AI Will Change Everything: Exclusive with Nandan Nilekani | Sonia Shenoy https://www.youtube.com/watch?v=zCYQEHBfYuU Apr 24, 2025
Is India ready to lead the global AI revolution? Nandan Nilekani, one of India’s foremost technocrats, to explore the impact of artificial intelligence on India’s future — from government infrastructure to job markets, education, and innovation.
Introduction
i nanda Nilikini is the chairman
and co-founder of emphasis founding
chairman of Aadhaar and a true visionary
behind India's digital revolution a
pioneering force in shaping India's
digital economy Nilikini has played a
defining role from co-founding emphasis
one of India's most iconic IT companies
to architecting Aadhaar the world's
largest biometric identity program
transforming industries and empowering a
billion people please put your hands
together for Nandanda Nilikini
thank you so much for being here
pleasure pleasure juu is an old friend
and I've known him for 30 years we both
began together so it's a pleasure to be
here at the embassy rate wow 30 years is
a long time '90s '90s I met him first
and both of you have come a really long
way
Where is India positioned in the AI world?
but today you know the topic is the
advent of AI and honestly I think
everyone from leaders to businessmen to
entrepreneurs to startup owners are all
grappling with what's happening with AI
right uh so can we start with where we
are at Nandan in terms of see everyone's
talking about India perhaps becoming the
use case hub that's right in Asia for AI
are we there yet and from your vantage
point where are we at no I think uh AI
obviously has caught the public
imagination especially for the last 2
years since uh chat GPD came out and I
think certainly it's a very exciting
time to be there and it's going to make
a huge difference to everyone uh I think
but the use of AI still at least in the
business side is still in the I I feel a
long way to go there and uh we have uh
you know there's been a lot of focus on
models you know the different companies
have models my own belief is that
there'll be thousands of models so
models are really commodities you know
they're going to be you've seen what's
happened from China and so on so lots of
models are going to come out and the
value creation will be on the
application of this technology to either
business or society or government and
that's where the value creation is and I
think India because of the enormous
talent we have the great companies here
who are all customers of uh embassy I
think we have real chance to take a leap
on implementation of AI to really make a
difference both in business and in the
society you've spoken about that many
times in the past as well as to how you
are also involved in doing a lot of work
in perhaps trying to scale AI across the
country right so at scale whether it is
for different users whether it's
agriculture finance fintech language
tell us a little bit about that and um
you know how long do you think this
journey would take well I think you know
to just step back a bit what we learned
when we built digital public
infrastructure like Aadhaar and UPI we
learned how to build at population scale
how to build things for a billion people
how to make it frugal and cheap how to
make it with small transaction size high
volume and so on and we are seeing that
other has 1.3 billion people 80 million
times a day authentication UPI has 450
400 million people 17 billion
transactions a month so we really create
scale so we have to bring the same
thinking to applying AI in India how do
we apply it at the scale of population
which nobody actually has done that at
that scale and to make it cheap to
operate you know how to make uh
inference cost cheap how to make
reasoning cheap how to make sure the
farmer in his village is able to use AI
at an affordable price so that's where I
think we are headed the first
fundamental thing we need to do in India
is language because it's a unique
country we have 22 official languages
and several hundred languages and
dialects you know you move you move 50
kilometer the you know the sort of the
language shifts so I have been
personally supporting uh building the
whole language AI for India and funded a
group at IIT Madras which has built the
largest database of Indian language
called AI for Bharat the idea is that if
we can get Indians to speak to the
computer in their language of choice
dramatically improves access so if a
farmer is able to speak in Bjpuri to his
phone and get the best information on
farming that's where the future is he
doesn't have to read or to key in
anything or touch anything he can just
speak so I think language is a
foundational AI for India everybody will
need I mean India in particular for all
users and then it'll be about finding
use case for example if you look at
finance NPCI has now launched voice
payments in on UPI so I can speak to the
phone UPI and give a instruction to send
money to somebody and that just a start
so you'll be able to give more and more
financial transactions by speaking to
the phone and there'll be some agent on
the phone that will do it for you so I
think we are going to see that kind of
thing happening so it key thing is how
do we take AI apply it at population
scale make it work in your own language
and make it work in a very relevant and
contextual way for your issues so when
you talk about applying it at population
scale the government is also doing a lot
right whether it's the PMISAN program
whether it's Vistar as well can you tell
us how far have we come in terms of what
Government adoption of AI and digital infrastructure
the government initiatives are and what
