All language subtitles for VICE.S07E00.Special.Report.The.Future.of.Work.2019.720p.AMZN.WEB-DL.DDP2.0.H.264-TEPES

♪ ♪

(rattling)

(Artificial Intelligence
voice speaking)

We're riding in the back
of an autonomous truck.

It's about to take a left turn
into oncoming traffic.

So, if their sensors--

(AI voice speaks)

Andavolu:
It's accelerating on its own.
It's breaking on its own.

It's steering on its own.

Trucking in
the United States is, like,

a $700-billion-a-year industry,

and it employs
1.8 million people.

Thus far, it's been pretty

immune to the changes of
globalization and technology,

but that's about to change
with technology like this.

(AI voice speaking)

♪ ♪

Andavolu:
There's an operator here,

there's a safety engineer here,

'cause they're
still experimenting.

It's still kind
of in development.

(AI voice speaks)

Andavolu:
Have you put your hand
on the wheel at any point?

(operator speaks)

Andavolu:
This is like a pretty complex
traffic situation.

There are cars merging.
There are cars passing us.

This truck is changing lanes
to go around slow traffic,

anticipating when
people are stopping.

And honestly, I just, like,
kind of can't believe it.

It's... It's driving itself,

and it's doing
a pretty good job.

(AI voice speaks)

Operator:
Here we go, making our
right turn into our driveway.

♪ ♪

Chuck Price:
We recognize that this is
a highly disruptive technology,

on the order of
10 million people,

and displacing
rapidly that many people

-would have a dramatic
societal impact.
-Mm-hmm.

We certainly
don't want to see that.
We're not targeting that.

We're focused
on relieving a shortage.

But what we're hoping
is that there will be
a natural evolution

into the jobs of the future,
just as there has been

in every other
technological change.

-What do you tell a trucker?
-We try not to tell truckers
things. We try to listen.

♪ ♪

(sizzling)

(patrons chattering)

This is Chuck.

-Hello, Chuck.
-Chuck runs a company that is

developing
and is gonna produce
self-driving trucks,

that are autonomous,
that do it themselves.

-You guys are truck drivers.
-Don Schrader: Yes, sir.

How do you feel about it?

If the truck doesn't
have a driver in it,

is somebody at headquarters
sitting behind a monitor?

Schrader:
No, there's
a driver in there.

No. In the future,

when there isn't, you know,
once it's fully tested...

-Oh.
-...and we don't need
to have a driver in it.

-Oh, driver assist?
-No, it's not driver assist.

It is purely self-driving.

So, like a GPS basically?

You're going from
point A to point B?

There's GPS in it.
There's a lot of tech,

-but GPS is a part of it.
-Okay.

Andavolu:
He took me into one
of his trucks today.

We did a 90-minute run
up and down on I-10.

Traffic was merging in.
It was changing lanes.

-It was braking, it was
accelerating. It was like--
-Schrader: Really?

So, does it do really good
when the cars cut you off?

A car cut us off!
And you know what?
It didn't slam on the brakes.

The truck just
kind of kept rolling.

Schrader:
How is your equipment
handling the high winds?

-We're actually surprisingly
good in high winds.
-Really?

How much weight
in the trailer?

We've gone empty and loaded.

-Empty and loaded?
-Yes, sir.

We have very complex
control algorithms,

and now we hold it
with such precision that

it is perfectly straight.

Yeah.
It's tough.

Andavolu:
You seem pretty impressed.

Actually, yeah, I am.

I wasn't all for it,
but, I mean, it's...

I gotta,
I gotta see this.

♪ ♪

Price:
That's laser.
That's lidar.

It's a laser,
like a laser radar.

Andavolu:
What do you love
about driving a truck?

Oh, gosh. (chuckles)

I love to drive.
For me, it's a privilege.

It's a privilege
to get out there,

get behind the wheel
of 80,000 pounds,

drive that thing down the road,
knowing, hey, you know what?

-I can do this,
you can't do it.
-No, I can't.

Schrader:
And I know
I'm providing the US.

I'm providing
the world with whatever

I got in the back
of my freight.

-I deliver your clothes,
your food that you're eating.
-Mm-hmm.

A lot of people
don't see that.

And it's a good
feeling as a driver,
as a human being.

Schrader:
You really gotta have faith

to rely on, is
this gonna kill me or...

Yep, and we'll have
to do a lot of testing

-to prove that.
-Wow.

All right.

That's definitely different.

Over the next year,
we're building a fleet
of 200 trucks,

and we're gonna be operating
them day and night,

just to validate it,
to prove it.

We have to prove to you.

We have to prove
to the regulators,
to the states.

We have to prove
to ourselves.

Schrader:
When you asked me
what would I do
if I didn't drive?

I can't honestly answer that

'cause I really don't know
what I would do.

I'd... I'd be scared.

♪ ♪

(beeping)

(engine rumbling)

♪ ♪

♪ ♪

Andavolu:
It's one of the first questions
that every kid gets asked.

What do you wanna be
when you grow up?

Previous generations
grew up confident

that no matter the answer,
they'd be something.

That's actually not
so certain anymore.

Automation already
affects most jobs,

but the pace of change

and the sheer capabilities
of artificial intelligence

are revolutionizing
our relationship to work,

something economists are
paying close attention to.

Technologies are tools.
They don't decide what we do.

We decide what we do
with those tools.

If we have more powerful tools,
by definition,

we have more power
to change the world
than we ever had before.

Andrew McAfee:
It took about 600 years

for global
average incomes

-to increase by 50%.
-Wow.

McAfee: Here we are,
from 1988 to now,

almost 50% or more increases
across the board of humanity.

This is because
of economic freedom

and because of
technological progress.

It just expands
the possibilities,

and therefore,
expands our income

so much more quickly
than we've ever seen before.

Joseph Stiglitz: The problem is
that as the economic pie's
gotten bigger,

not everyone has shared.

Wages at the bottom
today are the same,

-adjusted for inflation,
as they were 60 years ago.
-Mm-hmm.

So, all that growth

didn't go down to the people
at the bottom.

