opportunities.
Saved transcript
Smartest Way To Use AI: Powerful Words, Tools & Next Revolution | Sowmay Jain | FO439 Raj Shamani
Channel: Raj Shamani
So messaging app a billions of user base
is a very good opportunity WhatsApp or
telegram very human oriented with less
AI [music] tell me things are incredibly
fascinating around the world
there's one startup they charge people
after they die what they say okay you
died out of cancer right we will put you
in a refrigerator
we will make you I escaped
>> AI
there's like a big market in robotics in
China and even other places
I think where it's heading is a bit u
dangerous territory in terms of like
extinction event of humans very high
possibility
do you feel bias there's always a biases
I think in in AI because it's trained on
some data and then every data will have
some kind of biases you can't have like
a clean AI
>> what does a good prompt look like. So
depends how better you are conveying it
in your words.
One thing that works best when you
really ask AI to think hard, it does
think hard like go in depth, think hard,
use these words. So if you use these
words, it actually gives you better
answers. Yeah, just right answer.
[music]
Hey, so can you issue an invoice to
Rajamani for receiving 50,000 and
scheduleuler every 9:00 p.m. if you
haven't replied to the email? Yeah. PDF
and this is my invoice directly invoice
on the fly. I'm just talking to you on
the fly.
You're talking about Elon. You love him.
>> Yep.
>> His goals are very not normal. Going to
Mars
goal. If you are aiming for stars, you
end up being on the moon. So I think
that's a very big trait of him.
>> In your journey of building products and
actually attracting so many users, what
make people stick to product?
episode guest Jen founder of Upsurge
Labs Bindi AI or co-founder of
Instadrypto
investment future economy found
according to
Bitcoin
execute [snorts]
workflows
systemf
references
per week date fix. If you're curious to
see how AI can actually do things for
you, check out the link in the
description.
[music]
If somebody's watching this today, okay,
>> but AI has become a buzz word and now
[clears throat] it's reality also.
So all the software we are using today
is going to go
like very manual built for humans the
interfaces. So
apps even messaging app I think
messaging app is a very big opportunity
WhatsApp or telegram be very human
oriented with less AI. So something we
are also adding into bindi backroom. So
messaging app a billions of user base is
a very good opportunity.
Productivity apps you don't have to
work. I don't know why after an AI
people will work. People see it like
but I see it more positively. You really
don't have to work. You can just have
fun because
the only thing we need to involve like
we are connecting right now right that's
only thing as a human we need to do rest
all of the things is already handled
under the hood by some AI systems give
me an example like what
give me use cases use productivity
that person can make a lot of money.
>> One of the biggest problem right now is
context like
try to understand my I'm about to have a
loss in some of my financial assets at
let's say $2,000 and I talk to my
brother key price
>> so he will not take it as like uh answer
okay acknowledged he will be like oh
he will have some kind of emotions
to attached to that same stuff right so
AI
problem. I think context building anyone
can actually add it. The AI can
understand us more better.
GPD come back at me. That would be a
very big change. I think
is like
it will suggest you some places to go.
So now this is a
this is what lacking in the AI
but AI systems are not coming back at
us. Now this is a very good one like
they will talk to us and it reduces my
workload more because I'm not even
thinking to talk to AI now they they are
coming back and actually solving my
problem.
>> Suggestive
>> suggestive so instead of human to AI
it's AI to human and I'm just a
confirmation yes or no. So that's one
which you feel is incredible opportunity
productivity
sort of app
and next
that is one yeah that is one opportunity
right now
>> very big opportunity
>> okay um
>> what's the second second opportunity I
mean there's multiple industry and a lot
of opportunity but I will say if you can
make work invisible
work invisible like
all these thing needs to not be there.
So because all these are very logical,
very deterministic, AI should be able to
handle it. All they need to know is my
intent and they should just get the work
done. So if if there 10 people I'm
talking to, AI should know just this
skill. These two people are very
important. just take care of them
better, coordinate better and
systems.
I mean, internet as a whole is going to
be zombie land because
Obsidian Obsidian founder Obsidian is a
very popular notemaking app. that
founder was replying to the tweet done
by my bot and I read like oh that some
guy in San Francisco who leads this
project is replying to a uh
>> a bot
>> yeah and I and I was confused like okay
this is a very good example like key
slowly slowly how do I believe I should
need to coordinate with an AI how why do
I convince other person on internet so
the internet will just become a
consumption layer I will just go there
to acquire information but not to
converse because I don't know
So it's a and I had seen his firstand
already two months back
we are underestimating like how
dramatically the world is changing
but is there some way where people can
do it or there is no way which you can
think about
So I think every app that exist will
have a revamp.
So
simplify end users and just change it
and it's going to have value like
lot of apps we are just doing a lot of
manual work which doesn't exist and I
think one more thing is like um uh novel
tweeted this some months back that the
promise of AI is no interface like
that's when actually AI system. So
>> explain.
>> So
be I mean if you want to book a car, you
go to Uber. If you want to order food,
you go to Zamato.
>> Or you want to book a flight, you go to
make my trip.
>> Uh you don't need this division. You
need a one uh system which we call AI
system or AI assistant or whichever
whatever we call
>> whatever you call it something XYZ in
your in your phone. There's one app
called XYZ and then you just tell that
to the app
>> and it will take care of everything and
um let's go one more step ahead. It
doesn't even I think mobile will change
itself. So
design everything is concentrated
in the top of the iPhone
everything is concentrated in top layer.
They are
they are moving towards a miniature
version of devices. You will not need
like a book phone. Everything is
concentrated here. Could be attached to
here. Could be anything. I don't know.
>> So it will not even me using a phone. It
just stays with me and figuring out
things around me.
>> Yeah. That Jio launched that glasses.
>> Yeah, glasses are a very good example
like uh I'm fascinated.
>> Glasses, geo glasses, all that they've
launched.
>> Yeah, they haven't gotten to that stage
yet that I will use on a day-to-day
basis, but it have a real chance of
disruption. Yeah, they're using I saw
this video recently
building and that will give you history
of it. So you can use it as your let's
say in air inear assistant fulltime.