more do you think can be done yeah I
think the government has been quite
ahead if I don't know whether you
realize but in Aadhaar
authentication we use AI for livveness
detection so when somebody puts a
fingerprint finger we have to make sure
it's a real finger and therefore we do
liveness detection so it's been last
four five years it's there's AI inside
even the Aadhaar authentication but
specifically uh I think PMISAN has been
a great program because that allows
farmers to ask the status of their pay
uh you know the monies they get and
that's really has used you know it has
at least millions of users so that's
already a live case of a chatbot uh
using AI and Vistar is an implementation
for agriculture so that's based on a
open agriculture network and many states
are rolling it out in the next few
months and the idea is that the
knowledge for a farmer can come from
anywhere you know market data will come
from one side crop data from somewhere
else soil data from somewhere else you
know weather data from somewhere else
how do you bring all these various data
together in a way that's consumable by a
farmer in a language of his choice in an
instant conversation so that I think in
the next couple of years you'll see some
really good uh deployments of that and
then we'll I think India will learn from
that then of course there's the other
thing which is happening which is the AI
mission which uh minister wnav has been
talking and leading so they are talking
about supporting the development of
models so I think there's a lot of
action going on in fact recently I think
Mr
was talking about uh 14,000 GPUs and you
know a lots in terms of data sets and
you know everything that's required for
I think the other good thing they're
doing because AI is all about data and I
think there's a systematic plan to
create public data sets so the even the
AI for Barat is a language public data
set so the more public data sets are
there which are anonymized it's not
about personal data then more people can
train their models on that data he also
you know made this joke about how um
India should at some point have its own
LLM models i mean open AI need not be
called open AI anymore i think building
LLM is not a big deal you know uh at
Infosys we have built four SLMs 2
billion parameter SLMs we know the
technology of building at least SLM so
two billion parameters which you can
then take to 7 billion 13 billion 20
billion 70 billion so it's then is
question of more CPUs and more data to
crunch so I think that technology now
has become widespread and that also you
have to give credit to the deep sea guys
because they showed you could build uh
these LLM for much smaller amounts than
the guys in the valley and they have
published all the uh you know knowhow of
how they did it so now the knowledge has
become universal everybody knows what to
do so I'm not particularly too hung up
about saying we should develop an LLM we
should develop no H but if you can get
the use case first make it work at scale
then which LLM you use you can replace
one with the other so there are many
people here I mean most of the people
here are people who run teams people who
How AI will reshape the job market in India
manage large businesses large
organizations right the one big fear
that everyone has in their mind is what
happens with jobs and it's not so much
about the challenges as much as it could
be perhaps about the opportunities what
do you see the way forward not just for
people who are entering the job for job
force which is a huge number um you know
by say 2050 2035 or so but also from a
managerial leadership level what are the
changes that one should do well I think
that's a great question I think
definitely uh AI will improve
productivity of software development we
already seeing it with the copilot a
cursor there are zillian dev devon many
many each day there's a new announcement
ment of some product right so I think
it'll definitely improve productivity
and a lot more can be done with less
people but on the other hand a lot more
things have to be done because AI will
force the envelope on new applications
new use cases so it'll create new
opportunities the other thing is
enterprise AI where companies use AI is
very different from consumer AI so
consumer AI use chat GPD or Plexity or
Google Gemini or whatever but enterprise
AI is going into your firm and
transforming with AI That's a longer
hall because companies have lots of
technology they have duplicate systems
they have legacy stuff from many years
back the data is in silos so they you
know you got to bring all that together
so there's a lot of work to be done
inside firms to organize themselves for
AI so there's a lot of work to be done
while some part of it will get automated
a lot of new things will have to be done
so I'm comfortable that also I think the
Indian workforce you know we have the
largest Gen Z population in the world
and they're all very techsavvy and think
about it the Gen Z was are people born
after 1997 so they were 10 years old
when the iPhone came out uh you know so
they they've been born to this world
right so I think they will quickly adapt
to this AI world I have no doubt I think
so India will actually the number of
people who will use AI to actually solve
real problems this will be the biggest
place so how does India capitalize on
that what are the two or three things
that uh you know leaders or managers or
business owners need to do in order to
make sure that this demographic dividend
is something that we capitalize on well
I think first of all we have to do some
massive uh work on reskilling and
getting people ready for this AI
revolution and therefore firms that