Andavolu:
The future of work has
even caught the attention

of experts like Richard Haass,

and the Council
on Foreign Relations,

who've authored
a report on the threats

it poses
to geopolitical stability.

Millions of jobs are
beginning to disappear.

You've got now
a whole new generation,

whether it's
artificial intelligence,

robotics or autonomous

or driverless vehicles
that are coming along,

that will displace millions
of workers in this country,

in the United States,
but also around the world.

And then suddenly,
these technologies come along,

and they destroy
our existing relationships.

The stakes for us
as individuals are enormous,

and unless
we can replace all that,
what's gonna come of us?

Now, you drop
a few drops of blood
in the shark tank...

-Krishna; Which is AI.
-...which is AI,

-and machine learning.
-Machine learning.

Goolsbee:
And we're gonna
shut down factories,

we're gonna replace
truck drivers.

What if everything
in the country's owned
by three people

who are the ones
who invented the robots...

-Jeff Bezos. Mark Zukerberg.
-...and we're all begging them
for a crumb.

What can I eat today? Please,
will you hand me something?

You can see why
people get upset.

Even thinking about it,
your blood is gonna boil.

The discontent that is evident
in the Brexit vote in 2016,

the vote for Trump in 2016.

What is very clear,
there's a lot of discontent.

A lot of people have not
been doing very well.

It isn't clear
how long we have

before the political system
comes under enormous stresses.

We as a society have not even
begun to have a sustained

or comprehensive
national conversation.

And what worries me is

by the time we really get
around to dealing with this,
it's gonna be too late.

(beeps)

(sizzling)

♪ ♪

(indistinct chatter)

Andavolu:
Over 3.5 million people
work in fast food.

It's one of the easiest
jobs to get,

and a good first step
on the ladder up to a job

with better pay
and less grease.

But at Southern
California's Caliburger,

Gianna Toboni found out even
this first step is in jeopardy.

Our vision is that we want
to take the restaurant industry,

and more broadly
the retail industry,

and make it operate
more like the internet.

We wanna automate
as much as we can,

and allow merchants
that operate

brick-and-mortar businesses
to see their customers

in the same way that Amazon
sees their customers.

-There you go. So, you're done.
-That was 10 seconds, probably.

(sizzling)

♪ ♪

Miller:
This was our first robot
to work on the grill.

The entire fleet of robots
that we deploy

learns from the training that
takes place here in Pasadena.

So, unlike humans, we teach
one robot and you perfect it,

and then you can
deploy that software

to all the robots in the field.

What are the advantages of
automating a business like this?

For a restaurant chain,
consistency is critical
to success.

Right? Critical to scale.

These restaurants will be
safer when you automate them.

Humans touching the food less.

Labor costs is a big
issue right now, right?

It's not just
rising minimum wage,
but it's turnover.

They come in,
they get trained,
and then they leave

to go drive an Uber
or do something else.

I notice you still have
some employees back here.

-It's not just Flippy
running the whole show.
-That's right. Yeah.

Miller:
So, we currently
think about this

as a cobotic
working arrangement,

where we have the robot
automating certain tasks

that humans don't
necessarily like to do.

But we still need people

in the kitchen managing
the robotic systems

and working side-by-side
with the robot to do things

that it's not possible
to automate at this
point in time.

Toboni:
Natalie, how do you
like working here?

I really like it.
It's something different.

It takes a little bit
of getting used to,

-but I really do
like it. Yeah.
-Yeah. Cool.

(whirring)

♪ ♪

(whirring)

♪ ♪

Andavolu:
"Cobotics" is a new word,

but the idea
has been around forever:

Let's integrate
new tools into old tasks
and do them faster.

(beeping, whirring)

At Amazon's robotic-enabled
fulfillment centers,

thousands of newly
developed AI-powered cobots

are rebuilding how man
and machine work together.

We actually started introducing

robotics in around 2012.

-So just, like,
seven years ago?
-Yep.

Brady:
Since then, we have created
almost 300,000 jobs.

-Just in
fulfillment centers like this?
-In fulfillment centers

across the Amazon workforce.

-We have this first example
of human-machine collaboration.
-Uh-huh.

These are mobile shelves
that are drive units

of the little orange robots
that you see below?
-Andavolu: Yeah.

Brady:
You can move those at will,
any shelf, at any time.

And at the right time,
like magic,

that universal station,
it's gonna say,

-"Hey, I think
that object is right here."
-Andavolu: Wow.

Brady:
Now, she is gonna do
a pick operations,

that's scanned,
and if it passes that bar,
it's the right object.

-Look at them go!
These shelves are crazy!
-Brady: It's awesome.

Andavolu:
But it's interesting.
So, who is sort of...

Who's working
for who here? 'Cause, like,
the robots are coming to her.

She's taking the stuff out
and putting it in,

but then she has to say
to the robot, "Oh yeah,
it's actually it."

-Is it on her time, or is
it on the robot's time?
-(laughing): I love it.

It's... I love that question,
it's great.

-It is a symphony of humans
and machines working together.
-Uh-huh.

You put them both together in
order to create a better system.

Is there a day where
there's gonna be a robot

who can pick
and look through things
just as good as she can?

And is that a day
that you're planning for?

Are you already
planning for it?

Brady:
Humans are amazing
at problem-solving.

Humans are amazing
at generalization.

Humans have high value,
high judgment, right?

Why would we ever want
to separate that away
from our machines?

We actually want to make
that more cohesive.

♪ ♪

-What's cool is that Amazon
is growing big time, right?
-Yeah.

And it's creating
a lot of jobs.

It's one of the biggest
job creators in the world.

So, with automation working
hand in hand with people,

is it making
jobs better?

I think it is making it better.
First of all, our associates,

-they choose to come and work
in our fulfillment centers.
-They choose to be here, yeah.

And we're really proud
of the wage benefit

that we are offering
our associates.

-It's a $15 minimum...
-Uh-huh.

...that we instituted this year.
We're really proud of that.

They are the reason
that we're so successful

inside our fulfillment centers.