>> Totally. Yeah. I just tried it one week
back. The newest uh Meta glasses
>> and u it doesn't even have a voice. It
actually transfer
information transfer.
So I'm seeing like all these things
coming out like so we're dramatically
underestimating where we are heading. I
think it's going to be very cyborg kind
of world
>> and it's going to happen.
>> It's going to happen. If we don't do it
someone else will do it. So why not us?
>> So there okay again we're back to the
same question
opportunity.
>> I think the next good opportunity could
be robots.
>> What do you mean? [clears throat] M for
example [sighs] what if this is powered
by some AI system I'm moving here
>> yeah it comes to me immediately
>> but it's not only about that if I just
went down to let's say
it should not come to me now if it's
just a deterministic program it will do
that too but it needs to have some kind
of consciousness inside it
>> some kind of intelligence
>> yeah some kind of intelligence and
possible so it will move around with me
this is just a An example maybe a lamp
Pixar kind of lamp sitting around me.
>> So and these are a very good one because
uh uh this is a good opportunity. I
personally want to enter into this but
it's a very high turnaround. You can't
just move things around so fast robotics
right
>> but if dedicated time I will suggest to
get into the intersection of robotics
and AI because that's when the real
change will happen for us software
but when all of the things around us are
having some consciousness
>> via this artificial intelligence they
are moving around and
they're kind of coordinating with you in
some lights
>> by the
We also have a robotic team in like
Bangalore. The lamp actually reacts to
you. You say it like hey I had an
accident yesterday and uh it will get
depressed. It will show emotion via just
eyes. You you try to be angry it will
get angry too. It have ears moving like
this like a pat you know. So this um u
uh these are some examples like how how
world around us is going to change
software physically. But I think when it
comes to the physical form that's when
we will feel it a lot.
>> True. So that's one or examples. I like
this example.
>> Okay.
>> Give me more examples like this.
>> Intersection of
>> robotics and AI.
>> Uh yeah this is one but intersection of
bio and AI.
>> Explain.
>> Basically bioengineering your body. So
>> what do you mean by that?
>> So already crisper tech you can change
genes in you. M so
you want to change the color of your
eyes maybe I don't know if it's possible
when you're adult but when you're kid
you can change the color of ice this is
one like very real use cases and it's
very pos it's technically possible if
someone is disease readen like let's say
high chances they have some genes which
can cause cancer now you can edit those
genes via this crisper tech and AI is
actually advancing it recently I read
about it Like
so AI has created its own sequence of
genome genomic sequence um it's
complicated but okay in simple terms
so species also evolved like monkeys
apps and then a human there's like
evolution that happened right AI
species with a whole sequence of genes
and now this opens up a lot of like
you're kind of creating a new species
right you
So uh I mean the positive and negatives
are many but uh very big field because
AI up her aspect of your life is
automated by AI digitally physically
then the only thing that left is you how
you maintain your body and it also
linked to longevity now like
>> how you can live longer by fixing
um because dying is not natural you die
because one of your organs stop working
like most cases if it's not sudden death
you die because your one organs failed
and other organs like follow the root
right
>> m
>> now if you find a system bioengineering
your system where you can let that organ
sustain well basically you can't die
anywhere so uh again many people are
prediction that the person who's going
to live 5,000 years is already born
>> uh I don't know I mean maybe we will not
live long enough to see that [laughter]
>> but it's a very bold prediction
>> right
>> there's one startup which is very funny
Like they charge people after they die
what they say okay you died out of some
let's say cancer right we will put you
in a refrigerator
cell will not deplete now
we will make you alive
and I went to this um uh uh event in
Dubai. uh even Brian Johnson came to
that this was a NAS daily event and uh
there was the startup pitching and
people and I found like this is such a I
don't know it's a very good money-m
process like you are selling after you
die it doesn't matter right you don't
need money now like you're going to die
you don't need money why not just keep
pay them and put me in a rental cube
>> and just uh so there's lot
>> that's tech
>> yeah that's tech I mean that was crazy
death tech startups. That's
>> what are the more f Okay, let's let's do
this.
>> Tell me two or three more things that
that is incredibly fascinating around
the world
and you have seen
like death tech is very
>> I've never heard of this. I'm hearing it
for the first time.
>> I mean crazy crazy startups
can't immediately think maybe it will
come into the conversation. Okay.
>> Yeah.
>> All right.
Okay. Because you're talking about AI
and
AI will replace a lot of things,
right? Let's figure out
that's your startup
dayto-day life.
Can we do that?
>> Yeah, we can totally do that. I mean,
our moto is to like make work invisible.
>> Lovely.
simple language. What does your start
and it will respond you in some kind of
text.
>> It will not do that for you. What bindi
do is it execute that for you. So for
example JDP can you draft me an email?
It will draft you an email and then you
will copy paste in Gmail and send it.
What happens on bindi? You ask it to do
it. It will actually send the email via
Gmail account. Massive
prompting prompting skills needs to be
good because AI doesn't understand you
well. So
it will not the outcomes will be not
right but it will do it. M
>> so Gmail but it could be anything
um can you get the price of
I don't know HDFC bank and send that
price to my broker and then ask him what
would be the and then have a have a
check whenever he respond you back after
2 hours remind me back now it's a
chaining of multiple actions it will
first go to perplexity or Google to get
the price then ask your broker should I
buy or not and then set a reminder after
2 hours I need to check if that your
investment person has like gotten back
at you or not and when it get back and
you can so these are the workflows which
are not possible in tech with bindi you
can do all this it it actually execute
on your behalf
>> okay so let's let's make use cases by
use cases I love this I'm going to make
sure that by the end of this podcastes
right one of the first use cases says
let's see
right
how what is my I'm giving you a problem
so problem is
Mhm.
Cheapest flight.
Cheapest flight.
Cheapest platform
clear
to end travel
bookings
and
coupons are particularly
specifically
all of this end to end. How can I make a
workflow or a travel agent type flow?
>> Yeah, it's mostly words basically.
like you're adding words to word,
sentence to sentence as descriptive as
possible.
