HELLO FUTURE: Can AI End World Poverty?

Speaker 1 (00:08):
It is artificial intelligence going to end world poverty. Everyone
is debating out in Silicon Valley about the artificial intelligence boom,
but the real story is actually unfolding in the global
cullth Hello Future, It's me keV and this is a
dispatch from the Digital Frontier. The planet is Earth. The
year is twenty twenty five. My name is Kevin Surreali,
and I'm actually broadcasting this live from the World Bank
and the International Monetary Fund meetings the IMF. I honestly
can't believe they let me in here. I thought, I
don't know, but here I am, and everyone is not
necessarily talking about money. It's a big, fancy global gathering
of all the world leaders where they talk about money
and technology, but they're really talking about technology. And sitting
next to me is World Bank economist Gabriel. Well, demo
means and I'm so excited because honestly can't.

Speaker 2 (01:04):
Call you Gabriel. Yeah, that'd be great.

Speaker 1 (01:06):
He's like the head Hans show here when it comes
to AI and how the World Bank and the IMF
is looking at this stuff.

Speaker 2 (01:14):
In other parts of the world, we so much talk
about all of the bad.

Speaker 1 (01:20):
To be honest, that people think artificial intelligence is going
to do. But he's got this paper that him and
his team released that really dives into all of the
opportunities that artificial intelligence might be.

Speaker 2 (01:33):
Able to help us with.

Speaker 1 (01:34):
So, Gabriel, first of all, thanks for showing up to
meet the future.

Speaker 2 (01:38):
Thanks Kevin, great to meet you.

Speaker 1 (01:40):
So do you think artificial intelligence could help end world poverty?

Speaker 3 (01:45):
Well, I've been thinking a lot about how AI can
help us address challenges, particularly around education and healthcare in
the developing world, and we know that those challenges are immense.
We know that if you look at low and middle
income countries, seven out of ten kids in those countries
cannot read and understand a simple text by age ten.

Speaker 2 (02:09):
A simple text message, a simple text.

Speaker 3 (02:11):
So you give the US a simple passage to read
and understand, Wow, seven out of kids ten kids in
those countries can't do that by age ten, right, and
so that's obviously a massive crisis. And then on the
health side, we know that about half the world's population
lacks full access to essential health care. So these are
just two of the major challenges that we're facing, and
there's a tremendous hope that AI can help us address
those challenges.

Speaker 1 (02:40):
So this is kind of reminds me of one of
those science fiction movies which I love, which is where
the main characters I just watched stow Away the other
night on I think it's Netflix, which is great, great movie,
but where essentially they get their training through artificial intelligence
and they're able to have a knowledge transfer way faster
than when I was a kid and we had to
kind of do all of this stuff. Is that what
you're kind of talking about, which is using artificial intelligence
to more quickly educate young people in literacy.

Speaker 3 (03:09):
Yeah, that's one piece of it. There are a bunch
of different personalized tools for learning via AI applications, and
my group in particular has been trying a bunch of
those out around the world. We're really at a stage
where we need to figure out what works. It's not
that we can point to a particular AI solution right
now and say every country should be trying to adopt this.
We're really at the point we need to be experimenting, evaluating,
and that's what my group has been focused on.

Speaker 1 (03:38):
So what you write in the report making digital work
for people and jobs, I like they're focusing on jobs.
Is quote during the last thirty years, schools have increasingly
been using education management information systems in reporting and to
foster data driven decision making and education. These systems were
initially rudimentary, paper based and focused on data collect without
the aid of sophisticated analytical tools, which I guess is
a nice way of saying the blackboard, chalk and racers
like what I have. As the global impetus for education
reform and accountability has grown in strength, the emphasis has
shifted to developing more robust digital data systems to track
student enrollments, teacher performance, infrastructure needs, and learning outcomes. But
I guess, as you focus on artificial intelligence, in particular
in parts of the world where there is no digital infrastructure,
how do you seek to create it to begin with?