Andavolu:
This is Jav.

He's 23 and at the very
beginning of his career.

He dropped out of college
for financial reasons,

then left a job
at an elementary school

to become
an Amazon associate
because it paid better.

He still works there,
which is why he asked us
not to use his last name.

When I heard we was
working with robots,

-I thought
the idea was cool.
-Uh-huh.

It's something
I only imagined coming
out of a sci-fi movie.

But, um,
I guess for the most part,

I dislike
the stowing process.

Stowing,
so what's that exactly?

Just imagine just standing
there, just... all day (laughs),

-for those 10 hours.
-Yeah.

You know, you feel like
you have to move fast.

You have to do right
by the robots, you know?

Do the robots work for you,
or do you work for the robots?

Wow, that's
a good question.

I feel like I work
for the robots. (laughs)

At Amazon, data seems
to be like this huge thing.

Like, do they track
your data as a human?

They track everyone.
(chuckles)

How many products are you
actually moving in a single day?

I think the highest
I've ever stowed was...

-2,300 units. Yeah.
-Wow.

You feel like the robots
you're working with. (laughs)

♪ ♪

Oh!

Andavolu:
Something you wanna
keep doing for a while?

You wanna stick
around with it?

-What, Amazon? No.
-Yeah.

(laughter)

-That was quick, right?
-(laughing)

(panting)

What agency
do you have

when you step
into that building?

What am I gonna eat
for lunch? (laughs)

But it's funny
'cause it's like,

I'm hearing from
people at Amazon that

human creativity
and problem-solving

is still something
that they value.

-You heard that
from someone?
-I did.

I don't know.

I haven't been put in a position
where I can like, you know,

be creative.

I'm pretty sure
a lot of people haven't.

♪ ♪

I know that one day,
I would like to get a career

that I don't feel like
I need a vacation from.

It's the whole thing
of being useful,
you know, like,

that's part of being a human.

We all have to feel useful
for something, you know?

-Andavolu:
Do you feel replaceable?
-Jav: Yeah, of course.

I know that if I get fired,
there'll be another person
in my place...

-ASAP, so...
-(indistinct bus announcement)

Andavolu: Do you think
they're gonna try to automate
you out of your job?

Jav:
Yeah, 'cause...

they wouldn't have
to worry about

people being injured
on the job, you know?

So, if they could
replace us with robots,

I think it could be done.

(whirring)

Andavolu:
From 2015 to 2017,

Amazon held competitions where
college teams designed robots

to do more or less
what Jav does all day:

single out objects, grab them,

then stow them
in a specific place.

See how I can collapse the hand,

so I can get it
into where it is?

A robotic hand,
just I haven't seen anything

with that sort of dexterity.

Andavolu:
That's Tye Brady,

the same Amazon
exec we met earlier.

Sure, he hasn't seen a robot
perform that task yet,

but that's exactly why
it's the holy grail

for making robots as
physically versatile as humans.

♪ ♪

-Can I... Will it hurt my hand?
-Lillian Chin: No.

-(whirs)
-Andavolu (stammers): Hi.
Yeah, that's not bad.

Yeah, that's actually
kind of gentle.

So, we've spent, I don't know,
however many hundreds of PhDs

and decades trying to make
robots smarter at grasping.

We're starting
to get there with

-artificial intelligence
and neural networks...
-Uh-huh.

...but even still,
it's still very early

-in terms of our
ability to grasp.
-Right.

Right? And the Amazon
picking challenge is
a perfect example.

A bunch of minds working
on it for a long time,

and we still haven't
figured out

how to just pick
things from a bin.

And that's
a multi-billion dollar,

potentially trillion dollar
value proposition.

What's so
hard about it?

Computers are
smart, right?

What we think is hard
is very different

from what computers
think is hard.

So, we think that being
a chess grandmaster is
a hard challenge,

but the computer can just go
through all the possibilities.

Whereas this, there's
infinite possibilities
to grab that apple.

What happens
when we crack
the grasping problem?

-A lot fewer
Amazon employees.
-(laughing)

(whirring)

Andavolu:
This March, MIT
and Harvard debuted

a new concept of the grabber,

saying Amazon's just the kind
of company that could use it.

Amazon already
sucks up
nearly 50%

of America's
e-commerce
transactions,

and makes a dollar for every
20 spent by American shoppers.

There's a reason it's valued
at nearly a trillion dollars.

But Amazon and its
tech peers, like Apple,

Google, and Facebook,
employ far fewer people

than the richest
companies of
previous eras.

So, even if robots aren't
helping you find the right size

at a brick-and-mortar
Gap quite yet,

that doesn't mean
they aren't eroding
the future of retail work.

-Goolsbee: I spent
most of my childhood
in Southern California.
-Mm-hmm.

The Whitwood Mall
and La Puente Mall were the...

the center of social life.

Andavolu:
Austin Goolsbee is
a professor of economics

and was the chair of economic
advisors under President Obama.

-Retail was, kind of,
often an entry-level job...
-Mm-hmm.

...with probably 16 million,
15 million people

-in the United States
working retail.
-Yeah.

And this technology,

-if you wanna think
of it as that...
-Yeah, it's interesting.

...replaced a different
kind of retail.

Mm-hmm. So, could you see
a mall like this

an example of kind of
creative destruction?

I mean, you can see
the destruction.

How could you make
a living doing that?

You know,
the Hat World.

Hat World.
Well, there's two.

There's Hat World
and Hat Station!

-You know, they're competing
against each other.
-Yeah.

And now,

-it almost seems quaint.
-Mm-hmm.

From the '80s and the '90s,
and into the 2000s,

if you had, say,
a college degree,

the technology
has been great.

-Mm-hmm.
-And it's allowed you
to increase your income a lot.

-That's right, yeah.
-If you're the financial guy,

the number of deals you can do
has expanded exponentially

as the computing power
has gone up.

Those same technologies
have come at the expense

-of expensive physical labor.
-Mm-hmm.

That's the first thing
they tried to replace.