In fact, it can also move ahead and give
you an itinary when you go there.
It will plan out all based on your
prompt how well you are defining it. I
mean inputs are so better like
the only place it I think the current AI
systems in general fails is the
financial transactions because the
platforms doesn't open up their APIs to
make the payments that's also solving
but I think majority of the problem it
will actually build it. Uh the real key
here is how better you can talk to AI
how descriptive you are when you talk.
>> How do we do that? Let's say bindi. Can
you can we get a laptop and do it? Yeah,
I mean uh it's it's sim UX wise it's
similar like GPD. Can we open it right
now and write a prompt?
>> Should we do it? It'll be incredible.
People will love it. Yeah.
>> Okay. Can you get give me my expert
book? So I want to plan a travel. Let's
start with some work first.
>> Okay. You want to start with some work?
>> Yeah. Tell me what you want to do.
Okay, let's say
It's okay
every morning so that
okay can you check out my email inbox
and archive
all the
All the emails
with subject starting with let's say
invitations
>> h
>> invit these are like your calendar
invite
>> okay
>> you know so you probably don't want
those things in your inbox because it's
already you can see in calendar
>> okay
>> first thing it will just ask you to
connect the Gmail
>> so you double tap it connect and then it
will check your inbox and then it will
actually archive it. Simple. Done. But
it checked the email. There is no email.
What I can do? I can ask it to
send an email
to
hey, can you send a random email to this
email? Um uh in the subject start it
with title invitations. Uh and send send
it three emails to this email in the
screenshot. So it's quite smooth in that
terms. You can just vocally talk and it
will do. So every task
I'm just talking to it. I'm not opening
Gmail. I'm not opening uh our admin
panel to add your subscription.
>> Everything is all automated
[clears throat]
and
sent. So try again.
So this system at okay it found your the
email that I sent it send
>> and it will archive. So now this is a
process where you can it will ask
because it's a but you can also go yolo
mode and it will actually not ask on the
fly. Okay archived.
Now if you go to the Gmail,
>> it had already you received an email.
You will see a blank inbox but uh you
received an email in archive too.
>> Yeah.
>> Done.
>> Nice. Okay. So irrelevant archive. Yeah.
I mean depends invitations. My calendar
is full and I receive like 40 emails a
day. So I was like okay archive
everything. I don't want to see those
emails. I will check my calendar
directly. You can say anything with
marketing anything. So this is my
workflow but uh you can have your own
workflow.
>> Yeah.
>> So AI can work as a filtering here
system here.
>> Okay. Give me another use case workflow.
May
>> I want to travel to Dubai
>> and
my budget is
one lakh
rupees.
Can you make sure to plan an eating
for the same with flights and
hotels included
also suggest some cafes nearby hotel. I
love downtown.
>> Mhm. and um
go in depth and now one thing that works
best I mean it's sounds very funny but
when you really ask AI to think hard it
does think hard like go in depth think
hard use these words think hard and find
the go extra length you know
>> so if you use these words it actually
gives you better answers
>> yeah it gives better answer which is um
which is I mean I think that's how
They're trained. They use these words to
actually expedite their um like go extra
length to actually do the research.
Don't stop. I do a lot of these things
when
>> and this one is built on tragedies like
what's so we use a multiple system. Um
um
>> what is LLM?
>> So set for some cloud set for for some
element and chat GDP for background
tasks. Okay. So once the chat happens
there's a lot of background processing
happens. So cloud and GPD use
>> it would do something. My only worry is
like um uh the financial transactions
are not that um
>> proper. So I read everything Google
flights, Google hotels, Google maps,
Gemini prosequential thinking for
thinking hard perplexity.
>> So so it's showing me scan.
>> Now you can just enter to confirm or you
can just switch on the yolo mode from
it. I'll ask you next. So it's thinking
hard now.
>> Okay. City Dubai from Delhi outbound
>> return date.
>> Yeah. One of the fascinating things when
it does don't do the job right. You just
actually go anger angry like hey you are
so dumb. Why didn't you ask me the dates
I'm going to go? And it it actually next
time it think more better. I mean the
navigation of the thinking gets better.
Oh, so
now this is the one-off. You can even
ask it like give more prompt like hey
don't only search in one direction maybe
also try other places it's also
searching maps for the cafes because I
mentioned like you can you also search
for the cafes
>> but why is it asking me to confirm again
again
>> yeah I think you need to check something
maybe you this works when you start the
chat I did it I switched it in between
right so when you start the chat uh you
need to switch it otherwise the chat
only goes into normal mode
>> got it with a newer prompt then it'll be
yellow
>> yeah it will be yellow so you need to
set it up at the start
but it's um quite quite nice and now we
can see the system at scale when it's
system at scale you don't why I even
need to check all this
fascinating
but at the end I will just do it and I
will wait for mobile when you put it and
you forget it
this is the best flight at this time and
like everything so
like you don't
>> let's say forget this and let's say a
small business owner wants to create
bills
okay here's a here's a here's a problem
Mhm.
XYZ
PDF.
>> Can we do this?
>> Totally. Very easy.
Hey, so can you issue an invoice to Raj
Shamani uh and reply to the Rabau email
for receiving
50,000 rupees and set up a uh
scheduleuler every 9:00 p.m. uh if the
if if if the if you haven't replied to
the email
um and um stop the scheduleuler once he
replied. So keep reminding him like this
this is the invoice and is to be paid
and send an email to uh the above email
that that you had sent the random emails
to issue the invoice make it fancy um
make it clear uh if you are confused
stop ask me don't make any dumb
decisions [snorts]
>> okay
let's see
>> it's going to do it's I'm totally like
it's I think this is one of the easiest
I was still confused about this because
like five apps But it still did it
>> which is quite nice.
So it asking me question because I ask
you to actually you can take your call.
You know me better. Okay.
Okay. It added an invoice gen agent.
>> Creating an invoice
and then it will just uh simply send it.
So another best part when there is an
error
it actually changes the trajectory
let's try this now
>> now and this is good because that's how
human operate they make mistake and then
they get better.
