Latest / Low to Grow / Explainable AI and the Power of Quiet Thinking with Forbes Under 30 Joyjit Chatterjee

Transcript

Explainable AI and the Power of Quiet Thinking with Forbes Under 30 Joyjit Chatterjee · Low to Grow

0:00
There are so many young people in the world who are introvert
0:02
who think often before they actually take any action, before
0:06
they actually kind of speak, speak up in front of others.
0:09
So there is that kind of hesitation which might come
0:12
across as a lack of confidence, but actually a thinking more
0:15
before you speak up. That is actually a quality which
0:18
you can leverage to do things like.
0:21
Welcome to Low to Pro, the podcast transforming life's
0:24
toughest moments into opportunity for growth.
0:27
I'm Annie, a Folks under 30 technology founder who's
0:31
entrepreneurship journey brand parallel to a mental health
0:33
awakening. In every episode, I sit down
0:37
with inspiring individuals and delve into how they managed to
0:41
turn their personal or professional challenges into
0:44
opportunities for growth. If you're facing uncertainty in
0:47
your life, feeling down, or simply need a kick of
0:51
inspiration to keep moving forward, this is your space for
0:54
the honest and uplifting conversations that you will want
0:57
to hear. Hit follow so you never miss an
1:00
episode and let's dive in. Today on Lotic Road, we're
1:05
joined by someone whose story is as humbling as it is inspiring.
1:09
Dr. Georgette Chatterjee was once a quiet, shy kid, often
1:14
underestimated by classmates and adults.
1:16
He was even told that he lacks what it takes to survive in the
1:20
real world. But through personal patience in
1:23
a resilience and a quiet determination, George, it proved
1:27
that being an introvert is actually a superpower rather
1:30
than a weakness. Now, George.
1:32
It is a Forbes on the 30 honoree in Manufacturing and industry,
1:37
also working as a lead data scientist at EPAM and is a
1:40
global speaker on AI and sustainability.
1:43
He has worked on smart factories across continents, earned
1:47
international recognition for his research and also shared
1:51
stages with Fortune 500 leaders, all the while staying grounded.
1:55
Today is a conversation about the power of patients listening
1:59
deeply and just letting your work speak louder than your
2:03
voice ever could. Georgette, I am delighted to
2:06
have you here on Low to Grow. Who do you think will benefit
2:09
the most from listening to our conversation today?
2:13
There are so many young people in the world who kind of are
2:16
introvert, who kind of think often before they actually take
2:19
any action, before they actually kind of speak, speak up in front
2:23
of others. So there is that kind of
2:25
hesitation, which might come across as a lack of confidence.
2:28
But actually it's like you're just trying to be humble.
2:31
If you are trying to be grounded and if you are trying to give
2:34
others the stage before you kind of jump in and make your views
2:37
kind of evident, I think that's a lot of young people these
2:40
days. They would understand this, that
2:42
it really helped to understand that being introvert thinking
2:47
more before you actually take any steps further, before you
2:51
speak up, that is actually a quality which you can leverage
2:54
to do things like research, do things like innovation.
2:57
So how young professionals can actually transition from the
3:01
society which might treat it as a lack of confidence to actually
3:05
something, as you mentioned, like a superpower which will
3:08
actually help you transition and grow in your career.
3:10
So yeah. Wonderful.
3:13
Thank you. I am very excited to delve deep
3:15
into that with you. Let's start at the beginning
3:18
then. So JoJo, you've spoken about
3:20
being shy and introverted when you were a child.
3:23
How do you feel that those early experiences actually shaped the
3:27
way that you approach communication and leadership in
3:31
your work today? I would say, you know, like
3:34
during my childhood, because that's where most of our life is
3:39
shaped, right? When we are kids, that's where
3:41
the beginning of our life is shaped.
3:43
I was very shy, like really, really shy, really quiet.
3:47
I would hesitate to even ask simple things in front of
3:50
teachers and to speak up when I maybe was struggling to
3:54
understand something on stuff, stuff like that.
3:56
So I wouldn't really be that kind of outgoing in terms of my
4:00
personality. So that was something that was
4:02
by nature only, you know, like it, it wasn't something that I
4:05
had kind of learned or anything, obviously.