♪ ♪

Andavolu:
One virtue of technology
is that it's impersonal.

It's an equal
opportunity disruptor.

So, even as automation
and AI hit lower-wage jobs,

they're coming for
higher-wage jobs too.

And the people
at this MIT conference
are pretty excited about it.

(applause)

Eric Schmidt:
The biggest, I think, focus

for a while is gonna be
AI acceleration.

Basically, can we use
machine learning and AI

in fields that have
not had it so far?

What do we have to do?
Let's fund them,

let's hire the people,
and so forth,

to bring those tools
and techniques to science.

McAfee:
Most of us walk around

with this implicit rule
of thumb in our heads

about how we should divide up
all the work that needs
to get done

between human beings
and machines.

It says, look, the machines
are better than us
at arithmetic.

They're better
at transaction processing.

They're better
at record keeping.

They're better at all this
low-level detail stuff
than we are.

Awesome. Give all that work
to the machines.

Let the human beings
do the judgment jobs,

the communication jobs,
the pattern-matching jobs.

When I think about
the progress that we're seeing

with AI and machine learning
right now,

that progress is
calling into question

that rule of thumb
in a really profound way.

-Right.
-Because
what we're seeing

over and over is
that the computers

are better at pattern
matching than we are,

even the expert human beings,

and actually they've
got better judgment.

♪ ♪

Andavolu:
Judgment calls
are basically

all we do
at the office
every day.

We take the facts at hand,
run them through
past experiences,

give them a gut check,
and then execute.

And these days,
we're offloading judgment calls
to computers all the time.

Whether it's to take the subway,

take the streets,
or take the highway,

or what to binge-watch next.

And what's behind
these decision-making tools

is a technology called
machine learning.

Some machine learning
relies on programmers

presetting the rules
of the game, like chess.

Newswoman:
As world champion
Garry Kasparov

walked away from the match,

never looking back at the
computer that just beat him.

Others utilize what's
called neural networks

to figure out the rules
for themselves, like a baby.

This is called deep learning.

However it's done, programmers
use other recent innovations,

like natural
language processing,

image recognition
and speech recognition

to take the messy world
as we know it, and shove it
into the machine.

And the machine
can process more data,
more dimensions of data,

more outcomes from the past,
more everything,

and could even go from helping
making judgments in real time

to making predictions
about the future.

With vast amounts of data
available in the legal,
medical, and financial world,

and the tools to shove
them all into the computer,
the machines are coming.

♪ ♪

Gianna visited a tech
company called LawGeex

to see whether their new
machine-learning software

could put lawyers
on the chopping block.

We've built an AI engine

that was trained
after reviewing

many, many, many
different contracts.

-How many? Wow.
-Tens of thousands,
even more.

We decided to focus on

automating the review
and approval of contracts.

So, simplifying and making
that process faster.

The actual analysis
that happens on the back end

takes a couple of seconds.

The work is actually
going through the report

that the system generates,

and then fixing whatever
issue that is found.

So, the system has flagged
all of these items for you.

-Bechor: Exactly.
-Toboni: Okay.

Some of them are marked red,

meaning that they
don't match my policy,

and some of them
are marked green,

meaning that they
do match my policy.

It gives me all of
the guidelines about what

the problem means, and also what
I need to do in order to fix it,

but then I fix the problem.

So, didn't spell-check
start doing this in the '90s?

-Okay, that's...
-(both laugh)

-Like, how is this different?
-Yeah.

The way LawGeex works
is very, very different.

It actually looks at the text,

understands the meaning
behind the text.

-Oh, wow.
-Very similar

to how a human lawyer
would review it, right?

The only difference is
that with the AI system,

it never forgets,

it doesn't get tired,

and it doesn't need
to drink coffee.

-Hey. Gianna.
-Tunji.

-Nice to meet you.
-How's it going?

-You ready for this?
-Yeah, I think so.

-Let's do this.
-All right. I feel
like John Henry.

(laughs)

So, Tunji, you're going up
against this AI system

with Noory to spot legal
issues in two NDAs.

We're rating you guys
on both speed and accuracy.

And because this isn't
a commercial for LawGeex,

I need you to try your hardest

-to beat this computer.
-I'm on it.

All righty. On your marks,

get set,

go!

♪ ♪

(clicking)

(quietly):
So, we snuck a little phrase
into this contract.

It says, "In case of any
breach by the recipient

"of any obligations
under this agreement,

"the recipient will
pay Vice News a penalty

of $15,000 per event."

So, we're gonna see if Tunji

and the AI system can find it.

♪ ♪

♪ ♪

Poor Tunji's just still
working away over here.

So far, it's taken
him more than

double the time that
it took the computer.

And meanwhile,
Noory and I just got a coffee.

He's having
a meeting right now.

Pretty clear just how much
time this technology saves.

Done.

Toboni:
Okay, guys.

-The results are in.
-Tunji: Okay.

LawGeex:
95% on the first NDA.

Tunji, 85%.

Second NDA,

95% for the computer,

83% for Tunji.

You don't seem disappointed.

I wasn't disappointed
when the iPhone came out

and I could do more things

with this new piece
of technology.

So, this is exciting to me.

So, Noory,
did the computer catch

the phrase we put
in there about

Vice News?

Yeah, the computer
caught it,

and I was able
to strike it out.

Tunji, did you catch it?

I straight up missed it.
I didn't see it at all.

We take cash, check,

whatever you have.
(laughs) Just kidding.

So, McKinsey says that
22% of a lawyer's job,

35% of
a paralegal's job,

can now, today,
be automated.

So then, what happens?
These jobs don't go away?

People just take
longer lunch breaks

or take on more
clients or what?

We're kidding ourselves
if we think that

things are not
going to change.

But similar to pilots
with the autopilot,

it's not like we don't
need pilots anymore.

Could this technology
pass a bar?

No.

Okay, so we still need lawyers

to be signing off
on these legal documents,

even if they're not doing
the nitty-gritty of them.

Absolutely. Defining the policy,

serving escalation points,
handling negotiation,

and then also handling more
complex contract categories.