>> So um this is one of the example it
actually did the job but there's a error
beforehand. It also set up a
scheduleuler every 9 p.m. It's going to
send you an email reply to that test
email and remind you every day
>> and send it to that email that I
mentioned before. Let's see if you
received the email.
Okay, received.
Done.
It doesn't have the PDF attachment, but
hey, did you send the PDF attachment? I
don't think you have.
But done.
We have it.
>> Yeah. Done.
>> So, yeah. PDF.
And this is my invoice.
>> Yeah.
>> I mean, I can ask you to do more stuff
at the fast.
>> Yeah. But this is like the easiest
>> easiest fast and quick. I've gotten it
directly invoice just on the phone
>> on the fly. I'm just talking to you and
I can on the fly dump the voice. So this
this systems at scale is going to make
the whole layer of business and work
work invisible. In other words, you
don't need to work now. There would be
system which are listening you and me
all the time. And I said okay invoice it
already did. [laughter]
>> Isn't it scary?
>> It's scary but it's a new world. We got
to uh that's why I'm saying we
underestimating the way the world is
changing in a complete like it's
dramatically very changing and uh it's
going to be way different than how
internet changed the world.
[clears throat] I think one of the
biggest use case of this industry is
going to be dating.
>> Okay. How tell me AI dating.
>> So now
you don't basically people are swipe
fatigue like they have to swipe a lot
right?
Now this AI systems can you can find you
one date per week
>> and you can go on that per week date and
you haven't even went through the
filtering process. AI
they know you and the AI systems are
talking to other AI systems and they
figured out okay this could be the right
match and they matched us and we went
out. So this is a big case and and it
can make sure you don't need to like
swipe hundreds of profile. You can just
have one date per week and just go and
figure out and give me feedback. Based
on the feedback the system will evolve
more and then it can find much better.
So this is a very big industry I think
and it's it's it's very very very
like very visible this is going to
happen. Meta did launch some like a
pilot app around this this week.
>> So something's going to change. So
dating so somebody who knows you will be
able to find someone
>> will be exactly compatible to you.
>> Totally. So it's going to optimize that
process.
>> But isn't that dating that you usually
end up falling for someone who's not
compatible?
>> I mean ideally that's how it should be.
But I think the world is a bit different
[laughter]
[gasps]
>> or maybe when you go for compatible date
then you tell your AI
>> key bro this didn't work because of this
this I didn't find it exciting. find me
something exciting and then
>> and this matches you extremely opposite.
>> Yeah. And next week it will evolve and
you get something. So it's going to
optimize that process.
>> Um and it's a I think it's a very big
industry. Do you think
because artificial intimacy will become
so high possibility? High possibility
because um
um
because humans are I mean people want
convenience
>> and uh when this AI systems understands
you so better
then they will make sure you stay happy
or you are satisfied properly.
>> Yeah. and with robotics and there's like
a big market in robotics around this
like building humanoids and a lot of
things in China and even other places
and with consciousness in that it's I
think where it's heading is a also a bit
um dangerous territory in terms of like
extinction event of humans very high
possibility I mean people are worried
about this that it can extend human
because it will dilute the human
connection as well you are mostly
talking to AI persona who understands
you best
>> don't you think then the biggest
businesses will be built community
ultimately depends how what we want. I
mean there will be a good fraction of
people who want this
>> like who want to um be more connected
with humans but a big big population is
also like like [laughter]
they don't care they don't care so um um
but yeah I mean biology have its way to
like smoke out the people who are not
contributing to the species so the
people who are actually more connected
end up reproducing more and the people
who are art will end their legacy. So
biology have its way to like
>> like fine-tune anything. So
>> fair. What are the other use cases? AI
disruption dating you said entrepreneurs
AI use case I mean on the work and it's
mostly automation like there things are
automated um I think only two kinds of
job as a human will be left one is you
are curious about something
>> so you think and instruct your AI
systems
>> or you are creative so you ask AI to
create some kind of image some kind of
movie for
or some kind of aesthetically looking
better stuff. So we as a human going to
do just two things and I don't think
we're going to work. So
>> and then who's going to consume these
things? Us humans. But then
like why do you want to create when you
know
but um
I mean when I say create it's not
frame by frame I'm just giving it an
intent key I'm feeling very excited
today show me something related I'm
feeling very depressed show me something
motivating so intent
So, it's going to be
>> But it's going to be AI generated,
right?
>> It's But you can't tell the difference
between AI and human. I'm okay with
Okay. How will people earn money? How
will people connect with each other?
>> That's another topic. But I think uh
very good topic. No, very good topic in
a way. Like I also think the the concept
money as a concept might also not exist.
>> Um Okay. So
>> no no I'm truly so the only limiting
factor for AI is energy. We have enough
solar energy to actually power all of
the AI and when everything is automated
around us government will have to come
up with some kind of universal like
energy allocation
>> and AI system running for automating my
stuff. So now think about it like
you will plan out a two months or 3
months and four months. Now think of in
a world 20 years down the line you are
talking to your AI systems there are
robots flying down with woods and stuff
and stuff and within a week or within
few days they can actually make the
whole house you think it's very ex
exaggerated very possible
north
because travel time is very
So the time it took to travel reduced
the time it for us to come go from here
to the other part of the world is
reduced
everything is so getting advanced travel
time building time or time or the cost
is getting optimized we'll get to a
point
system even in physical world it will
make it drones stuff humano roids
properly creating a house that I want
for me physical change not just digital
>> so up when you're entering into this
kind of world the trajectory is going to
be very different like the base level
needs are already done you're going to
be eating enough food because government
systems are like I mean government needs
to be good though I'm assuming the
government is like very good
>> just allocate some energy solar energy
nuclear energy could be any energy but
they will just allocate that energy and
everyone have base seek necessity
fulfilled who I want to go extra it
could be more like space travel Mars
that will come into an elite behavior or
status quo behavior
very possible I mean looks not but if
you see the last 50 years the prime
minister of India doesn't have the first
prime minister doesn't have the comfort
of switching on a but uh switching on a
switch to start the heater there's a
process in heating water and like today
we just wake up switch on and just go
into we have the moving air condition
which is cars
200 years back people don't have it I
mean we don't see the change that
happened because we world born in a new
world the change that's going to happen
is way different everyone is going to
live a billionaire's life that's for
sure and the what billionaires doing
today is going to do more extra official
thing like
>> space travel
>> space travel or I don't know building a
tower in one week or
could be different. So we heading
towards that world and growth is
exponential.