4:07
It was something that my personal nature was always like
4:09
that I used to speak whenever I felt like this is like the right
4:13
movement. These are the right people.
4:15
Like a comfort zone, right? I had a particular comfort zone.
4:18
Maybe it might be my parents, maybe some best friends whom I
4:21
would be most comfortable in speaking to, but not really to
4:25
strangers, not really to new people, students, other
4:28
classmates, especially the ones who are more outgoing.
4:31
They would mock me and they would tell me this nature, it
4:34
won't help you in your future career.
4:36
It won't help you in your life because really, if you don't
4:40
speak up, if you can't like put in your views or if you don't
4:43
ask questions, then you you will struggle to grow in your life.
4:47
It came across as something that was more destructive, you know,
4:50
like it, it would actually hamper and it, it could lead you
4:53
to depression, it could lead you to anxiety and stuff like that.
4:56
So that was kind of, you know, like the moment that made me go
5:02
into the low point of my life, like during childhood when I
5:05
realized that, OK, like, society views introspection and
5:10
quietness as a weakness rather than something that is just
5:14
coming across from our nature, which we can't really change
5:18
when we are kids. Do you feel that there was a
5:21
turning point where you actually realize that you're naturally
5:25
quiet? Nature was actually an advantage
5:28
and a strength that you have. Yes.
5:31
So I would say that once I moved to university, that was the
5:36
point when I realized that there is a bigger picture behind this
5:40
nature, which means that I have got certain skills which some
5:44
other people might not have in terms of giving others the
5:48
opportunity to speak up rather than being the loudest person in
5:52
the room. It makes me understand things
5:54
deeper and I can go into the intricacies into the details
5:58
which other people cannot. So that was where I got really,
6:01
really interested into research right from my undergrad, right?
6:04
So as soon as I moved into UNIF in my undergrad, I kind of
6:08
published quite a few research papers, which is quite rare
6:12
there for undergrad students. And that got me interested in
6:14
participating in conferences and giving presentations, stuff like
6:19
that. And obviously that led to the
6:21
PhD and then even more upliftment as a researcher, as a
6:25
scientist. So yeah, I think that low point
6:29
in itself, it became like the strength when I realized that,
6:32
OK, this is something that could be a trait of a scientist
6:36
growing up in their life. That's so interesting.
6:39
And I remember you actually told me when you were a child.
6:43
Other adults told your parents, or rather they asked your
6:46
parents, how will you be able to survive in the real world?
6:50
Do you remember as a child when you heard that, how you actually
6:54
felt? Oh yeah, it was terrible.
6:57
It felt like I couldn't do anything literally for days that
7:00
that thing was in my mind for like several days.
7:04
Like I even like struggle to eat honestly, like if if someone
7:08
tells you that right, like when, especially when you are a kid,
7:10
when you don't know how to tackle these things.
7:15
Any negative perception that comes across from not just other
7:20
children, other students, but even teachers, right?
7:23
I, I think the sad thing is even teachers thought that, OK, this
7:26
is like something that would be negative, negative thing enjoys
7:31
it. But fortunately my parents were
7:33
quite supportive. My grandfather, who is no more,
7:36
but he was like, he was very much like me.
7:39
I think like, so maybe it comes from genetics as well.
7:42
He was also a writer, author, you know, like he used to edit
7:46
books, he used to write stuff and stuff like that.
7:49
So I think like that quality of being a writer, being a
7:52
researcher and all of that, it comes from genetics and kind of
7:56
my fam family was quite understanding that.
7:58
OK, maybe Joy just grandfather was also quite quiet and like
8:02
him during his all his life. They gave me the opportunity to
8:06
not take that as a negative thing, but rather take that as
8:09
an opportunity to work on myself.