invest in the infrastructure and the
training and the learning and you know
all that that's very important you
cannot do this without investing in your
human capital to get them ready for this
AI revolution so that's like table
stakes for uh everyone also I think it's
very clear now that this AI stuff is
okay for some things but where human
messiness is involved it's not really
all that great so you know you still
your ability to lead to align different
people to the same goal to collaborate
to form teams those human skills empathy
th those skills will never be done by AI
so I think we have to get especially our
leaders to be ready for much better
human skills to lead in this world of AI
so leaders have to be ready with more
human skills uh I hope all the leaders
here are listening but what about the
young u you know perhaps young
entrepreneur or young student who's
entering the workforce how does he
upskill or what are the skills that he
would need in this new AI world the two
skills are human skills people skills
and thinking from first principles you
know ability to get to the conceptual
basis of what you do thinking from first
principles very important because the
middle stuff middle stuff will get
automated you know the you know the
coding you'll use some co-pilot or
something so I think that's both human
skills and thinking from first
principles is what we need but also the
good news is that you are going to see
in India an entrepreneur surge you know
10 years back in 2015 we had maybe a few
thousand entrepreneurs now we have
150,000 entrepreneurs and my estimate is
by 2035 we'll have a million
entrepreneurs because we're creating a
virtual cycle of entrepreneurship you
know companies are going successful then
people leaving starting new companies
that whole thing is in a virtuous cycle
so all those guys will use AI to solve
very key issues in India so I think if
you imagine if a million smart people
solving problems with AI something's
going to happen right i hope so i mean
at this point in time people are
approaching it more with fear and little
bit of you know look it's a massive
change it's a massive discontinuity in
the way we did things and uh I think uh
so people are rightly concerned but I
think we have to embrace it it's look
it's here we can't stop it it's the
genies out of the bottle we have to now
Role of AI in transforming education
work on mastering it applying it for
good purpose so you were talking about
you know the advent of AI at scale right
in India uh you spoke about language you
spoke about agriculture there's another
area that your focus is on which is
education as well tell us a little bit
about that and how do you think because
I mean I have a young child and I do
read a lot of articles about how in the
west uh people are now or at least
schools are now programming using AI to
give differential education to children
depending on what their capabilities are
so just education in India at scale how
can we say and you know actually when
you think about it they there's a set of
children who have access to the best
education they're able to read right
they can they have a iPad to put them to
sleep and all that so that's I don't
know whe is doing that but thankfully
not so what I'm saying is that they are
well served but the real challenge in
India is children who can't read or
write or do math
Because in a public schools while they
go to school they don't learn so our
focus is on them if they are able to
read and write then they can join this
other category of children so we are
using AI to enable children who can't
read or write or to or do math to learn
those foundational skills and they're
using AI to help diagnose what are the
gaps in their knowledge and helping them
to practice it's really like using AI to
practice those skills and it's being
rolled out now in four states uh so I
think if India's future is its human
capital the human capital is not
educated enough to do basic things read
right arithmetic you know then there's
no future right so that's what at least
when you say education we're focusing on
these foundational things because the
people who have get the phones and get
the apps there there's plenty of stuff
for them yeah I agree so there are you
know um a lot of industry veterans
sitting here and in the next panel as
well we'll be speaking to people across
industries whether it's automobiles
insurance CRM etc um talking about
sectors where do you see the maximum
disruption happening because of AI and
Which sectors will be disrupted the most by AI
what do you see as the changes say over
the next five to six years I know five
to six years is a very long time because
lots is happening can't predict that
long but but we are seeing I mean you
know because at Infosys we work with
thousands of companies around the world
we are seeing activity in every sector
certainly financial services is a big
one because that's an area where there's
no physical stuff right it's all digital
so people are applying it for uh credit
they're applying it for chat bots wealth
management assistance you know all kinds
of things so in each area I think there
is uh stuff happening and uh I think
what's exciting is that the applications
are so varied right so a a
pharmaceutical company can use it to
discover new drugs faster or a company
making batteries can use it to discover
new materials faster and so on and so
forth the fact that you can accelerate
the pace of innovation using AI by
discarding all the dead-end options and
honing on what will work is a new thing
that we have now so that that'll be big
and from your vantage point when it
Can India become a global AI powerhouse?