>> So
world new world replaced because a lot
of jobs will get replaced. A lot of
people will get replaced. A lot of
skills would get replaced. I think we
need to think I we need to think it
differently. We at uh we are not meant
to work. So replacement comes
we as a human are not meant to walk
this system at scale you don't need
money [laughter] but okay you will need
some kind of
really clear.
So I think immediate solution to that is
just adapt to AI. I mean uh one of the
most important skills is how well you
can prompt the AI.
AI
I had seen we had seen like a lot of
users come and give a oneliner stuff and
say it doesn't work. That's the most
important skill as of today you need if
you're in college if you're doing a job
start to learn how to talk to an AI
there's also a technical term to it
prompt engineering you need to prompt it
well in order to make it work. Whoever
does this best is a so in our company we
make sure that two two skills are uh um
really uh there before we hire how well
you can talk to AI prompting and high
urgency you need to have that kind of
urgency to do something I think these
are the two skills you need to operate
in the new world to be valuable in the
new world high urgency and uh good
prompting skills
>> okay what does a good prompt
look like
>> yeah it defines things uh very well I
And again very personal to everyone how
they talk but or what kind of response
they need but the [clears throat] first
paragraph needs to to
for look for any place in
anywhere else but
depends
how better you are conveying it in your
words.
So prompt engineering means first is
okay
I mean it's a process like first is you
define
dos then don'ts and then you enter
that's a good is that a good prompt or
is there more to it
>> yeah this is a very good foundationally
like so you need to say what you need to
get done and the negatives and positives
of it what like what
what not to include, what to include,
what to include and um
do you feel
like
I mean that's that's going to happen
because uh for this corporates main
so for that they want that they make
sure you feel good just like how social
media works. So
>> they want it to happen. So I think
systems are going to evolve to that. And
also every every uh AI system have their
own biases. So you know uh Elon tweeted
about it like
um if we make AI can be anyone. If we
fine-tune it better on the right wing,
it become resist. Like if we if we fine
tune the AI on the type of content which
is anti-liberal it become resist
>> and if not it become completely liberal.
>> So you can't actually build a nonbias
AI. So there was this meme
Hitler. I'm not sure if you heard on
Twitter it was happening but uh the
Grock turned into Mecca Hitler and
started like talking like that. So
there's always a biases I think in uh in
AI. Uh, it's going to stay I mean
because it's complet it's trained on
some data and the data will every data
will have some kind of biases. You can't
have like a clean AI.
>> Got it. You're talking about Elon.
>> Yep.
>> You love him.
>> I I mean I haven't met him personally
but I I really appreciate him.
>> Why Why do you think Elon
I think um I mean there's a lot to it
but um I think the
the major part like
you need to obsessed over something for
a long span of time.
>> So
many people fail on it. I mean and his
goals are very like not normal going to
Mars. I mean anyone who want to live a
normal life he can just stay happy on
earth right but as a bad goal like you
want to like achieve this I mean those
kind of traits are if you are aiming for
stars you end up being on the moon so I
think that's a very big trait of him
like which that's why he's able to
handle like five different startup all
in its own like doing the own innovation
neurolink boring company and um SpaceX
and even in the SpaceX they have cluster
of team working on different defenses,
Starlink. So, I like I like how how he's
able to handle all of them while while
not burned out. So, um I admire that and
I think the most important part is
you know what drives people are like the
hope right I mean they need to they need
to believe like this can be possible. So
he had made the bar so high as a single
human you can do I can do this
>> anyone can do anything. So that opens up
the wide range of opportunity. You will
not start doubting. You be like okay if
someone did it we can do it. So the
indirect butterfly effects of that act
is crazy. I think that's why I think
many people admire him and I think it's
a it's a net positive like it's going to
really add some value to the world.
>> Fair
entrepreneurial journey.
>> I think the first time founder is just
figuring out stuff like how to do it and
um I was very much interested. So so
Kota Rajasthan.
>> Kota Rajasthan by the way. Tell me about
your family.
>> Yeah. Um so
I mean I was born there but
uh keep changing cities jobs and stuff.
So I kind of stayed in everywhere like
um from Gujarat to Mumbai
um Hyderabad
um and even in UP I was 5 years I was in
a guru school in Manglatan Gurukul.
>> Yeah.
>> Okay.
So
>> what do you mean by this? So you stayed
in Gurukul but you studied CBSC and in
your extra time
proper scriptural studies and scriptures
and u rituals you're doing it.
>> Okay explain.
So this is fascinating.
>> Yeah this is
yeah and the and the most different part
is there's no internet like internet
connected.
in 8th standard my mom like sent me to
the gurugul and um from 8 to 12 I was
there host
>> you can call it hostel but uh it's a bit
different in a way like it's just not
hostile it has some kind of uh different
>> explain
like early in the morning
to before school I'm talking about
before school
puja part and main the main uh focus was
um suadhya. So swadhya is like you're
reading all these scriptures.
>> So I'm a big fan of like few scriptures
has shaped my thinking process is like
moka praashak samsar and and we used to
read it every day for an hour like yeah
school like
we do our breakfast and then go to
school and when there's no holiday in
school we spend extra time in temple. So
I mean at that point of time we like ah
why have you do it
>> but later we figure out that was a very
good like uh it actually shaped a person
out of me like the discipline like I I I
don't think if
connected like in this whole peak of 8th
to 12th I would have had that kind of
mindset or sharp mind that is right now
regular school like from 7 8 So day uh
uh one and two
>> and then you come back and some some
special days you have to again go back
and normal days you can spend it your
time as you want. We spent it by reading
books because that was the only way we
can uh get information uh on internet.
>> Yeah. Internet. I mean they didn't give
the access to the internet. So
but it was also very regulated. So um
then evening we food before sunset we
don't eat after it.
