You can’t care for patients; you’re not human! | Maneesh Juneja | TEDxPorto

When it comes to improving the health of the world, many eyes are looking at recent advancements in Artificial Intelligence (AI) with hope. Some even believe that AI will solve every problem in health care. Our homes will become smart enough to know that we haven’t been active enough and will instruct the fridge to suggest healthy recipes personalized to your nutritional goals. Your self-driving car will track your health during your commute to the office and schedule doctor’s appointments. You will grow old surrounded by robot companions who can take care of your every need, never feeling tired, never complaining but anticipating your needs in advance. Technology companies, large and small, are looking at health care and wondering, “How can my algorithm make this better, cheaper and faster?”

Humans are inconsistent and make mistakes and don’t always smile at you if they are having a bad day. Children born today, growing up with digital assistants, maybe the first generation who believes that intelligent machines can "care" just as well, if not better than any human doctor or nurse.

Health care is about humans, it involves listening, empathy and compassion. Machines can’t care for us; only humans can. However, what if some patients prefer interacting with a machine rather than a human? Should we be holding back progress because we don’t understand this?

This talk was given at a TEDx event using the TED conference format but independently organized by a local community. Learn more at https://www.ted.com/tedx

When it comes to health care today,

what's natural is that
it's a deeply human, personal experience.

Kindness, empathy, compassion,
eye contact, human touch -

when we're feeling sick,
when we're unwell,

when we need help and reassurance,

we turn to human doctors,
human healthcare professionals.

That's what's natural right now.

However, I want to tell you a story

about something that happened
to me and my family

which made me rethink
what's going to be natural in health care

and whether humans
would always be at the center

of the healthcare experience.

So last year, in England,

my mother called a service
on the telephone

provided by the healthcare system,

where you can phone up,
you describe your symptoms,

and they tell you whether to go
to the emergency room,

whether you go to urgent care

or whether it's something
you can take care of at home

using over-the-counter
medicines, for example.

So middle of the night, and she called up,

and they took her through
the list of questions,

and they're rushing her through the list.

She's in her seventies.

She's sick, she's confused,
it's the middle of the night.

And she was telling them,

"Look, I'm really old, I need time
to think about my answers."

And they're rushing her,

and she felt that the entire experience,

despite a human being

being a human being
on the end of the telephone

in the healthcare system,

it was missing of compassion,
empathy, kindness.

Now, contrast that with a few weeks later.

I was also unwell, and I decided
to use the same service.

But this time, there's a version
of that telephone hotline

but provided by an app -
without any human being.

And I put in my symptoms,

and the end result was it advised me
I can take care of myself at home;

nothing to worry about,

"You don't need to go to a facility."

But what I found was really interesting

was compared to my mother's experience

of the human being on the telephone
rushing her through the questions

is this machine, using this app,
I had time to think about my answers.

I had time to respond.

I wasn't rushed.

And it really felt like
a compassionate experience.

And it made me think:

Are we at a point

where one day we might have
machines that can care for us,

machines that can do
what humans currently do today?

When we think about the greater challenge,

we currently have 7.5 billion
people on this planet.

Many of them don't have access
to any healthcare facilities,

and even if they do have access,
it's often unaffordable

or there're long queues.

And if you think,
by the end of this century,

it's forecasted that we're going to go
from 7.5 billion to 11 billion people -

how are we going to provide
safe and affordable health care

to those 11 billion people
by the end of this century?

And on top of that, there's another trend,

that healthcare systems are trying
to reposition themselves, saying,

"Instead of just coming
to us when you're sick,

what can we do to help you
prevent you from getting a disease?

How can we help you stay healthy?"

So self-care

and taking more responsibility
for our own health on a daily basis.

And that's where AI is coming in:
Artificial Intelligence.

So you'll hear stories;
influential people saying,

"We got to be really scared about AI

because if it develops
and we let it develop,

it's going to end up ruling us

and we're going to become
slaves to these machines."

But I believe, especially in health care,

what if we can develop AI

that's compassionate,
that's kind, that's empathetic,

and it can fill this gap
in helping 11 billion people

in the future at the end of this century

have access to health care
whenever they want it, at the right time?

I want to share my experiences.

I've been trying a number of chatbots.

You may have tried chatbots
for customer service on a website.