8:11
Maybe this is something that I can work on and I can start to
8:16
speak to strangers. I can start to collaborate with,
8:19
you know, other people during my education journey and stuff like
8:23
that. So yeah, like the parents helped
8:25
me a lot in recognizing this as something that I could actually,
8:28
you know, like, improve. Based on your experience, then,
8:32
Georgia, to someone who was a quiet child, if some of our
8:35
listeners out there have young children and they notice that
8:39
maybe one or two of their children tend to be more on the
8:43
shy side, so similar to how you were when you were younger, how
8:47
would you advise the parents then to help their children to
8:51
grow out of that shyness or to really take ownership of that
8:54
shyness? I would say that definitely
8:57
don't push, push the children to like completely change their
9:02
nature. You know, that would be the
9:03
very, very first advice for the parents that let them be their
9:07
authentic self. There is nothing wrong with
9:09
being shy. There is nothing wrong with
9:11
being an introvert. So it it is actually something
9:14
which they can leverage as their biggest superpower, as their
9:18
biggest quality as they continue to grow in their life.
9:21
What could be improved is maybe yes, like speak at the right
9:24
moment. Like instead of being completely
9:26
quiet all the time, you know which I realized that you need
9:29
to speak at the right time, right?
9:30
Otherwise will will take you for granted.
9:33
So speak at the right moment, the right opportunity whenever
9:36
you see like, OK, this is the right point when I need to jump
9:39
in, then jump in. But otherwise, give others the
9:42
space. Also give give others the
9:44
opportunity to speak up first before you kind of jump in to
9:49
that. So I would say that let them be
9:51
their real self, Let them be authentic, but kind of inform
9:55
them and help them to become more outgoing by maybe
9:59
encouraging them to participate in, you know, like events,
10:02
encouraging them to go out in competitions, be it poetry, be
10:06
it singing and all of that. So hobbies, I think they help a
10:10
lot. I used to play guitar and all
10:11
when I was, you know, growing up.
10:13
That helped me a lot to come out in front of the public, right,
10:16
and increase my public public perception in terms of being
10:21
someone who is kind of more interactive with the public and
10:23
the society in general. So yeah, encourage them to do
10:27
whatever hobbies your children are interested in.
10:30
Jojit, I know that you were selected for reforms and a 30
10:34
list in manufacturing and industry.
10:37
Could you share with us a bit more about your journey there?
10:41
What took me into Forbes under 30 is definitely my research,
10:45
you know, so it started with my PhD at the University of Hull,
10:48
which was in AI for the renewables industry with like
10:52
wind farm operators and organizations like the Offshore
10:56
Renewable Energy Catapult in the UK to develop AI algorithms that
11:00
can perform explainable predictive maintenance.
11:04
So not just like highly accurate maintenance, but also like you
11:07
can get like transparency and accuracy both in your decisions.
11:11
I used variety of concepts ranging from causal inference to
11:15
natural language generation to knowledge graphs to bring trust
11:18
and confidence into the AI models, which is something the
11:21
industry, especially as of today, it's very, very much
11:25
looking forward to. No one wants to use black box AI
11:28
models. They want trust and confidence
11:30
in your decision making. So that research got quite some
11:34
attention. It was published in lead leading
11:36
journals, well sighted. After my PhD, I continued my
11:41
research trajectory in explainable AI.
11:43
I worked at Racket, which is one of the leading consumer goods
11:46
companies, which makes Jet all eyes all and all those products.
11:50
So I worked there and I actually applied the same explainable AI
11:54
foundations into the work in the factories of the future of the
11:58
future, how we can reduce waste in the factories, how we can
12:02
improve or optimize the shelf life of products that they make
12:05
in R&D and those kind of areas. It kind of got that attention of
12:10
Forbes and like I, I think that's, that's where it led to,
12:16
you know, like the manufacturing and industry category.
12:19
And what really helped, I think to get that attention was my
12:23
participation in several conferences, writing several
12:27
research papers in leading journals, even organizing
12:31
workshops and social events at leading AI conferences like
12:34
ICLR, new ribs and those kinds of things.
12:37
So that really helped to, you know, bring together like
12:40
industry plus academia plus public sector, like everyone
12:43
could see that, OK, there is some guy who is working on AI
12:47
that connects to sustainability, that connects to explain
12:50
ability. George it for some of our
12:52
listeners who might be just starting to understand to learn
12:56
about what AI is, what does black box AI mean?
12:59
And also, how would you explain what explainable AI is to
13:03
someone who is completely new to this area?