>> Okay.
>> So you have to stay. So now current
lifestyle is so different than what you
don't have to eat uh and you don't get
food after that. So basically you are
like all
>> and then you go to temple again spend
that 2 three hours this was a normal day
>> every day
>> every day
>> and 2 three hours temples again
scriptures
>> yes scriptures and if someone want to
spend extra time there's also extra
classes so
>> and what do they teach you apart from
just books and scriptures
focus lessons from the ancient or from
our gurus or from yes
>> which is the ancient wisdom they would
try to teach you that
>> yes totally that so um spirituality I
think is the main part they wanted to uh
like make us read. So I think that was a
very interesting part of the life.
>> Um and um for guru kebab
>> guru kulkad then I wanted to um and now
it's on your own like you get that um
whatever skill set you want or mindset
you want and now you can operate on your
own. So I wanted to get into fintech.
>> Um actually finance I was a big fan of
Warren Buffett. Even at Google days I
used to read like of Warren Buffett and
a lot of finance book um the annual
Burkshire Hathway newsletters letters
and stuff
>> shareholder letter.
>> Yeah shareholders letter. So so I wanted
to get into investment and finance. So
>> I was 17 and I started writing on
internet ka kora I have like a good
followers. uh people started following
me at a 17 and I was like okay
[laughter]
like taking my advices on stock market
and stuff um and I didn't have any of my
DMAT account because I was not legally
then I turned 18 that's when I started
doing and whatever pocket money I had I
started investing all in like um and but
I had the high energy I want to do
something like I want to build something
so I went into like building some apps
fintech apps and uh this is a very
interesting story but that's what pushed
me to the crypto Let me tell you.
>> So,
>> so I started building apps and figured
if you want to be successful in finance
or fintech industry, it's highly
regulated.
>> You can't do it immediately.
>> You need to go through an big approval.
There needs to be a credibility in you
like you had worked in 5 years 10 years
in this sector.
>> Uh but time and this is also termed as
high urgency but I want to do it a very
immediate like
>> yeah right now.
>> Yeah. Right now like I can't wait for I
can't plan out for next 5 years and 10
years. Um and uh then 2017 the bitcoin
was going up
>> and uh it was getting some good traction
and I thought and so I built an app
which was making like $1,000 a month.
>> Sweet.
>> Yeah. Just a uh just a app very
notification service of BC
notifications.
Um then overnight I switched from this
to crypto just because I can code and I
can deploy contracts on Ethereum
blockchain. I enter with Bitcoin via
investment but I can write my code and
put it on Ethereum blockchain. So
blockchain
deploy like open blockchain and there
are people who are actually depositing
millions of dollars
>> in that contract. They don't even know
me. They don't know my age. They don't
know where I live. And yet they are
using the software because smart
contract like smart contract is like
it's a open blockchain. You can see the
code
>> there's no like you can verify it.
>> Yeah.
>> And you don't need to rely on me to use
that. And uh that was a very defining
moment like I still used I remember I
was at home and I was saying mommy
[laughter]
and my mom used to think is he doing
scam
um um and uh that's how it uh entered
and u I think uh we started getting a
lot of global recognization because we
are early in our career and they're like
they wanted a story.
>> Okay, two kids somewhere from India
build something used by big big players
globally. So we got a lot of like media
coverage in crypto community at that
time and that's how we started the whole
journey and I think six seven years I
all in there building the
>> what was the first investment first
funding first investment
>> oh yeah it's also very fascinating I
mean um um so we had a good investor
base in the first round from Nal Raiki
Kant Balajinasan Coinbase Idea Collab
and a few more good players so
Or how did you convince these guys?
First round me
>> first round me I mean I think um the
best part this is also very interesting
like for any founders want to raise fund
I think
>> you don't have to raise fund in a way
like you want to raise fund I mean
like you need to work towards it u
what's rare in the world is the best
good founder there's a rarity of good
founders there lots of founders but the
good founder which can define things are
rare
>> and what's you tell me what's the
investor's job. What's their job?
>> To find hunt for talent, huh? Like they
want
>> a rare talent.
>> Yeah. Rare talent. So that's the
investor job. They want they it's their
job. It's not your job.
>> You do your stuff, build the best thing
and you they will find their way to
approach you and actually fund you. And
that's what happened.
>> And they found you up front.
>> I mean uh there was communication going
on but we didn't did anything. We didn't
have a deck. Even this round we raised
it on a tweet. We didn't have a deck or
a plan or anything you know. So um um
they are there to find the right founder
which can change the future. So you need
to be that you need to be that person.
>> So how did that happen? Tell me it's
fascinating story
or
no. [gasps] So first we get the lead
founder I mean lead investor. It
doesn't. So all of the other investors
are the follow on investors. They don't
directly come in unless you're raising a
friend friendly angel round which is
small
>> but we um so we got the lead which was
Panta.
>> Okay.
>> Um they are a big crypto I mean they are
a big crypto fund in US and uh
>> yeah I mean we connected to them via the
mentor we had and the mentor I mean the
mentor in a way like he wanted to
acquire us.
>> Okay.
>> And we rejected it. It was a good offer.
we can go to LA and we can work in their
office in LA and stuff but we like um
not too interesting. We did even had a
uh so he end up becoming our adviser.
>> Uh like okay we and here's uh I think
Edward Monata he's like a CEO of
Blockfolio and he got acquired by FTX
later. So another story
>> um but uh
>> um um from there uh we rejected that
offer and actually we did a rejection a
lot. I mean we did rejected a lot of
offers.
>> Okay.
>> On the way while we don't even have a CR
in our bank.
>> Yeah. [laughter]
>> We're just kids. So we went for uh we
went to SF and there was this um uh one
guy I think Tom his very lean position
at Dragonfly Fund. He used to work
somewhere else before and he told us you
did a very huge run before you completed
this round like you should have raised
earlier and um we were just meeting Noel
after that. And uh there was this
statement by this guy like um uh you
should you have should have not did such
a long sprint because we are so popular
zero fund raise.