-Mm-hmm.
-You have a new
generation of lawyers

that are much
more tech-savvy.

The ones that can actually
leverage technology

are the ones that
manage to prosper.

Andavolu: Unlike the law,
medicine has always been
intertwined with technology.

But it's relationship with
robotics isn't just graceful,

it's miraculous.

Toboni:
What is the most
groundbreaking thing

about how far we've come
with robotic surgery?

So, robotic surgery
allows us to really

treat tissue
in a more delicate way...

(whirring)

...allow us to be much more
efficient in suturing,

decreasing bleeding.

We are improving outcomes,

shortening hospital stay,

and that's why we're using
more and more now.

So, the robot is
enabling surgeons,

and by enabling
surgeons is giving

access to more patients
to minimal invasive surgery.

Horgan:
Okay, we're good to go.

♪ ♪

Toboni:
That's great, right?

'Cause one of the huge problems

in our health care system is
that not enough patients are

getting seen when
they need to be seen.

Not only that, not every
patient get the same care.

You may have great
surgeons in one area,

but not in other areas
with the same experience.

The robot is flattening that.

-We're ready?
-Man: Ready.

♪ ♪

Horgan:
I think that the biggest
groundbreaking part

is the fact that I am
operating on a console,

and the system is
on the patient.

We have all the degrees
of articulation

you would have in your hand

inside the abdomen.

That's revolutionary.

Man:
Geez. That's close.

Horgan:
Okay!

I'm gonna go out
and have my espresso.

Toboni:
The way robotic
surgery is evolving,
do you see any jobs,

like the job of a technician
or a physician's assistant
going away?

No, but what we have seen is
the opposite--

is them being
much more involved.

Okay, stapler.

Assistant:
Loading up the seam guard.

Horgan:
You need a nurse that knows
how to move it around.

You need a scrub tech
that knows how to load
the instruments,

how to clean the camera,
and how to move the arm,

-how to undock.
-Toboni: Right.

Horgan:
I like it a lot there.
What do you think?

-Assistant: Yeah, looks good.
-Horgan: Only good?

-Assistant: Perfect.
Unbelievable.
-Horgan: Eh?

Horgan:
Look at that.

Toboni:
How do you think
automation will

evolve in this area?

The anatomy changes
from patient to patient,

but with machine learning,

the system will be able
to recognize different tissues.

-Mm.
-And I'm sure
that in the next year,
we'll see the system

-telling you that's cancer,
that's not cancer.
-Wow.

And that's where
the benefits are gonna be.

♪ ♪

Andavolu:
Finance has been cashing
in on the benefits

of machine learning for years,

hiring so many programmers

that it's virtually
indistinguishable from tech.

Michael Moynihan
visited Goldman Sachs

to see how they're leveraging
AI's predictive capabilities

to make more money.

How has your job changed,

and how has this
industry changed?

Over time, there's been
a lot of automation.

I think it lends itself
naturally to trading, right?

If I'm trading Google,

if I have to make
a decision whether

I'm going to buy it
or sell it at a certain price,

there are hundreds of variables
that go into that decision.

And you can code an algorithm

to assess all those variables,

and when you swing
that bat 1,000 times,

it's going to do it more
efficiently than a human would.

-(bell ringing)
-So, when you look on the floor
of the New York Stock Exchange,

there aren't a lot of
traders down there anymore.

Moynihan:
If I went down there today,
what would I see?

Levine:
You wouldn't see
a lot of people.

You might see them clustered.

Moynihan:
Little tiny clusters
of people?

Levine:
It'll be a cluster.

Moynihan:
So, to be in this industry now,

do you need to understand

lines of code and what they
do and how to produce it?

Sinead Strain:
I'd say more and more, yes.

If we look at this firm,

the number of engineers,
technologists that we have here,

we're probably the largest
division within the firm.

♪ ♪

So, you have
a math background?

Yeah, math
and computer science.

-But yours is economics
and computer science?
-Yes.

Moynihan:
But you learned this on the fly?

Jameson Schriber:
Yeah, in school or here.

More likely, most of
the things were picked up here.

Can I see like,
you know, an algorithm,

give me some
sense of code?

Phillips:
I'll show you a really
simple example.

So, in this case, I'm gonna
run a volatility function,

essentially an algorithm,
and this is showing me now,

on a 22-day rolling day basis,

what's the volatility
level of the S&P 500.

Essentially, we would
write code to do that.

So, in this library,

this is the volatility function.
It's actually fairly simple.

-But, we--
-Wait, I'm sorry.

That's fairly
simple?

This is all documentation.

This is actually a fairly
simple algorithm.

It's essentially the code
that's generating
what you're seeing.

Looks terribly
complicated to me.

People like
you guys still

need to exist to create
these things, right?

I mean, are they
self-sustaining?

Or can you write
yourself out of a job?

-I think we're... Yeah.
-I think that would be hard.

♪ ♪

Moynihan:
Formerly a tech CEO,

Marty Chavez is
now global co-head

of Goldman's
securities division.

It strikes me, obviously,
that this is an industry

that has been on
the forefront of using

AI, computers, to...

you know,
make big decisions
and make a lot of money.

What has that done
to the kind of, you know,

the job market within--
even within this company.

It's creating new jobs that
couldn't have existed before.

Whole new businesses

now exist for us
and out in the world

that wouldn't
have been possible

without the technologies

that have arisen
over the past few years,

whether they're machine
learning, cloud services,

open-source, right?

All of those
activities go into,

for instance,
our new consumer

lending and deposit
taking activity.

You know the counterargument
to this, don't you?

These are jobs that are being
created for smart people,

-educated people,
rich people.
-Mm-hmm.

Other people out there,
who are being made
redundant by robots,

don't have the skills.

-It's only for
a rarefied few.
-Mm.

How do you respond to that?

Technological change is
disruptive,

and there's
a lot of pain,

and it's something
that we must concern
ourselves with.

What do you do during
the disruption,

which has been continuous
since the agricultural
and industrial Revolutions,

and I expect it will continue

and will accelerate
in all likelihood.