13:06
In very simple terms, I would say that everyone is using like
13:12
ChatGPT right now. Everyone knows of it as AI, but
13:15
behind the scenes it's using neural networks.
13:18
And neural networks are very complex, like machine learning
13:21
models, actually deep learning models which work very much like
13:25
how your human brain functions. So they are based on the
13:28
operations of the human brains. The analogy is like that, but
13:32
they are kind of algorithms or models which can make highly
13:35
accurate decisions. But as of today, it's
13:38
practically infeasible to understand the workings of how
13:41
these large models work or operate.
13:44
That's like the black box behind this.
13:46
So you you can use ChatGPT, you can type in a prompt and you can
13:49
get a response. You also can get the reasoning
13:52
in the latest models that are out there.
13:54
You can use O3 and stuff like that to get the reasoning.
13:57
But that is something that the model doesn't really understand.
14:01
The text that goes into the model, the data that goes into
14:03
the model, the model only sees it as a bunch of numbers.
14:07
It is just tokens for the model and it doesn't really understand
14:11
the things in the same sense as a human does.
14:14
So that's like the black box. And in the industry, for
14:17
example, the wind industry, like I was talking about in my PhD,
14:21
you would have data from wind turbine sensors which could lead
14:24
to decisions on faults, right? You could use that data to build
14:28
predictive models that can tell you that in the next 24 hours
14:31
the gearbox of the wind turbine is gonna fail.
14:34
But what actually leads to that message, like what is making the
14:37
model think that in next 24 hours the gearbox would fail, is
14:41
where explainable AI comes into play.
14:44
That is where techniques you could use to decipher that these
14:48
are the most important parameters from my wind turbine
14:51
sensors, which actually contribute to the fault.
14:53
And same goes for factories. I can predict with an AI model
14:56
that the product that I'm making right now, it is going to have
15:00
more waste produced in the next year.
15:02
But what actually leads to that waste?
15:05
What is the root causes of that waste and how I can actually go
15:08
ahead and make changes at the ground level in my factory so
15:13
that I am minimizing the waste and maximizing the profits and
15:17
revenue for the business? So that's where explainability
15:20
comes in. That's where causality comes in.
15:22
And you actually don't just make predictions, but you can get
15:26
actionable insights out of the AI in the Gen.
15:28
AI age like it's, it's quite important to use these models to
15:33
make sure we use these models in the trustworthy and responsible
15:36
manner. In your opinion, George, it what
15:41
are some of the current misconceptions or myths about AI
15:46
that really, really annoys you? One of the biggest
15:49
misconceptions is that Jenny I is all AI.
15:54
Jenny I is such a buzzword in which billions of dollars of
15:58
investments are being poured in as AVC.
16:00
You will see so many varieties of Jenny I products that have
16:04
been released into the market and which makes a lot of the
16:07
people especially at the senior levels, right like the VPS and
16:11
the C level leaders, they might not know that Jenny I is not
16:15
something that can accomplish everything that we need to
16:18
change in our business. So the hype and the buzzword
16:21
that has been created and on Jenny I, it's like it has
16:25
created the biggest misconception that Gen.
16:28
AI can solve every single problem that we have in the
16:31
business. To be honest, in my experience,
16:34
most of the problems that we face in everyday life is
16:37
something that we don't even need Gen.
16:39
AI for. How the R&D operates, how the
16:43
supply chain operates, how the logistics operates, how the
16:46
finance operates, different business functions.
16:49
They actually have either structured or unstructured data
16:53
which you can use to build traditional machine learning
16:56
models or even statistical models and apply data science on
17:00
top of it. And then where Chennai could
17:02
help the most is when you have to build that human interface to
17:06
any predictive model to create like a conversational layer or
17:10
to create like an interactive question answering system.
17:12
That's where it could be quite powerful.
17:15
But if you want to do something trustworthy and you can't like
17:19
really use an LLM to do predictive maintenance in a
17:22
factory, which would be very, very risky.
17:24
And you might just get more false alarms and missed
17:27
detections than actually saving money by using AI.
17:32
What is the area of focus for your research or your work right
17:36
now? I'm quite focused in using AI
17:40
for different real world use cases in the connected lab
17:43
space, like in the labs of the future, how we can use AI to
17:47
optimize the new product development process.