>> Yeah.
>> So um um and then uh it's more of
network and connection. We went to an uh
New York event and um just shaked hand
with Paul who is like the lead in uh
who's leading the panther fund. Just a
two-minute conversation, nothing else.
We came back. Coinbase the biggest
exchange wanted to lead our first round
and then other investors came and like
okay uh what about we do it
>> and there was a race between those
investors like who can fund first
[laughter]
>> race versus Panta
>> yeah I mean it was not so visible but
the point is once you get one investors
u so [sighs and gasps]
oh okay why do investor invest
I mean most of the time investor invest
because other investors are investing
simple so if you can build that vype or
vibe uh kind of a mystery around
yourself. Uh raising ground is not a big
issue. I mean you do need a substance or
a credibility from your history. Um but
but once you build that hype uh there is
a risk capital anyone is ready to risk
like even this is like a zero a 1%
chance let's allocate some funds.
>> So uh it converted into a completely new
kind of um um um um uh a group of people
wanted to like just be a part of it. Uh
and we also fortunate at the back desk
like I mean they were the best people we
can work uh I still remember walking
alongside novel in SF street
>> and then about novel meeting
>> so um
>> who introduced you to novel
>> okay that's also a good connection so
panther invested and Coinbase was
already wanted to invest everything was
friendly there's no nothing aggressive
um and then at that time Balaji was the
CTO of Coinbase if you remember
>> so um that's how we got in touch with
Balaji Coinbase did participate in the
round as a follow on so they were also
part of our they also backed us so quite
appreciate that but that's how I got in
touch we got in touch with Balaji and
from Balaji we got in touch with novel
and um um uh uh got chance to walk with
him in the SF street we were very young
with our bags you know and uh on the way
back we give him pa like the milk pa
because you know desert and we were like
first time going out Um and uh it was a
good uh good uh good. So I think what I
appreciate
>> what did he ask you?
>> I asked our vision and mission what we
are trying to do and stuff and some
basic questions but I don't think he
went too much to technical. I think he
was backing the founders compared to
getting to know what the project.
>> Yeah. So what did he find in you? What
do you think?
>> Uh I mean I think we need to ask Novel.
>> What do [laughter] you think? What do
you think he backed?
>> I think we are very early in our career
and uh we took the risk. Nobody at our
age like I was 20, my brother was 18
actually think of doing all these
things. So might have seen we are very
self-driven. High urgency I think is
again a very critical skill everyone
need.
>> So you just went to him and be like this
is what we're building. This is what
we're doing.
>> Coinbase and Pantherra
>> we have active offers from both of them.
>> I think these are more followons. So
they like uh lead already already
already gets confirmed and
>> the lead is Panthetherra, right?
>> Yeah.
>> So it got confirmed. So you went to
Novel and be like my lead is confirmed
already.
>> Panther is investing this much in us.
How much do you want to invest or is it
more like hey we want advice to build
this?
>> Yes. I mean it's more like a follow on
then like uh we are raising around but
we also want you now the followons with
it doesn't work like they want to
invest. We like okay we need you. You
can be very helpful to us. You have a
good network. So with followance is more
like how we can because uh the the
maximum amount of funds you get is from
the lead. So you solve the money
problem. Now you need to solve the
network or credibility problem. So
that's where the following comes. That's
when you make sure your angels who are
putting the small checks uh are very
well curated.
>> Baji is big in crypto community. Nal is
big in that community.
>> People like these
product building to it's again founder I
don't I mean we can we can't expect
investor to do it because if they're
doing it then what we are doing it so
>> I think it's a first I think their
backing in itself is a very big thing
they believed in us that's I think
majority part is like solved like okay
there's someone who is believing in us
who who had built something influential
before so that's the m and uh another
benefit is just networks
You can just DM them anytime and you can
just like hey can you connect with us
this we're raising another round can you
connect with us on other investors. So
it's it's on you how you are actually
leveraging their network and how well
you are leaging.
>> Did you leverage?
>> Uh yeah I mean I did text texted here
and there anytime I needed some help.
>> Help.
>> Yeah it did help. I'm meeting Balaji
next uh month October in network school.
There's an event
>> Singapore. So I'm I mean we kind of
connected and uh can text anyone and if
it sounds appealing they will actually
um
>> do they help? Does it has it ever helped
you till now?
>> Yeah it did help. It did help. I think
uh angels help the best like um um I
mean first I'm the backd is one and the
networks they have they're wellconnected
people. I mean if you can tell them hey
we need to connect with them and this is
what we want they will. I mean that's a
very big value prop because getting to
know some third person directly but
someone is referring you. So let's call
it like biology introduces to novel.
Novel is not never directly investing in
us but now there's a referral. This is
the first case. So I mean if you
articulate it like it they it does help
the network does help not too visible
but it does help.
>> So at some stage some scale it will help
eventually.
>> Yeah it will help. I think uh one more
thing you need to be very grounded.
uh you raise your first and you're like
okay I'm g I can do something but you
never know how life will turn around
like uh you don't need a uh who you
connected today how they're going to be
helpful down the line or how you can be
helpful them it could be anything but
you need to be very grounded um if you
really want to be so um yeah even at
earlier stage that was also like okay
then we got some roller coaster rides in
the start we're like okay [laughter]
we got to be but uh yeah it's very very
important very important.
>> Tell me one failure you faced.
>> I I don't think I see failure as a
failure. I because
>> but one setback that you had
>> setback I mean I did had a lot of
problems which we see as a challenges
and we need to overcome it.
>> Um
>> what was the biggest challenge?