Speaker 3 (04:32):
Yeah, that's a really good question in fact, So just
two weeks ago we hosted what we called a knowledge
symposium with George Washington University, where we had people presenting
the latest research on education, healthcare, job applications around AI
as they applied to developing countries, And a real highlight
of that was a presentation on an AI tool that's
been used in Kenya to advise healthcare workers on their
diagnosis and on the treatment that they're giving to patient,
and it shows some really impressive results. This is with
the Penda Health clinics in Arabia would encourage people to
take a look and find the research on that they
find a pretty substantial decrease in the errors by healthcare
workers using this AI tool. But the tool only works
because it's in the context of clinics that have electricity,
have Internet connections, and have digital record systems already in
the healthcare club. So it's just basically sort an add
on to the digital systems they already have. Now, those
systems you don't see in lots of other places in
countries where we work, especially in low income countries many
countries and subser in Africa. So those systems require all
the infrastructure, which often we don't have. So for that reason,
I think we need to be thinking about what are
the AI solutions that can work more in simpler context,
that can work just on a mobile phone in areas
where maybe people don't have Internet access. So we're talking
about AI solutions that can run directly on your phone,
not necessarily using the big models that you have to
have the Internet connection for.

Speaker 1 (06:08):
So some of these solutions especially when you look at Kenya.
For example, let's stick with Kenya, a lot of it.
What you're saying is beneficial not only to Kenya, but
it could also be beneficial to other global health concerns.
If we know, for example, more information not just about Kenya,
but a more total picture of health issues around the world,
or literacy rates around the world, that can be beneficial.
So folks here in America.

Speaker 2 (06:35):
As well, sure potentially.

Speaker 3 (06:37):
I mean we're looking at developing applications that have wide use,
and I think there's the potential for there to be
that kind of knowledge transfer from developing countries to the
developed world. And a slightly different topic, Kenya is actually
a pretty good example in that way because Kenya was
the pioneer for mobile money. There's a system called Mpesa
which was launched probably twenty years ago in Kenya that
was really the predecessor for Venmo and all the mobile
money systems that are now very commonly used in the
United States and other high income countries. So I think
there really is the potential for there to be pioneering
work done in developing countries that ultimately could have application.

Speaker 2 (07:19):
That's such a great example. I think a lot of
people forget that.

Speaker 1 (07:22):
I mean, and it's not just about creating just for
something in another part of the world.

Speaker 2 (07:29):
It's also learning and being able to.

Speaker 1 (07:32):
Adopt that for our part of the world as well.
Specifically when it comes to poverty. What have you noticed
in terms of some of the biggest challenges as it
relates to addressing some of the needs for poverty and
education and how artificial intelligence can be helpful for it.

Speaker 3 (07:50):
Yeah, so I think, I mean, of course poverty has
complex roots, right, but certainly lack of education, lack of skills,
which then create challenge for people getting good jobs. I
think that that's sort of number one on my list,
and that's why we're very excited about the possibility that
AI can help us address some of these education challenges.
We have a bunch of different experiments that we're doing
now in different countries. The ones which I think are
the most promising are those that are focused on supporting teachers.
You know, there has been in the past a lot
of excitement, a lot of hype around educational technology, which
ultimately didn't pan out, and the big lesson that we
take from those experiences is that you can't.

Speaker 1 (08:33):
Cut the teacher out of the program, right, nobody wants
to learn from a robot exactly exactly.

Speaker 2 (08:38):
Maybe you want to play with the robot or like
have the you know, I don't know.

Speaker 1 (08:42):
But at the end of the day, it's the same
thing like when I call Comcast, I want to talk
to a human.

Speaker 2 (08:47):
I don't want to talk to you know, press one
or whatever.

Speaker 3 (08:50):
And the teacher has so much you know, even if
you've got an AI tool which can help kids learn,
the teacher is so critical to motivating kids. I mean,
we know that from our own experience. If you think
about your own educational experiences, you think back to teachers.

Speaker 2 (09:04):
I still remember them who really motivated you, right, and.