And so, sitting back
and complaining about it,

sitting back and doing
nothing about it,

don't seem to be options.

At the same time,
I don't think the answer is

-to stop the progression
of technology.
-Mm.

Haven't seen that work.

(brakes screeching)

(PA announcement):
This is a Queens-bound
M local train.

The next stop is Myrtle Avenue.

It's worth noting at this
point that even if AI tools

are better at making decisions,
high-level judgment jobs

aren't about to be automated
away any time soon.

And let's be clear,
the people who are

most excited about
AI's creative potential

aren't the ones
getting replaced.

So the question becomes,
as this disruption accelerates,

how do you benefit from it
if you're not already rich,

or white, or male,
or have a spot at the top?

Let's put it this way.

It's a lot easier for all
of us to get directions,

but it's becoming
increasingly hard

for most people to find
their way to a stable career.

And there are not a lot
of people whining about it.

There are a lot of people
who are racing to catch up.

♪ ♪

We visited a class
at Per Scholas,

a non-profit that
skills up people

in New York and other cities

around the country for free.

(indistinct chatter)

(chatter continuing)

Student:
I have had just about every
terrible job you can imagine.

Fast food, grocery store,
stock clerk.

I was a supervisor
at a retail store
while I was in college.

I was a mechanic.
I worked in the industry
for over 10 years.

Then I went
to substitute teaching.

Then I went into solar,
doing sales.

And then I was a writer,
and then I became
an English teacher.

Reservation sales agent.

Customer service,
marketing.

Now I am here
at Per Scholas.

-I should do that
in CodePen, right?
-Yeah.

The job market is really
shifting towards a gig economy.

What do you have
that you can work for yourself,

or work for someone else?

A year ago,
I would've told you
that I was gonna go

to grad school and get a PhD
and all that good stuff.

But the reality is
when I graduated,

and I was looking for jobs,

one of the main things
that kept popping up

was software engineer,
coding, tech.

Do you think
you would've made the move
if this place wasn't free?

No. (laughs)

Because just
graduated college,

so, you know, Sally Mae
is still knocking on my door.

-They're knocking
on my door right now.
-(knocking on door)

-Is that them?
-You see?

-(laughter)
-I just said their name,
and here they are.

When I put
the comma, right?

And I put the P, does
that mean parent and P?

-I'm a mom of three.
-Mm-hmm.

So, my last job, I was working
at a busy call center,

and I was in a place where
I felt you know, undervalued.

Robots was gonna come
and take my job,

and I would've
been without a job,

and what would I've been able
to pass onto my children?

"Hi, how are you doing?"

With this,
I can give them a skill.

They can be in a better place
in the next 10, 20 years,

opposed to how long it took
for me to figure this out.

Do you think you're at
a disadvantage as far as
the job market itself?

You look at the tech industry,
and it's like,

other than south Asians,
east Asians,

it's like, there are not
a lot of people of color.

Yeah, I wanted
to touch on that.

This is actually why
I decided to specifically

go into coding because
this is one where,

at the end of
the day, it's just

how good your
applications are,
your codes are.

Growing up, I used to think
this was just magic,

or like it's done
by all the smart kids
in California, you know?

So, I wanna prove
that it can be done

by some kid in Queens
who just wanted to do it.

(indistinct chatter)

Kelly Richardson:
The perfect world is

that we have enough investment

that we could grow to meet

-both the size of the demand
and the size of the supply.
-Mm-hmm.

Richardson:
We have more employer
partners willing to hire
than we have graduates for.

We have more students
or applicants

-applying to Per Scholas
than we have spots for.
-Right.

The constraint to our
growth is resources.

The domestic investment
in workforce retraining

-is so small, and...
-Mm.

...the impact automation
is going to have

-is not going to be equitable.
-Mm-hmm.

That it's largely
people of color,

largely women who are
in the current low-wage

occupations that are
gonna be displaced.

There really should be
some critical thinking

and some action that
legislators are taking

to invest in
programs like this.

For decades, the federal
government has repeatedly

taken action
to fund reskilling.

I am proud today
to sign into law

the Job Training
Partnership Act,

a program that looks to
the future instead of the past.

Giving all Americans
the tools they need
to learn for a lifetime

is critical to our ability
to continue to grow.

So, the bill I'm about to sign
will give communities

more certainty to invest
in job training programs
for the long run.

Andavolu:
But in today's dollars,

that funding
has fallen
for years.

President Trump campaigned

on bringing jobs back
to American workers.

A Trump administration
will stop the jobs
from leaving America.

Andavolu:
And he signed an executive
order enacting, what he calls,

the White House's Pledge
to America's Workers,

installing his
advisor and daughter,

Ivanka Trump,
to lead the charge.

It's one of my favorite
words, reskilling. Reskilling.

We're calling upon government
and the private sector

to equip our
students and workers

with the skills they need
to thrive in the modern economy.

♪ ♪

(indistinct PA announcement)

(horn beeps)

Andavolu:
The Trump
administration's strategy

on reskilling America's
workforce is a lot

like its strategy for other
big problems America is facing.

Rather than increasing
public investment,

they prefer to see private
industry fill the void.

For the past nine months,
the Trump administration

has been twisting the arms
of CEOs to promise funding

for worker education
and training.

And so far, more than 200
companies have signed on,

with Toyota being the latest.

Ivanka:
Toyota signed our pledge

to America's workers for 100,000

enhanced career opportunities,
apprenticeship, um...

-new jobs...
-New development.

New development,
workforce training.

I think James was
a little bit inspired.

He's just increased that
commitment to 200,000,

-which is unbelievably exciting.
-(cheering)

Bevin:
The American worker is great,

but the Kentucky worker

-is greater still.
-(cheering)

(applause)

Does government have
a role? Absolutely.

But is it to define what
the workforce looks like? No.

If the state were
to develop a program

for Toyota's workers
of the future,

it would be a failure.
I'm telling you, straight up.

Toyota knows what Toyota needs.

But the risk, perhaps,
is a reliance on a company

that isn't owned by you and me,
like the government is.