>> At different stage it was very different
like I mean it depends at early stage of
my education it was like I need to study
or I need to make something. I failed
everything like academically I didn't
pursue I was doing CA I didn't went for
second level I was doing CFA failed
level one in fact I even failed uh BCOM
[snorts] so I was very like uh
academically drained like so but I was
very passionate about other side of
things so it pays off later on very
stressful period but possible
>> and while building the startup finding
our users another very big outcome like
the How much brand visibility you get or
whatever you get like you're known by
millions of people or you raise the fund
raise round doesn't matter the ultimate
test is how many users you can get how
many people you are changing life for in
other ways if you say are people using
your product if not all these are like
very temporary maybe year or two you
will die down one of the ultimate stress
test for a startup is how well you can
get the users how sticky they are how
many lives you're changing in that so I
think with The first startup we did
for a subsequent like for a few years
figuring out was a very big problem like
figuring out like how to like actually
game up. We launching something it gets
popular for a week or two again the
viral viralities r goes down. I I'm sure
you might also have some many moments in
your life too. So you can relate. Um but
a very very big I think it's things are
very stable right now but initially when
we starting out very tricky. What makes
in your journey of building products and
actually attracting so many users,
[snorts] what have you learned? What
makes people stick to product?
Because there is there are so many
products and whatever you're making
>> maybe starting you were innovative
>> but in 2 months there'll be like 500
meters of you
>> right. How do you make people stick to
your platform? What do users really
value?
I think the uh the initial growth of any
startups comes from users directly
connected like you need to listen them
you need to incorporate those feedback
quickly in your product and make them
happy. So initial set of 100 to thousand
users how happy they are is going to
decide the next trajectory of your whole
whole company.
So um we did a lot of those things. Um I
think one of the major feature we we
were like very viral about on Twitter
was refinance. I mean
build a system from one protocol to
another one button click. So very very I
mean small thing nobody will do it but
there was this 100 to thousand user who
wanted this
>> we build it and uh we listen them and we
incorporate that in the product I think
that's what matters the most got it well
is there any one lesson last lesson that
you want to share it with all the
budding entrepreneurs
that what should they keep in mind if
they want to build something and which
category they should be building and
thinking in today's world
>> there's a lot that can come into the
mind um but whatever industry you guys
are building like whoever is listening
this um I think one of the most
important thing you need to do is stay
highly urgent about whatever you're
doing like high urgency is a very
important skills I mean it looks very
cliche again but you need to be obsess
about the problem. You need to have that
kind of urgency. Uh because that's the
only skill that is going to make you
stand out. Uh looks very normal but I
can tell it based on where I am right
now. The AI industry, how everything is
panning out. The only skill set is going
to make sense is how what level of
urgency you have in in in in within
yourself. It's going to it's going to
make the whole that's going to
differentiate the winners from the
losers. Let's say that way. And in what
industry do you feel the next billiond
dollar company can be built?
uh I think
all of the industry there's a reason
we are going through a paradigm shift
you know from from industrialization
to hardware we got a new wave of
billionaires Microsoft Apple from
hardware to software we got a new wave
of billionaires Snapchat Facebook all
these cloud internet software
>> internet companies. Now we are moving
towards intelligence age like now
internet and human. This is the current
form. Now it's going to be internet
intelligence layer and then human. So
we're going to see a new wave of
products and companies popping out. So I
think it's haven't incorporated well in
our life. So wherever whatever interests
you just try to think how this new
species artificial species that we are
building intelligence can incorporate in
your lifestyle and the next wave of
billionaires are going to be from this
intelligence age. So what is a good way
to start learning or maybe thinking
about how can I incorporate intelligence
in my world or start thinking about how
to build products and intelligence
because someone listening this
>> for you it might be very obvious and
easy to just say that okay you're going
for a paradigm shift and this is going
to we're going from internet to
intelligence and whatever right
>> but for a lot of people it's it's not as
easy as as it comes to you right
>> what should they be looking forward to,
what should they be reading, what should
they be uh maybe
surfing on internet or pages to follow
or the news and articles to read or
maybe just what is one prompt they
should put in maybe chat GPD or claude
that they start thinking from that
framework of building something big.
>> I think the start from basics of
intelligence.
>> Yeah. uh um uh yeah at some point like I
think at this point everyone needs to
somehow get exposed to this uh new
systems intelligence.
One of the important thing as I
mentioned earlier too I think uh is how
well you can talk to the system. So uh
personally uh I I can say my best model
like I my best model is some like I can
mention it like sonet 4 I mean it's bit
technical but you need to have that kind
of tuning with your own models like
which models you vibe with well so um
I don't know the immediate step but
whatever you're doing you need to wire
your mind to just talk to this like you
need to talk to this just like how we
were doing this bindi stuff and I was
wired to like say whatever
Everyone needs to have that kind of
wiring. The software can be anything
chat GDP or cloud to start with but
ultimately you will find your own model.
This reminds me of a meme like you know
um there's a meme going on like uh the
future models are going to be these
models. [laughter]
>> We're going to date all these
algorithms.
>> Yeah.
>> So um u like u yeah. So fascinating
world. So just start thinking about how
well can you talk to your software.
>> Yes. How well you can
>> and how well they understand you. So all
the problems and the thinking and the
next step that you want to take
>> you have that conversation with it and
>> let that language model search entire
internet come up with solutions and
things for you
>> and on that framework you start building
or creating the next level of things
that you want to do.
>> Is that what it
>> I mean it's a it's a good start. Yeah.
Okay.
>> And the rest you will figure it out on
the way. But um yeah, that's a very good
start.
>> That's a good start. Perfect. Thank you
so much, brother. Thank you so much for
coming here and spending time with me.
>> Yeah, very interesting conversation.
>> Thank you.
>> Thanks for having me.
>> I speak to a lot of founders
>> and I love this Elon Musk analogy
because a lot of founders quote this
>> because as a founder you have every time
you see an opportunity, you want to do
something about it.
>> Yes.
>> Right. And because you're like, "Oh, I
don't want to lose this. I know that
this company is going to grow and I've
put in a good team to let that grow on
autopilot. I can handle another project
as well.
>> Okay. But
>> so that's also a trap. But also depends
again.
>> Exactly. That's a trap. And why is it
also a trap? A lot of people keep
quoting Elon on this. But
>> Elon did [music] that after unlocking
billions of dollars in value for
shareholders. So shareholders started
trusting him with multiple projects.
>> Yeah. Exactly.
>> Thank you so much.
Number
two.
and number three episode.
One conversation can change someone's
life. I'll see you next time. Until
then, keep figuring out [music]
[music]
[music]