comes to global not dominance but once
again coming back to the point that we
started off with India becoming a use
case hub for Asia for the world for the
world not Asia world for the world are
we there yet you think or is it it'll
happen because it's you have the talent
you have the problems to be solved the
AI knowledge will permeate everywhere
there'll be thousand millions of
entrepreneurs they're going to solve
problems there's no question about it so
what is Nandanda Nilikini going to be
doing next in this uh field of AI i mean
you have always been at the forefront of
every innovation over the last 40 years
whether it was it whether it you're
betraying my age now whether it was it
whether it's Aadhaar UPI AI I mean any
revolution and you know you're right
there in the front no I I like to be
relevant and you know keep ahead of
everything so there are three things I
have three four things I'm working on
which if you want to really know AI of
course is a big part but it's not just
AI I'm working on an idea called the
digital energy grid and how how will
energy get transformed uh from a model
where you have a big producer say
thermal plant that sells to every house
where every home is now a producer of
electricity because they have a rooftop
solar they have a EV car so they can
store it they can buy it they can sell
it they can sell it to the neighbors
suddenly Energy is going to become like
UPI is going to be millions of small
transactions renewable energy huh
largely because of renewable every home
will have a rooftop solar every home
will have a battery and you know price
of energy will fluctuate wildly because
in the daytime when the sun is there
prices will drop at night it'll be
expensive and so on so I think this kind
of a dynamic thing of energy is
something so I've just come out with a
we have come with a paper with the IEA
on this on digital energy so that's one
other thing I'm doing and then uh we're
looking at the idea of the fintet which
is how do you tokenize assets because
just in India for example 50% of assets
are in the form of land i'm sure makes
uh jitu very happy but land is illquid
you know you can't easy if a guy guy guy
guy guy guy guy guy guy guy guy guy guy
guy guy guy guy guy guy guy guy owns a
half an acre in a village he can't sell
it easily it's paperwork all that so if
you can tokenize land assets we can make
it unlock the asset you can sell it you
can get a loan against it so
tokenization is a very important way of
bringing in fungibility and liquidity of
assets so we're working on that and then
another area where from Infosys we're
working on is high performance buildings
uh you know Infosys actually has built
55 million square ft of we are like a
real estate company only 55 million
square feet of office what is high
performance buildings what was it
essentially uh how do we make buildings
super efficient consume less electricity
less water less more sustainable yeah
more highly and that now we have
published all that 15 years of learnings
on how to build high performance
buildings and we feel that in the next
15 years India is get going to get built
out by people like Juu uh how do we get
those buildings to all be super high
performance and reduce their
environmental footprint so that's One
more thing that I do how I I think Do
you have more than 24 hours a day i mean
how many things i don't I don't do any
of this somebody does it i just guide
them you're just the you just get that
brain wave and then the Ureka moment and
then you pass it on i delegate well I
don't do anything myself
okay but that's very inspiring um so uh
coming back to just some final thoughts
before we move on when we talk about AI
and we talk about uh you know large use
case in terms of AI uh any great any big
challenge that you see for India for
leaders for business owners that might
that looks tough right now to perhaps
overcome and something that the
government should work on no I think no
this is not a government thing I think
how do we make the AI accurate it does
not hallucinate it does not have bias
That's the challenge of this AI stuff
because it's okay you know in a consumer
app if it gives one out of 10 wrong
answers it doesn't matter but for a
business a business the brand of the
business is behind it's it's reputation
it's all that is there so when large
firms apply AI they have to think about
reputational thing and therefore they'll
be far more cautious about rolling it
out because if something goes wrong
it'll rebound on them that's one point
the second point is AI is a unique
situation where regulations are
happening even before the AI is coming
you know in most of the areas regulation
like in data privacy internet began 2000
data privacy law came in 2018 18 years
later here in real time only every
country every state is issuing
regulations so when large firms design
AI for the world they have to make sure
how does it work and in a responsible
way as per the rules of that region so
if you're in Europe you may have some
rules UK some rules India some rules and
so on so it's a whole new dimension of
being regulatory compliant from day one
across multiple jurisdictions these are
new things for technologies technology
have not thought about regulation so
much as they now being forced to
absolutely and things are changing by
the day earlier things were changing by
Nandan Nilekani’s world view: Future of tech, India, and humanity
the decade then by the year every day
every morning you get up some something
new either there's a tariff or there's a
AI model
and then the next day either the tariff
gets you know reversed or something else
happens in but what is your worldview by
the way I mean there's so much
uncertainty with respect to tariffs and
what's happening in the west where do
you think all of this is going I don't
think anybody knows if anybody claims he
or she knows then I don't think it's
true I think we just have to ride ride
it but you know there's so much in our
control I mean you know we can build
great companies we can build uh new
solutions we change lives of people
that's all within our grasp so I'd
rather focus on things we can control
than try to worry about things that I
have no control over okay one final
piece of advice since you have so many u
you know like I said so many leaders
here one piece of advice that you'd want
to give them with respect to AI and how
they should sort of move forward i think
they have to be keep themselves current
and curious because the the avalanche of
knowledge now is extraordinary right
we're getting so many things and how do
we uh you know go through all the noise
and find the few signals in the noise
because noise tends to distract and then
get that's why the first principle
thinking is very important and really
keep on top of what's happening that's
very very important for today's leaders
because if you take the wrong turning
could go off into a culdeac so I think
it's very important to be on top of
what's happening and distinguish the
signal from the noise and focus on the
signals all right ladies and gentlemen