And this one is called Wysa -

it's developed in India -

a compassionate AI chatbot.

And what I found when I was using it

is that it actually feels

like you're talking or you're chatting
to a real human being.

And the founder of this chatbot,
Wysa, is a lady called Jo Aggarwal -

based out in India -

and in her TEDx Talk last year,
she talked about what changed for her:

that after some users
had been using in this,

she got a message from one of the users,

a 13-year-old girl who said
that she had suicidal thoughts

and what helped her cling on to life,
at that point, was Wysa,

this app on her phone.

Think about it.

Imagine the people out there
around the world.

They don't feel like
they have friends or family

or even a psychiatrist or a psychologist
to turn to in that moment.

What if they could turn
to a machine that cares?

Is that a bad thing
just because it's not natural?

Think of how many lives could be saved
when it comes to mental health.

You've got the rise of smart speakers

from Google, from Amazon,
even from Apple now.

So children now growing up

are finding it perfectly
normal and natural

to have a conversation
with a smart speaker

and ask it the news, ask it the weather,
possibly even get information on health.

And then, if you think about
what the potential for this,

researchers are already working
on algorithms behind the scenes

that will take this voice data,
you're having a conversation,

and it could even potentially detect
conditions like Alzeimer's and Parkinson's

just from your voice data.

Think how powerful
that could be on a global basis,

just from these speakers as they evolve.

It's even coming to toothbrushes.

So this is me,
brushing my teeth in London

with the world's first toothbrush
which is powered by AI.

It has sensors in it, monitors
my brushing frequency, my habits,

how much pressure I'm putting.

But what surprised me

was that this toothbrush
actually cares about my sleep as well.

So it noticed that I was brushing my teeth
consistently late at night.

And it says, "Hey, maybe you want
to go to bed earlier

and have a good night's sleep."

How would you feel about this world

where even your toothbrush
is now caring for you?

Not just about your teeth
but about your sleep.

What next? It's going to tell me
what pizza I should eat tomorrow?

(Laughter)

So potentially in the future,
we might have machines that care for us

because they can recognize and respond
to our emotions in real time.

This is what researchers
out there are working on.

They have this vision of machines

that are as compassionate
and kind and empathetic

as any of us as humans.

So it might not feel artificial anymore

to deal with a machine
in these scenarios in health care.

So there's a company
out in the US, Affectiva.

They're working on emotion AI
for automotive, for example.

So picture this scenario:

you're stuck in rush-hour traffic,

here in Porto,

you're feeling a bit sad,
it's been a long day,

you're stressed out.

How would you feel if the car could detect

that you're feeling sad
from your expressions on your face

and how you're looking
and how you're feeling?

And then, maybe suggest,

"Oh, maybe I'll play you
some music to cheer you up,

or maybe I'll change
the fresh-air setting in the car."

Or maybe if it can detect
that you're feeling that tired, it says,

"Pull over at the next rest area."

So think about even these cars
that potentially will care for us,

these machines.

How do we feel about
more and more machines caring for us?

[What do people want?]

And it's important to understand
what people want.

It's not just about technology,
people saying,

"We can develop this. Let's roll this out.
It's great for humanity."

Let's ask people.

So there was a survey done that asked
people in a number of countries,

"How willing would you be -

would you be willing to engage with AI
and robotics for your healthcare needs?"

They asked a number of people
in a number of countries.

The country where people were most willing
to engage with health care and robotics,

with AI and robotics
for their healthcare needs

was Nigeria, in Africa.

94% of people surveyed in Nigeria said,
"Yes, I'd be willing to engage with AI."

So when you think about
where are the places in the world

that are going to try these new advances
in terms of caring machines,

where is it going to hit first?

Who's going to be most open to it?

It could be countries we don't associate
with innovation and technology.

And in fact, in that same survey

in the UK only 39% of the people

said they'd be willing to engage with AI
and robotics for healthcare needs.

And then, what about
when we might prefer -

even if you got a human being
available in health care -

when might you prefer a machine,
to deal with a machine?

So there's another survey,
this one done in 2016,

they asked smartphone users
around the world.

And they found that actually 29%
of those people surveyed

said for sensitive matters,

they'd prefer to deal with an AI
medical adviser than a human being.

Think about that.

What could drive you to prefer
a machine than a human being?