Speaker 3 (09:08):
So either through kindness or fear, hopefully more through their kindness.
So the programs are most excited about our programs that
are supporting teachers to use AI in the classroom so improved.
We have a program like that. It's an experiment where
teachers are being trained to use AI in ways that
make sense to them to make them more effective in
the classroom. And one thing I've realized actually just quite
recently is you know, we've been focused on these particular interventions,
but already AI is being used out in the wild
at incredibly high levels. You know, we know this from
our own experience. But just a few days ago a
study came out. This is the OECD does an international
survey of teachers and they asked them a bunch of things.
One of the things they asked them was are you
using AI as part of your teaching program? And this
was done in twenty twenty four, so this is not
even a really recent survey, and incredibly high percentages of
teachers in a bunch of middle income countries said they
were already using AI. The OECD average was about a
third about a third of teachers, but in Colombia, in Brazil,
in several other middle income countries, a majority most teachers
said that they were already using AI in the classroom.
So I think it's it's really critical that we figure
out how to help them use AI effectively.

Speaker 1 (10:31):
See this is what I find fascinating and probably why
you're one of the head Han shows at the World
Bank and I'm just an interviewer. Is as you're talking
and you're describing something that was very simple to you
but was an AHA moment for me, which is the
best way to combat poverty is through education, you know,
And it goes back to that philosophy. And I'm no
theologic by any means, but you know, you teach someone
to fish for a day. See, I can't even do it.
You give a man OFFICI sheets for a day. You
teach a man official heats for a lifetime. Education What
I'm hearing you say? And correct me if I'm wrong.
By by emphasizing on education, you're encouraging opportunity to combat poverty. Correct,
And the best way to do that is to upgrade
the education systems so that students are able to.

Speaker 2 (11:25):
Compete in the global economy.

Speaker 1 (11:27):
And artificial intelligence can be a mechanism to do that,
particularly in training teachers.

Speaker 2 (11:33):
Correct.

Speaker 3 (11:34):
Yeah, that's the hope. Now, I think it's a long road.
We still have a long way to go to figure
out the most effective.

Speaker 2 (11:39):
Way to do that.

Speaker 3 (11:40):
And I often find myself, you know, sort of in
the middle road between skeptics, skeptics and enthusiasts. You know,
there's a lot of hype, there's a lot of belief
that AI is going to solve all of our problems.
I don't think that's the case, at least not in
the short term. But I do think it has massive potential,
and that's why we're working to figure out how to
best realize that potentially.

Speaker 1 (12:08):
There's another example, Gabriel from the report quote Education authorities
in Botswana piloted the use of SMS and phone based
tutorials to teach basic mathematic skills to pupils and primary
school grades four to six. Within six months of the
launch of the pilot exercise, absolute and numerous cy rates
had declined by thirty one percent relative to pupils who
had not participated in the program simply by doing text
mathematics classes. But the other lesson is that you're teaching
young people in emerging markets how to engage with technology
as well, so it's a double lesson.

Speaker 2 (12:49):
Yeah, that's exactly right.

Speaker 3 (12:51):
We often talk about the importance of digital skills because
now we've always talked about basic literacy, basic numeracies or
basic math skis is being absolutely critical. It's clear now
that being able to use technology effectively is also an
essential skill that everyone needs to have.

Speaker 1 (13:10):
Awesome, Well, Gabriel, is there anything else you want to mention?
I know you're going to stay on for another episode
as well, but is there anything else specifically when it
comes to combating poverty and artificial intelligence and emerging markets
or anything that gives you hope for the future.

Speaker 2 (13:27):
I guess leave us on a good note.

Speaker 3 (13:29):
Well, as I think there's just tremendous action out there.
There's tremendous energy going into figuring out how to make
AI solutions work, and I see that very much at
the World Bank. So my group is using some of
its money to provide small grants to these country teams
to run these experiments. We had overwhelming interest from country
teams around the world in accessing these grants. We had
a tremendous surplus of demand because people had so many ideas,
so many possibilities about how they might use AI to
address these human development challenges.

Speaker 1 (14:08):
That's awesome, all right, Well, Gabriel Demo means he is
a World Bank economist.

Speaker 2 (14:13):
Thanks for coming.

Speaker 1 (14:14):
On to say hello to the future about hopefully using
AI to end world poverty and really focusing on education
us all. Thank you.