17:49
So whenever you are making new products, then you are trying to
17:52
ensure that the quality of the products is the highest, it is
17:55
safe, it is efficacious. You can leverage a lot of data
17:59
that comes from your historical laboratories, the data that
18:03
comes from formulators and analysts to who make all these
18:07
new products, which can be used as a knowledge pool to really do
18:11
something impactful and it could help you create your simulate
18:14
new scenarios, new products that OK, I'm making a new product
18:18
with these active ingredients. What would be the impact if I
18:21
change this, if I change my container, if I change my sample
18:25
from from a tablet to a capsule and stuff like that.
18:27
That is one of the areas. Plus I am also quite heavily
18:31
involved these days in grounding AI models into domain specific
18:35
knowledge. Quite important in the Gen.
18:36
AI age because we cannot really use a foundational model like
18:40
GPT 4 to understand the context of our own business, of our own
18:45
industrial specific use cases. Grounding it on knowledge by
18:49
using knowledge graph. I am quite interested these days
18:52
in creating my own knowledge graphs.
18:54
For example, for different use cases, I can model a wind
18:58
turbine as a knowledge graph, as a system with different sub
19:01
components with different alarms and different connections
19:05
between these alarms to the sub components and things like that.
19:09
Then I can connect my knowledge graph to an LLM which will give
19:12
that LLM the historical context or the knowledge that okay, this
19:16
is my domain specific knowledge or the foundations which the LLM
19:20
should refer to whenever it is making any decisions.
19:23
And that's where Agent TKI again, a big buzzword comes into
19:27
play. What I am kind of doing right
19:29
now is trying to connect it all to additional ML plus knowledge
19:35
graphs so that we build models that are, you know, like safe
19:39
and like I was telling, explainable.
19:41
So any, any decisions I get from an LLM, I can actually attribute
19:45
it to the responses I have in my knowledge graph.
19:48
So that I know that OK, there is actual enterprise data that
19:51
tells that this is the way to resolve the fault in a wind
19:55
turbine, or this is the way to optimize the incorporation of
19:59
vitamin C in a new tablet that I'm making.
20:02
I'm going to ask you to briefly explain what some key terms are.
20:06
So could you explain to our listeners who might not be that
20:09
familiar with AI, what foundational models are?
20:12
Could you also explain why is necessary to ground AI
20:17
frameworks? And lastly, could you also
20:18
explain what AI agents are? Foundational models are
20:23
basically your large models that are trained with any almost
20:27
anything and everything from the Internet, right?
20:29
So GPT that we all use. It's a foundation model which is
20:33
built on huge amounts of data from across the Internet, which
20:36
might even include Reddit, which might include Wikipedia and
20:40
stuff like that. So it has got a lot of
20:42
knowledge, but it is not something that is specifically
20:47
going to have your competitive advantage because it doesn't
20:50
know your data. It doesn't know your problem
20:52
statement. So that's like a foundational
20:54
model for you. Why we need to ground these
20:57
models in our own data? It's because of hallucinations.
21:00
So as as you know, ChatGPT we all use a lot of the time, there
21:05
is a huge possibility it will hallucinate, it will make
21:08
plausible real sounding responses which might actually
21:12
be completely inaccurate. And in factual it is tolerable
21:16
when we are using it to generate images or do some fun stuff,
21:20
experimental stuff. But when we use it at the
21:23
corporate level, at the enterprise level, it comes with
21:26
huge implications. There is EUAIA and there is
21:30
regulations that come in usage of AI models like FDA and EMA
21:35
and stuff like that. There is like huge regulations
21:38
that are coming up these days and huge penalties for
21:41
businesses if they do not adhere to these regulations.
21:44
So that's why it's quite important to choose your own
21:46
data to ground these models. Maybe it would be a lot less
21:50
powerful. It cannot do everything in this
21:52
world, but it can do something very well.