It's owned by shareholders.
Is there a problem with having

a private response
to a public problem?

Who has more of a vested
interest in getting this right?

The government
or the private company?
The private company does.

It is in the best interest of
Toyota or any other company

to train the best
quality people,

pay them as much as possible,

give them the standard of
living and the quality of life

that makes them wanna come

and retire 20,
30, 40 years later

from the very same company.

It's in the company's
interest until it isn't.

And the thing about a company
is, like, we don't elect

their executives,
but we elect our
government officials.

But the idea that we're
gonna rely on government

and "elected people,"
to come up with rules

for training people
for jobs that you will

volitionally want to buy
the products of is crazy.

♪ ♪

Andavolu:
Under the pledge,

they're promising 200,000
new opportunities.

They're promising to reskill

and retrain 200,000 people.

They have the capacity
to do that.

What happens
if they break the promise?

Again, they're
gonna do their--

Can you mandate
the government does it?

If they don't do it,
it's not like we're gonna
not reelect their CEO.

They don't do it, they might
have some bad publicity--

You think it won't
affect their CEO?

Here's what happens.
If they don't do it,

if they're not making
these investments,

there's not a chance
that they survive.

(applause)

♪ ♪

(indistinct chatter)

Andavolu:
This is Landon and Tim,

two buddies who work together
at the Toyota factory.

Tim's just retired,
but Landon sees himself
working at Toyota for decades.

It must be kind of nice
to at least know that

at a high level,
they're thinking about

your jobs, the things
that you guys do,

the opportunities
that you guys have.

Does it strike you that way?

Tim Smith:
With Ivanka,

I just looked
and said okay, it's PR.

Do people want to be reskilled?

If their job depended on it,

and they know it's coming,

I would say sure. Yes.

It would be depending on
your personal circumstances,

that could be very hard

because if you work,
you know, a full shift

and you have a family,

you may only have an hour
or two of free time a day,

are you supposed to go drive
to a training facility

and spend a few hours
a day there before
you go do your shift,

-you work your shift?
-Mm-hmm.

I don't think very many
people would do that.

Do you guys like your job,
or do you like working there?

(chuckles)

-You go, Timmy.
-(laughing)

-I actually love my job.
-Yeah.

Um, you know,
because I've always

loved working on cars,

so my job was pretty
much up my alley.

-There are parts
of my job that I like.
-Mm-hmm.

It's, you know, you're
creating something.

There's no way it could exist

without someone
putting it together.

(laughing)

-Exactly.
-And so...

-Yeah, I have a wife
and two daughters.
-Right.

-I wanna spend as much
time with them as I can.
-Mm-hmm.

They're young and they're not
gonna stay young for long,

and I hate missing the time
I miss with them at work
already.

I just wanna spend as much time
as I can with them.

I wanna retire.
That's the... that's the goal.

Andavolu:
For people in
the workforce today,

reskilling boils down
to doing more work
just to keep up.

To Andrew Yang, a former
job creation specialist,

and now a long-shot
presidential candidate,

the spotlight on reskilling
hides a larger imbalance

between the goals of workers

and the goals of
their employers.

We are so brainwashed
by the market that otherwise

intelligent, well-meaning
people will legitimately say,

"We should retrain
the coal miners to be coders."

Yep.

Yang:
We are trained to think

that we have no value unless

the market says that
there's a need for what we do.

And so, if coal miners
now have zero value,
then the thought process is,

"Oh, we have to turn them
into something that does have
value. What has value? Coders!"

And then, 12 years from now,

-AI's gonna be able to do
basic coding anyway.
-Sure.

So, this is a race
we will not win.

The goal posts are gonna
move the whole time on us.

What the solutions?
What are you proposing?

We start issuing a dividend

to all American adults,
starting at age 18,

where everyone gets
$1,000 a month.

So, basically
a universal basic income.

-Yes. We've rebranded it
the Freedom Dividend...
-Okay.

...because it tests much
better with conservatives
with the word freedom in it.

(laughing)

And it's not a basic income,
it's a dividend.

Which is to say the economy
is at a surplus,

so everyone deserves
a piece of the pie.

Yeah, no. All of us are owners
and shareholders of the richest

society in the history of
the world that can easily afford

a dividend of
$1,000 per adult.

People need meaning, structure,
purpose, fulfillment.

And that is the generational
challenge that faces us.

It's not like the Freedom
Dividend, giving everyone

$1,000 a month, solves
that challenge. It does not.

-But what it does is it--
-Buys us time?

It buys us time and
also channels resources

into the pursuit of
meeting that challenge.

It ends up
supercharging our ability

to address what
we should be doing.

Andavolu:
Universal basic income, once
a marginal political fantasy,

has been embraced by more
than a few rabid capitalists
in recent years.

We should explore ideas
like universal basic income

to make sure that everyone has
a cushion to try new ideas.

It's free money for everybody.

Enough to pay
for your basic needs:
food, shelter, education.

I don't think we're
gonna have a choice.

It will come about one day.

-And I think--
-Out of necessity?

Out of necessity, and...

and I think that cities
should experiment.

Andavolu:
Michael Moynihan went
to one city that already is.

-Moynihan: Mayor.
-Good morning.

Michael Moynihan.
How are you?

Moynihan:
Stockton's mayor won election
at just 26 years old,

taking office five years after
the city declared bankruptcy.

So, you grew
up here?

-Born and raised.
-Born and raised?

-This is home.
-And you left

for a brief period
to go to Stanford?

Left for four years,
I came right back.

Moynihan:
With funding
from Silicon Valley,

he's launched a pilot program
that's giving 125 residents

$500 a month for 18 months.

Why do Silicon Valley
guys like this so much?

Can't speak for all of them,
but I think a lot of them see

how detrimental it
would be to society

if there's a mass number
of people who are automated

without any way to make
a means for themselves,

without any way to provide
for themselves.

I mean, is it people in tech
kind of paying indulgences,

and saying, "Hey, we're
kind of screwing this up.

We feel bad about it.
Here's some money"?

