AI Superpowers for Sustainability

i were to ask you do you care

about the environment about

sustainability

most of you would probably agree

however if i did pose the question

is artificial intelligence ai

falls for good many of you might not be

so sure while it’s been

a hype subject for some time

some aspects of it are quite

controversial

many people believe that humans would be

replaced by by robots that they would

take

over our jobs make our life on this

planet more difficult

and reduce the chances

for survival of mankind

so what is true let’s find

out i invite you to

join me on a journey to see

the the details underneath to go beyond

some of the mysteries and misconceptions

and look at the data like we often do in

data science

so let’s take a start

when i was a young boy growing up

in munich one thing that i was

fascinated by

was the tv series night rider with the

ai enabled car kit

kit was able to reason make his own

decisions

and chase the bad guys and eventually

turbo boost

out of any tricky situations that would

arise

for me it’s always been a

goal of mine to create solutions that

help mankind

make life easier more sustainable

little did i know that eventually i

would realize my dream

of building such solutions in my

professional career

that took me around the globe today

i’m a ai thought leader and director for

and fintech at pwc and i have the

pleasure of

working on solutions like that with

clients in asia

u.s and europe and it’s

from that angle that i kind of brief ai

on a daily basis

so it comes very natural to me but i

understand that

i should define the term for the rest

let’s start with the artificial part

that one is quite easy to understand the

more tricky one

is defining intelligence many have tried

and when we look up in merriam-webster

we can find

five different definitions including

logical

mathematical kinesthetic linguistic and

so on

but let’s start with the first one the

ability

to learn or understand or to deal with a

new

or trying situations such as climate

change

and sustainability

so not many of you

might agree that data is the new oil

but it can’t be ignored that artificial

intelligence

will fundamentally change the way that

the world does business

to quote my favorite author on the

subject kaifu lee who said

that pwc estimates ai

deployment will add 15.7

billion dollars to global gdp by 2030.

of course you might not believe these

big numbers

but you would still understand the scale

of change that is about to happen here

and it’s exactly those

powerful tools and powerful methods that

has that kind of fascinated me for a

long time

to kind of see how those could be used

to speed up our work in sustainability

and do something good for the planet

when it comes to the sustainable

development goals that many of you know

it’s

kind of been around for for

some time and it’s for

my time in a silicon valley

company in 2014 where

my extended team had the honor of

contributing to the early drafts to them

that i kind of started being fascinated

by the work of the u.n

in in this in addition it was my parents

who from an early age instilled

the care for the environment the

recycling

the kind of awareness of what happens

around us

something that i took with me on my

journeys and my work around the globe on

five different continents that i would

always look out for

the environment but also how people

recycle locally and what they do

to protect when you look closely you

will

find out that only free namely

13 to 15 of those goals are related

directly to the environment

so the climate action the life

underwater and the life on land

but when you take others into account

you realize that

all of these goals are interlinked in

for example

looking at a person having

enough clean water enough

food on the table enough

job security enough income

only that kind of person can really care

and make enough

effort to protect the environment and in

that way

we next look at what impact ai

might have on those goals

so on the first instance ai

might have a detrimental effect to

sustainability into the carbon footprint

when you think about some of the big

data centers and also training ai models

when we look at this graph in particular

we see that taking a flight

from new york to san francisco is only

a small impact on the carbon footprint

but training a ai model would be 300

times as much

now for me having trained many

ai models with my team this is really

comparing apples and oranges

what you see in this kind of ai model is

probably something

that thousands of people worked on

together in a corporate setup

that is a one-off language model on a

vast scale and that cannot be compared

a individual taking a flight however

when we take this into account we kind

of have to recognize that the public out

there

still has a fear of ai like we mentioned

earlier

and that there are many misconceptions

that we need to

really fully understand and then debunk

and with that we move on to more recent

research

earth.org and spring in nature

mapped out the 17 sustainable

development goals

into three different categories in terms

of the environment

that we heard earlier economy and

society

what is quite telling here is that while

ai might have a detrimental effect on

some

the potential or

the ability to enhance the chances of

achieving those goals

is much more powerful so when you look

at it

there’s three to two to three times

um the chance of having a positive

effect with the eye

than there is to to reduce or making

these targets more

more difficult to achieve

and with that let’s look into some of

the research that

my colleagues in the international

network have done throughout the

lockdown last year

when you look at sustainability graphs

and climate change you typically see

a increasing curve namely looking at

the temperature going up by certain

degrees of celsius

and you look at modeling in terms of

what that might mean and at which point

it might be irreversible now what we’re

seeing here in

in contrast is ai

helping all the regions around the globe

to reduce their greenhouse gases and

average by

four percent which is quite phenomenal

some regions even going down to six

percent so we can say

that there there will be a tremendous

impact that

if my colleagues that model that are

correct and they modeled it in quite a

detailed

manner so let’s look in terms of

what other impacts are there one thing

that we see

and i’ve been debating this over the

last 10 years pretty much

is the impact of ai on jobs so

the debates are typically will ai

replace

or create jobs one thing that i realized

by debating this all over the place is

that it

very much dependent on the industry

the the sector and the country that a

person is based in to see

whether there would be an impact a

replacement

so while it is true that some lower key

jobs will be becoming obsolete it’s also

true that other jobs will be created

such as mine in terms of being somebody

who’s explaining ai

but equally other people would be

maintaining and creating those models

and other people would benefit from the

fact that certain industries are

pushed by this now to kind of

close the data piece off let’s look into

what are the four key industries that

are benefiting from the use of ai

starting with with energy we

measured the the impact here in terms of

smart grids and controlling system

to really particularly for renewable

energy to predict

usage patterns then also looking into

transport where

ai can help to enable

supply chains as well as autonomous

driving

and in fact reducing some of the biggest

traffic jams around the world

by controlling those systems but then

also to move on to

agriculture where with the help of smart

sensors

alternative data satellites and a

ai system we can predict weather

patterns we

can look into reducing pesticides we can

look into reducing use of water

which brings us to the last point in

terms of the water industry

here ai can help by

spotting pollution early on by

predicting rainfall by

kind of making sure that the the water

is used in the most efficient way

so overall what we have here is a system

where

without the new insights that ai

generates

it’s very difficult to make those

changes without it

in some ways when we want to follow

these sustainable development goals

we don’t know which path to take so we

really need those insights

to come full circle we

seen some of these changes that are

possible also the details behind the use

cases

but when we go back to merriam-webster

the second definition of intelligence is

the ability to apply knowledge the

things that we just heard of and

kind of learned to manipulate or to

improve

our environment such as that we take on

the things that we’ve

just experienced and try to change for

the better

and with that we are thinking in terms

of what can we do

at the local level to

connect sustainability and ai

so whose responsibility is it really to

connect those two forces it is those

people that

build the models like me and my team and

many many others

but also the people that are using the

system in terms of flagging up

inconsistencies biases restrictions

things that are restricting the use to a

certain

people group to make sure that it’s used

in the right way

so firstly when we call to action at the

local level

it’s for us to collaborate

to link academia to link the corporates

the public

all together to create a better and

holistic understanding

how ai can really help sustainability

for me personally kind of led

a lot of engagement in this and coming