21:54
That is where the power of grounding your model comes into
21:58
play. And agentic AI basically aims to
22:01
make sure that LLMS are something that can work
22:05
autonomously. So instead of us having to go in
22:08
and type in a question every time and get stuff accomplished
22:11
every time with a prompt, LLMS can work autonomously when they
22:15
are agentic. So they can actually approach
22:17
your data, they can approach your systems, they can actually
22:20
work 24/7. They can keep an eye on any
22:24
tickets that might be raised in your company, any issues that
22:27
are going on and stuff like that.
22:29
So you could have an HR agent that looks into HR tickets.
22:33
You can have AAR and D agent that is going to, you know, look
22:37
at your R&D data from the labs and it is going to keep an eye
22:41
or audit your R&D data to make sure that all the quality
22:44
control, all the quality assurance and test like that
22:47
they are being done properly. That's the power of agent like
22:50
the autonomous nature and the ability to work without human
22:55
intervention or with minimum human intervention.
22:57
It's quite important to realize that these things, if you leave
23:01
it without any human intervention at all, then there
23:04
is a huge possibility that it will lead to very, very, very
23:07
costly failures or scenarios for you.
23:11
So it's quite important to make sure that the agents, whenever
23:15
you are using it for something quite critical, for example, you
23:18
are using it to give refunds to a customer, then it's important
23:22
that you keep an eye and make sure that people cannot misuse
23:26
your AI agent and make it do things that you do not intend to
23:30
have it do in front of them. OK, understood.
23:34
In a way, it sounds like foundational models.
23:37
For example, if we take ChatGPT, that's probably something that a
23:41
lot of people are quite familiar with.
23:43
It's a type of conversant AI because it's a chat box, You
23:46
message it and it messages you back.
23:48
It's been trained on a foundational model, which means
23:51
that it's been trained on basically most things that's
23:54
publicly available out there. So it can be compared to a
23:57
worker being trained in a huge variety of different techniques,
24:01
but not really specializing in anything.
24:04
So there's worker, it can do work, but if you ask it to
24:07
create a very specialized product, it will really struggle
24:10
because maybe the skills that it needed to have hasn't been
24:14
focused yet. And this is where your work such
24:17
as what you were doing in grounding these models come in.
24:20
So it's really actually teaching this AI more specialized, more
24:24
focused, more relevant skills for the particular job that you
24:27
want it to do. And then finally, with the AI
24:30
agents is actually turning AI models or turning AIS into
24:35
almost like individual workers, but know what they need to do
24:39
and they can just do it. And you don't really need to
24:41
have 24/7 oversight or to have as much hand holding as
24:45
possible. Would you say that's like a fair
24:48
comparison? Yeah, yeah, definitely.
24:50
So it's, it's basically kind of ensuring that you go from like
24:54
just models that work with a developer or a data scientist,
24:58
kind of always keeping an eye on those two models that would be
25:02
more accessible or democratized, right.
25:04
So I think AI agents is really helping to democratize these
25:08
models and bring it into use by people who are not like orders
25:11
traditionally or not from the technical background.
25:15
Outside of work, what do you think one of the most exciting
25:19
potential applications of AI tools into humans day-to-day
25:24
lives could be? AI should be reusable in
25:28
scenarios where it couldn't really reduce the amount of
25:32
things that we humans have to do.
25:33
Not really doing everything for us.
25:35
Because I feel that honestly, if AI is doing everything for us,
25:39
from like writing an e-mail to reading an e-mail to writing a
25:43
presentation to even reading a presentation, then no one will
25:46
do anything. That's not the future we as
25:48
humans would want or for our future generations.
25:51
But what we really want is definitely cases where AI is
25:56
really, really great at to help help people who might be having
26:00
diseases and performing diagnosis and supporting
26:04
doctors, the medical specialist and reaching out quick
26:06
decisions, quicker turn around times in terms of you know, like
26:10
giving the patients their future prognosis and diagnosis and
26:13
stuff like that. It is really adept at drilling
26:17
down into the data. There is so much data around the
26:20
world which we have from our daily life as well, which could
26:23
be used to create so many different useful things.
26:26
Especially from the multi model modality bit like it can help
26:29
them to understand things that are going on around the world.
26:33
If some person is blind, then they can actually use the AI to
26:36
interpret what is happening in their surroundings and it would
26:39
be a completely game changing thing.