I know some people are
talking about robot taxes

and things of that
sort, and I...

I'm in very much
agreement with that.

That they have
a responsibility

to society.

It's voluntary now, but it might
be at the point of a gun later.

Yeah, voluntary
to start and pilot,

but absolutely
I think to scale it,

it's not gonna be
about generosity.

It's gonna be
a matter of policy.

♪ ♪

Moynihan:
The criticism that
most people get

for UBI type
experiments is that,

all right, you're just giving
people money, a handout, etc.

That's a start,

but what's the long-term
goal for jobs in Stockton?

The hypothesis
we're working on is

that folks have their
basic needs met,

and then create
the workforce of the future.

Primarily starting with
the kids in our schools now,
but also with the adults,

to give them opportunities
for retraining, reskilling,

but also supporting people in
their entrepreneurial pursuits.

Folks with a lot of potential
but historically folks

who haven't been seen as
important enough for investment

or important enough
for government
to really partner with,

so that's what
makes people excited

about the work we're
doing in Stockton.

So yeah, this is
the downtown marina.
Watch your step.

Moynihan:
There was basically
nothing here

when you were a kid,
when you were five,
six years old.

Tubbs:
I never really came out here

till I came back
for city council.

-Yeah.
-This wasn't part of my
Stockton, but I think for me,

this spot's so important 'cause
it represents real potential.

-There's not many cities
that have... this.
-Yeah.

When we talk about the future
of work, we're talking about

how do we ensure that
those left behind today

aren't further left
behind tomorrow,

and that's my biggest fear.

The folks who are making
the least now

are the most likely
to be automated out of jobs
and making anything.

So for me, a basic income isn't
even about the future of work.

It's about getting
our foundation set
in the present,

so that when
the future work happens,

we have a firm foundation
on which we can pivot

and figure out what we can do
with and for people.

Andavolu:
What's attractive about
UBI is its simplicity.

But that's also what makes
it vulnerable to critique.

Daron Acemoglu:
I don't think these
Utopian ideas

of universal basic income,

robots do all the production
and then everybody

stays at home
with a decent income
and plays video games.

I think that's
a very dystopic future,

and it won't work,
and I think it will lead
to a huge amount of discontent.

We really have no option
but create jobs for the future.

Politicians have to start
engaging these issues

to figure out
what's politically feasible.

It's gonna trigger
fundamental debates

about things like
a universal basic income

that everybody ought to get
money and the question is,

okay, where's that
money gonna come from?

Andavolu:
According to a progressive
think tank's analysis,

a UBI program giving every
American $10,000 a year

would cost the government
more than $3 trillion annually.

The entire 2018 federal budget
was just over $4 trillion.

Universal basic income isn't
the only policy being floated

to shore up the workforce's
shaky financial foundation,

whether it's using federal
funds to guarantee jobs
for anyone who wants one,

giving tax breaks to companies
to create more jobs,

or strong-arming companies
to keep their factories open.

What all these policies
and strategies tell us

is that the broad consensus
is that right now...

Protesters:
We say fight back!

Andavolu:
...things aren't working.

♪ ♪

Brynjolfsson:
We've designed a system where,

as the technology
creates more wealth,

some people, bizarrely,
are made worse off.

We have to update
and reinvent our system

so that this explosion
of wealth and productivity

benefits not just
the few, but the many.

Haass:
This challenge
of new technology

is not taking place
in a vacuum.

This country's
already divided.

It's divided geographically.
It's divided culturally,
politically.

We've gotta be prepared
for the fact that it could

actually take
the social differences

we already have,
and make it worse.

There's gotta be
a sense of urgency here.

A higher level of disconnection,

alienation, more declines
in social capital,

more groups of
people left behind,

more geographic
areas left behind.

-Protester: What do we want?
-All: Justice!

-Protester: When do we want it?
-All: Now!

McAfee:
This is not a recipe
for a stable, prosperous,
happy society.

My worry is not that the robots
will take all the jobs.

My worry is that more
people will be left behind

and will feel left behind
by what's going on.

Stiglitz:
If we continue in the way
we've run our economy

for the last 40 years,
it will be disastrous.

So, when you talk about
the future of work,

we're kind of talking about
the future of the whole system?

That's right. I mean,
you cannot have

a prosperous economy
without prosperous workers.

Voltaire said
that work saves us

from three great evils.

Boredom, vice, and need.

Haass:
So much of our identities
are tied up with our jobs.

People ask you how you doing,
who you are, what you do,

and that's essential
to our sense of self.

If we have millions,
or tens of millions,

of chronically,
long-term unemployed,

what's gonna become
of those people?

It's not simply
a question of how are they

going to support themselves.
What are they gonna do?

(horn honks)

(engine rumbling)

(drill whirring)

This is your NOx sensor drift.

-One box.
-Conversion efficiency.

Yep. One box gone.

(rattling)

You know,
someone might say

robots are coming. Might as
well hang up the keys now.

Does that go
through your head?

It's gonna happen.

-It's gonna happen, I mean--
-What's gonna happen?

With these vehicles
driving by themselves.

Change is good. Some change
ain't good, you know?

I mean, that's gonna be a lot
of people out of work, and...

Andavolu:
What do you think
you're gonna do?

♪ ♪

There's gonna be a crash.

I think there's gonna
be a lot of outrage.

-Riots more or less,
you know what I mean?
-Sure.

'Cause they're gonna fight
to try to keep their job.

I mean, would I do it?
Yeah, I would do it.

It's gonna be
a chain reaction.

In reality, you gotta
look at the economy

and what is it gonna
do to the economy.

What is it gonna do to
the American people?

Not just the industry,
I mean to... to the world.

What's gonna happen you got
the whole world pissed off?

(hammering, clatter)

But me,
I wanna die in a truck.

-Really? (laughs)
-I'm gonna die
in a truck, brother.

I've told all my friends,
I've told my family.

-That's when I retire is
when I die in a truck.
-Yeah.

I mean, I been doing
it for too long.

It's in the blood.

♪ ♪

♪ ♪