;28;08

I, I've certainly found in experimenting

with it, Gary, that you can ask it like,

01;05;28;09 - 01;05;33;03

you know, be compelling and in a short way

it can be more compelling and less flat.

01;05;33;15 - 01;05;37;16

But you know, the piece

that is interesting and gives me hope.

01;05;37;16 - 01;05;41;03

There is more on the investigative

journalism piece

01;05;41;11 - 01;05;44;28

and the, the, the tool for it

to be able to make those connections.

01;05;45;11 - 01;05;48;07

And you said it,

01;05;48;07 - 01;05;51;16

but it's really only as good

as the prompter is, right?

01;05;51;16 - 01;05;51;23

Right.

01;05;51;23 - 01;05;55;17

Like, yeah, it

it needs a human to ask the good questions

01;05;55;21 - 01;05;57;15

and give you the answer

to that good question.

01;05;57;15 - 01;05;58;27

And I mean, I tell you anecdotally.

01;05;58;27 - 01;06;01;19

It needs human to digest it.

I mean, that's right.

01;06;01;19 - 01;06;05;16

It'll give you a vast amount of data

like and it's your job as the writer,

01;06;05;16 - 01;06;06;29

the journalist, the human. Yeah.

01;06;06;29 - 01;06;10;27

You know, to kind of figure out like,

oh, here's here's my story again.

01;06;10;27 - 01;06;12;02

I, I'm always playing with it.

01;06;12;02 - 01;06;15;02

I'm always like,

you know, yeah, here's here's my notes.

01;06;15;18 - 01;06;17;03

Give me a good opening.

01;06;17;03 - 01;06;20;04

I'm like basically,

oh for every time doing that.

01;06;20;04 - 01;06;23;06

But you know, it's like, oh,

but that's an interesting word to use in.

01;06;23;06 - 01;06;23;25

Oh right.

01;06;23;25 - 01;06;25;29

Good place to start.

I hadn't really thought of that.

01;06;25;29 - 01;06;26;09

Yeah.

01;06;26;09 - 01;06;30;23

So again this is you know you know

Microsoft calls their chat bot copilot.

01;06;32;01 - 01;06;33;13

And I think that's the perfect

01;06;33;13 - 01;06;37;16

what I like about that

is it's the name tells you what it is.

01;06;37;27 - 01;06;39;29

You know, you're flying the plane.

01;06;39;29 - 01;06;41;25

You're the one who knows this,

01;06;41;25 - 01;06;45;22

but you got this really good assistant

right there that could do some work.

01;06;45;22 - 01;06;46;24

You know, I have to go to the bathrooms.

01;06;46;24 - 01;06;49;05

Okay, you fly for a minute, but just for,

you know, a little bit.

01;06;49;05 - 01;06;50;24

I'm. I'm learning this thing.

01;06;50;24 - 01;06;53;22

I'm taking it off.

01;06;53;22 - 01;06;56;03

And so, you know, it is your copilot.

01;06;56;03 - 01;06;59;22

And again, whether it's, we're talking

about investigative journalists or,

01;07;00;03 - 01;07;05;18

you know, more or less

any creative, professional, job,

01;07;05;18 - 01;07;09;27

it really is this amazing, tool

like reports, right?

01;07;09;27 - 01;07;14;03

It could write a really good,

efficient 20 page report

01;07;14;10 - 01;07;18;06

that reflects more or less

every 20 page report that's written on a,

01;07;18;07 - 01;07;21;09

you know, white paper

like thing on a basic subject.

01;07;21;09 - 01;07;25;26

So if I was in that that's

the kind of writer I was, two things.

01;07;25;26 - 01;07;28;23

One, I'd be using and again,

you're steering it.

01;07;28;23 - 01;07;30;06

It's you're the prompt engineer.

01;07;30;06 - 01;07;33;27

You're the one who is going to have

to figure out how this is useful.

01;07;34;03 - 01;07;36;19

You need to edit it,

you need to rewrite it.

01;07;36;19 - 01;07;37;06

And then yesterday.

01;07;37;06 - 01;07;40;09

But you know, instead of it

taking two weeks for write that report,

01;07;40;13 - 01;07;41;28

maybe you could do it in 3 or 4 days.

01;07;41;28 - 01;07;43;18

But the other thing

I'd be thinking is like,

01;07;43;18 - 01;07;45;29

I might want to find a different career,

01;07;45;29 - 01;07;49;17

because if my big company has 20 of us

who writes reports,

01;07;50;01 - 01;07;53;01

it's not too hard to imagine in the

not so distant

01;07;53;01 - 01;07;56;05

future that instead of 20,

they might have 2 or 4 of us.

01;07;56;17 - 01;07;57;26

Yeah, yeah.

01;07;57;26 - 01;07;59;16

No, that's that, that's well said.

01;07;59;16 - 01;08;01;14

But I do love the,

01;08;01;14 - 01;08;04;28

I just wanted to kind of close that out

with the notion that,

01;08;05;08 - 01;08;06;11

you know, to me,

01;08;06;11 - 01;08;09;15

and I don't think you'll find

this controversial at all, like journalism

01;08;09;15 - 01;08;14;00

and especially investigative

journalism are a public good, right?

01;08;14;00 - 01;08;16;10

Like they do

something of benefit to society.

01;08;16;10 - 01;08;19;09

And I worry about a future

where they're not there.