26:41
It would be a new life for them. That is where AI could be very
26:45
helpful. Things like Excel and you're
26:48
building a spreadsheet and stuff like that and you kind of have
26:52
to do lots of calculations and create lots of formulas and
26:56
things like that in which AI could really help you.
26:59
It could assist you to create this formulas in a shorter time.
27:03
It could help you to actually focus on the analytics rather
27:07
than the developments. It would help you through
27:09
decipher the data. You can perform deep research on
27:12
the data set and use it to extract meaningful insights for
27:16
your company. How do you think that someone
27:18
who is socially anxious might be able to use AI tools to help
27:23
them better go about their daily life?
27:26
A great example would be things like the advanced voice mode
27:29
that you have got in ChatGPT these days.
27:32
You can literally ask it, right? To speak to me like a teacher,
27:35
speak to me like a boss in my company, speak to me like
27:39
stranger and stuff like that. So you can actually use it to
27:42
simulate a person, a human who you have never met before, for
27:47
an introvert, for a shy person, that is perfect for knowing
27:51
beforehand what might it be like to interact with such a person?
27:54
What might be it like to interact with the CEO of a
27:57
company? What might it be like to
27:59
interact with a potential investor and stuff like that.
28:01
So it's really be useful, especially I think for young
28:04
professionals, for young people who haven't seen life in and
28:07
out. We, we are very much at the
28:10
beginning of our career. We are very much at the
28:12
beginning of our journey. We haven't seen like the
28:14
struggles that come later on in life and that we learn from
28:17
experiences. So I think with the AI, we can
28:20
actually understand that. OK, like how is it really like
28:24
to be in the shoes of ACEO? How is it really life to like be
28:30
in the shoes of maybe someone who is kind of, you know, like
28:34
older and you have like children and how do you interact with
28:37
your children and stuff like that?
28:38
So how how do I experience that without even having reached that
28:42
time or that age? So I think that is where, yeah,
28:44
I could really help the introvert people to become like
28:48
more extrovert. And from my experience, as you
28:51
start to speak to people, right, you kind of start to become more
28:55
outgoing. There is no way it cannot
28:57
happen. Speak every day to the advanced
29:00
voice mode, just for 15 minutes every day.
29:02
And I think literally in one month you will be much more
29:06
extrovert and much more outgoing.
29:09
That's a very creative way to use the existing tools.
29:12
And as you were speaking, I was actually thinking, yes, that is
29:15
so right. It's a way for us, I guess, to
29:18
also be able to more deeply emphasize with people whose life
29:23
experiences might be very different to our own.
29:26
And I think that's definitely a great way to garner more
29:30
understanding and more patience, especially when we have to, you
29:34
know, for example, work in a very multicultural group or
29:37
perhaps work with a group of people who are just naturally,
29:40
by default and by their own experience, very, very different
29:43
from us. What would he say to young
29:46
people who might feel like they don't fit the mold, whatever
29:50
that may be? Believe in yourself, trust your,
29:56
trust your skills, trust your ability to change the world.
29:59
Like literally every one person, every single person in this
30:03
planet, they have some unique skills that we are born with
30:06
which you can use to change the world.
30:09
Like every single thing that you do in your everyday life that
30:12
can actually help you to create an impact that goes bigger that
30:16
that can actually speak, speak more than words, you know, like
30:20
actions speak louder than words. So you can actually write your
30:23
own story every single day. So that is what I would
30:26
encourage young people that you might be introvert, you might
30:29
not fit the expectations of the society or every single person
30:33
in the society. It doesn't really mean that is a
30:36
weakness in you. It is something that could be
30:39
the positive quality, a positive trait in you.
30:42
You just have to leverage that. Whatever is your weakness,
30:45
identify how you can turn that into your biggest superpower.
30:49
Fittest introversion maybe generally, maybe you are a good
30:54
writer. Maybe you you might be a good,
30:56
good thinker. So how you can become a
30:59
scientist, how you can potentially win the Nobel Prize,
31:01
right? Like have big dreams.