I was thinking, maybe it's mental health,
maybe it's sexual health.

How many of you have had
an embarrassing condition,

and it's either delayed you seeking help
or you haven't looked for help at all

because it's just too embarrassing?

You don't want another human being

to judge you because
you have that condition.

But there are tough decisions ahead;
it's not going to be easy.

Again, like I said,

it's not a case of just
this technology is here, let's use it.

We have to be very, very cautious

because, again, this is our health,
this is our well-being.

And there was a video for a robot
that's come out of London.

And in this video,
their promotional video,

it showed somebody reaching
for a midnight snack.

Many of us have had midnight snacks.

We're not supposed to -
it's bad for our health - but we do.

And in this video,
this robot that was in the kitchen

noticed that they're reaching
for their midnight snack

and prompted, as they opened the door,

"Hey, sure you want to have
that burger or that cake?

It's an extra two hours to burn it off
in the morning to go for a run."

And that person then decides
to follow the robot's advice

and not have that midnight snack.

How do you feel about a world
where you'll stay healthy,

you'll be prevented from having
so many diseases

because these machines
will limit our choices on a daily basis,

whether it's in the car or the office,
at home, wherever we are?

How do you feel about your freedom
being affected there

with machines that care?

And then, underpinning this revolution
is going to be day to day about us -

our preferences,
our behaviors, our habits.

And think about the recent scandal
with Facebook and Cambridge Analytica,

and then privacy and trust.

Who's going to be having access
to all this data?

Is it a health insurer?

Are people going
to be using it to penalize us

because we're making decisions
that are not healthy?

There's so much to think about
on this road ahead.

And then errors as well.

So the possibilities are that, actually,
if we turn to machines that care,

maybe they're not going
to make mistakes

like human healthcare professionals do;

they're not going
to give us the wrong drug,

they're not going to misdiagnose us.

But equally, these digital systems
that are going to care for us

are not perfect either.

So just this week, I was trying something
from Microsoft called Cortana.

It's a general-purpose
assistant on my phone.

I said, "I'm bleeding,"
and it says, "Okay."

I said, "I'm bleeding to death,"
and it says, "Interesting."

(Laughter)

So we have to be really
careful in this journey:

who we trust, what we trust.

And accountability is really important.

So today, if a doctor makes a mistake,

we can go and say,

"Hey, you made a mistake,
you gave me the wrong medicine."

Who are we going to hold accountable

when these algorithms,
these robots, these chatbots -

these machines that care for us -

get something wrong?

How do we even know
when they've made a mistake?

There's millions of lines of code.

Who's going to regulate them?

Another thing I was thinking about

is that potentially could it make
the world more inequal?

In terms of could it only be in the future

that the very rich people have access
to human healthcare professionals?

For the rest of us,
we all have access to health care,

but it's via a machine
or a type of machine.

Think of Japan as well.

Now, in some ways
they're ahead of the curve.

They've got an aging population,
a shrinking workforce,

and they're also leading the world
when it comes to research of care robots,

especially for the elderly.

And I met a lady from Japan
a few years ago; she said,

"I'd rather be cared for by a robot nurse,
than a foreign nurse." Why?

Because they don't want
to open themselves to immigration.

So what does that mean here in Europe?

You've got Brexit.

Part of the drive of Brexit
was immigration, a fear.

We want to reduce immigration.

So does that mean here in Europe
we're going to end up in a world

where we do have these machines
that care for us

because we, suddenly,
we're afraid of foreigners?

We won't have enough people to care.

Then what about loneliness,
an epidemic of loneliness?

So in the UK it was suggested

that we give these smart speakers
to the elderly to combat loneliness.

Is this the kind of world
we want to build,

when we have a connection
to a machine rather than each other?

Is that really the way forwards?

So we have to have balance.

We can't be too afraid;
we can't be too excited.

We have to take a middle path
when it comes to this future.

But if it done well, if we do it right,

if we really involve everybody
in the conversation,

I believe that, potentially,
machines that care

could increase the humanity
in health care around the world.

Think about those 11 billion people
by the end of this century.

So at what point of the century
will it become natural

that when we're sick,
when we want health advice,

we turn to a machine
rather than a human being?

And, finally, I want to ask you:

Are you ready for a world
where we have machines that care?

Thank you.

(Applause)