01;08;19;09 - 01;08;22;28

And you, you've given me hope

that, you know, whether it's that

01;08;22;28 - 01;08;26;29

or some of the other kind of creative

tools for it, they're not going away.

01;08;26;29 - 01;08;30;18

And it can actually be a force

for good as we look at what it can do.

01;08;30;19 - 01;08;31;22

Yeah, I hope so.

01;08;31;22 - 01;08;34;06

But, you know, there

is that other trend line that, you know,

01;08;34;06 - 01;08;35;28

we don't have enough investigative

journalists.

01;08;35;28 - 01;08;39;13

There's, you know, most local newspapers,

01;08;39;23 - 01;08;42;24

you know, have few,

if any, investigative journalists.

01;08;42;24 - 01;08;46;12

But, you know, investigative

journalists have uncovered extraordinary,

01;08;47;07 - 01;08;49;12

thing and Watergate,

01;08;49;12 - 01;08;52;20

Jeffrey Epstein,

you know, kind of the list is long.

01;08;52;20 - 01;08;55;28

And I think at this moment, getting back

to a point you made before, like,

01;08;56;09 - 01;08;59;05

you know, Elon

Musk could say what he wants to say

01;08;59;05 - 01;09;02;13

and stuff, but investigative journalists

are the ones who figure out

01;09;02;13 - 01;09;07;07

in the New York Times, he's done

a great job of this saying like, wait,

01;09;07;21 - 01;09;11;16

but there's all these investigations

that the federal government, and say,

01;09;11;16 - 01;09;12;09

how does that work out?

01;09;12;09 - 01;09;17;01

Like, wait, you get contracts, in fact,

with, you know, rural broadband.

01;09;17;01 - 01;09;19;13

Like there's I don't know where it is.

01;09;19;13 - 01;09;22;13

It's not my topic,

but there's talk of Starlink

01;09;22;13 - 01;09;26;02

that's, you know, a must,

you know, Musk, product,

01;09;26;14 - 01;09;30;15

you know, of, of Starlink

being brought in, to give us,

01;09;31;26 - 01;09;32;26

rural

01;09;32;26 - 01;09;35;26

broadband, bringing broadband

to rural areas.

01;09;36;02 - 01;09;39;08

Well, not broadband, bringing us,

you know, wireless, bringing Starlink

01;09;39;08 - 01;09;44;23

instead like that could be worth billions

and billions of dollars, to us.

01;09;44;23 - 01;09;47;01

So, you know, is that what's happening?

01;09;47;01 - 01;09;50;20

I don't know, but I think it's

an investigative journalist role

01;09;51;00 - 01;09;55;01

to unearth that and say, like, okay,

I don't mean to be a cynic here,

01;09;55;01 - 01;09;58;01

but there is

this other set of motivations,

01;09;58;02 - 01;09;59;09

that the public should know about

01;09;59;09 - 01;10;02;12

when they listen to

what Elon Musk says about X or Y.

01;10;03;04 - 01;10;06;27

Well, and if the biggest barrier to that

right now is while it's just it's cost

01;10;06;27 - 01;10;09;14

prohibitive

to have too many investigative journalists

01;10;09;14 - 01;10;12;02

for a local paper or for any paper,

and we say, well, wait,

01;10;12;02 - 01;10;15;16

I can decrease the cost

of an investigative journalism

01;10;15;16 - 01;10;19;02

piece by,

you know, ten x suddenly, you know,

01;10;19;25 - 01;10;23;14

in my utopian world

that's back on the table.

01;10;23;14 - 01;10;24;25

And that could be really interesting.

01;10;24;25 - 01;10;27;22

You know, ten x might be a little bit

too utopian for my taste here.

01;10;27;22 - 01;10;28;27

Fair enough. But, you know.

01;10;28;27 - 01;10;30;25

I can dream. Yeah.

01;10;33;05 - 01;10;36;03

But, you know,

I, I keep on coming back to this term,

01;10;36;03 - 01;10;40;01

I, I stole it from Reed Hoffman,

you know, kind of superpower that,

01;10;40;01 - 01;10;44;21

you know, you

it does amplify your intelligence.

01;10;44;21 - 01;10;47;11

It does amplify your ability to do things.

01;10;47;11 - 01;10;48;21

And, you know,

01;10;48;21 - 01;10;52;00

I'll just kind of restate something

I said earlier, like investigative

01;10;52;00 - 01;10;55;22

journalists using AI are more powerful

01;10;55;22 - 01;10;58;29

than investigative journalists

who don't use AI.

01;10;59;00 - 01;11;02;09

So I do think that's kind of

01;11;02;10 - 01;11;05;10

not an equalizer,

but I do think it can help.

01;11;05;25 - 01;11;08;14

You know, what's what's a

01;11;08;14 - 01;11;09;10

help preserve?

01;11;09;10 - 01;11;12;28

What's a diminishing, resource

in our, in our country.

01;11;13;16 - 01;11;14;22

Well said.

01;11;14;22 - 01;11;17;07

Gary, I want to say a big

thank you for today.

01;11;17;07 - 01;11;19;03

This has been really, really fascinating.

01;11;19;03 - 01;11;21;23

A lot of really interesting insights

I'll be taking away.

01;11;21;23 - 01;11;24;21

So I really appreciate you coming on

and the time you spent here today.

01;11;24;21 - 01;11;26;14

Oh my pleasure.

Thanks for having me. This is fun.

01;11;28;06 - 01;11;31;03

This. Is.