31:03
I think like I never I never imagined like even for one day
31:06
in my life that I will be in Forbes under 30, You know, so it
31:10
was like have big, big dreams and kind of never give up in
31:14
your life. That's the biggest encouragement
31:17
to young people. Who knows what you might become
31:20
one day, right? So just just keep come on going
31:23
and let your work speak for itself.
31:25
And it will reach the right people at the right time.
31:28
And then that's where you can speak up and you can share your
31:32
views with the right people. And you might be speaking to
31:36
some of the biggest people, biggest leaders in the planet,
31:40
the best leaders in the world. They are the ones who give space
31:43
to their subordinates, right? So they let their subordinates
31:47
or their colleagues speak first. And they are very humble.
31:50
They are very grounded. I think this is like a very,
31:53
very much a unique trait which could be turning you into a
31:57
leader as you grow in your life. For yourself, what kind of
32:02
legacy do you want to be able to leave behind?
32:06
What I want to demonstrate to young people is that you can
32:10
actually shine in the big world that we have around us by
32:14
actually staying as you are. What I try to do is kind of jump
32:19
in at the right moment where I feel it will help the company
32:23
the most. It will help maybe the society
32:25
the most for other people to see that.
32:28
OK, like coming from a shy child, you can actually reach
32:32
this position in life. You can actually be like Forbes
32:35
under 30. That's just the beginning.
32:37
So there, there's so much more that could be done.
32:39
If I can become ACEO or something someday of a big
32:42
company, like I think that would be something I would love to
32:46
accomplish. But yeah, let's see where life
32:48
takes us, right? Wonderful.
32:50
Well, I will have to re interview you in five years time
32:54
and see just how much further you have gone.
32:56
Thank you so much for sharing your thoughts George.
32:59
Let me finish by asking you our podcast staple, which is what is
33:03
one thing that you think will allow more people to have better
33:06
mental health? Because we talked a lot about AI
33:10
today, AI, I feel it could be useful.
33:13
It could be harmful as well when it comes to mental health.
33:16
So if you are just completely using AI for every single thing
33:21
in your life and you are losing human interaction, that's not
33:24
good. It could lead you to depression.
33:25
It could lead you to, you know, like just be doing everything
33:29
with AI. But where AI could be helpful is
33:34
it could help with your mental health as well.
33:36
So for example, imagine a lot of people, especially the older
33:39
people, they might leave alone in today's life, there is not
33:43
much social interaction between everyone.
33:45
Like everyone is busy in their everyday life and they don't get
33:49
get to spend that much time with their friends and family.
33:52
So that's where AI could be very helpful.
33:54
You have the companion at the fly of a button.
33:57
You just click on that button and then you have an AI that can
34:00
speak to you and it can just be your best friend.
34:05
It can be your philosopher. It can help you to act like your
34:09
grandchildren, for example, and it can give you that happiness
34:12
which you might not be experiencing right now because
34:15
you are away from your friend, friends and family.
34:19
In a nutshell, use AI in the ways which gives you happiness
34:23
rather than in the ways which makes life monotonous for you.
34:26
Stay confident, stay grounded and humble and keep on working
34:31
on life like keep keep on doing things that you'll have to do.
34:34
If you like, like to write papers, write that.
34:37
If you like to write books to do that.
34:38
If you are good in sports, then do so, but do something that can
34:43
kind of will give you a competitive advantage and that
34:45
will set you apart in society and be 50 years from now, people
34:49
will remember your name, you know, like that.
34:51
OK, This person was the contributor for so and so thing
34:55
in the society and he made this change in the society which
34:59
which the world would always remember and your name would be
35:02
carved in history. What a wonderful way to end our
35:06
conversation, George It thank you so much for taking the time
35:10
to share your personal journey, the comments that you received
35:14
as a child, which actually inspired you to keep working and
35:18
stay true to who you are. And so thank you for educating
35:21
us about some of the key terms within AI.
35:24
It was a pleasure to have you on the podcast.
35:27
Thanks a lot. Yeah, nice speaking to you as
35:29
well. Cheers.
35:30
Thank you. That's a wrap for today's
35:33
episode of the Low to Grow podcast.
35:36
If it resonated with you, leave a review and hit follow to help
35:40
more people to find important conversations.
35:43
Keep growing and see you next time.