Hype Pod - The Hype Network Podcast

Arun: Welcome to episode
50 of the High podcast.

This is a special, special episode.

This is 50.

This is 50.

Wow.

Wow.

Let's, we didn't, we made the 50,
I dunno what the stats are, but

50 I feel like is half a century.

I like that.

You know, we're just, we're rocking
and rolling, so I'm really excited.

We got myself, Arun.

We got Vance here today and we
got special guests, Anto Patrick's

here, and so I'm really excited.

A AI Anto.

Ai Anto.

Is that your nickname?

Is that what we're gonna call you today?

Well, we got a guru.

We talk about AI a lot
on this podcast, but.

You know, we really dunno
what we're talking about.

So it's good to have an industry expert
here to tell us a little bit about it.

Um, you have a really,
really cool background.

I just gotta read off some of this 'cause
I feel like our listeners need to know.

We're, we'll let you do a spiel,
but you are the founder, the current

founder, and CEO of Cosmic Brain ai.

You worked at Xai, one of
our favorite platforms.

Um, you worked at nasa, you worked
for the F1 Aston Martin team.

Um.

And you're doing AI for all of
them, if I'm not mistaken, right?

You were doing some form one way
or the other, one way or the other.

That's really exciting, man.

Tell us about how.

A little bit about your
background, how you got into this.

Anto Patrex: Oh, sure.

Uh, Astro Martin was just an intern.

I was just an intern, so humble.

Alright.

Not, not a full-time job.

Yeah.

Um, yeah, so, uh, I was always interested
in building stuff from a very young age.

Like started building, using
physics engineering, uh, building,

started building a lot of products.

Right.

Uh, after my, during my, uh, you know,
time at Astro Martin, that's when most of

the ai, uh, kind of kicked, kicked off.

Like OpenAI, Sam was there
building OpenAI, one of the

biggest AI companies out there.

And, uh, from there I transitioned to
like ju uh, GPUs and super clusters and

NASA was going big on GPUs as well, right?

So most of my experience comes from
super clusters where the brain of

AI is actually in, where the models
are getting trained, the algorithms.

All work in the behind the scenes, and
that's kind of the experience that I

carried on forward to X AI as well.

So at that time, XI was building
the largest super clusters,

a hundred thousand GPUs.

H 100 is one of the
best, uh, GPUs out there.

So I was like, oh my God, this is a great
opportunity for me to actually be, uh, the

firsthand person dealing with the models.

Uh, and, you know, let's explore
the universe as Elon says.

So, so, uh, it was a great
experience working for them.

Arun: Yeah.

Fantastic.

And so you went from XI to start
your own company, cosmic Brain.

Yeah.

So Cosmic Brain, let me see if I
understood it and you can correct me.

Um, you guys are building world
foundational models, right?

And so you guys are approaching
the data side of things, right?

You guys are creating real world
models that you're, um, like robotics

can train off of for industries and.

For robotics in general, right?

Yes.

Yeah.

Anto Patrex: Yeah.

So, uh, we have came from lms mm-hmm.

To ISU text to ISU videos.

I think we are entering
the era of physical ai.

So physical AI is much bigger
than any other AI combined

that we ever seen before.

And it's, it's a very
data intensive process.

So this is gonna be the
most compute intensive, GP

intensive project there ever is.

Right.

So there is two different
ways of looking at a GI.

So as per OpenAI or as per Sam's belief,
uh, a GI will comes from reasoning models.

So if AI is able to ask you
questions and uh, it's able to

understand what you ask it, right?

So that is some sort of a GI, but the
way Elon thinks about it is definitely

in the terms of spatial intelligence.

So if the AI is able to see us.

Talking, there's mic in front of
you, there's table here, right?

And take a collective decision,
uh, looking at all the parameters

together, that is actual a GI.

Mm-hmm.

So physical AI is part of that.

So we are getting into a space where we
have a ton of video data, millions and

millions of hours of video data, and
we can actually take 40 d simulations

based on that and train models to
actually understand the physical

world and make decisions out of it.

So, and the first application
or the real world application

comes in the field of robotics.

So robotics are nothing but uh,
another version of human beings.

They have arms, they have
legs, uh, they see things.

So camera is the, uh, primary
input, uh, for data, for robots.

And we are building the brain for these
robots to actually function it so that

it can understand the world as we see it.

Arun: Mm, that's really cool.

Yes.

So you guys are creating these examples.

That maybe it's harder to
test in the real world.

You guys can create these simulations
and like expand the number of

areas that a robot can train
on in a shorter amount of time.

Anto Patrex: Yes.

So, uh, as I said, there is data
is the most important factor.

Uh, it's a golden key when
it comes to creating robots.

Uh, most of the data for this comes from
a point of view of the robot functioning.

So if you look into how human robots
are trained these days, right?

There is a ton of humans, 20 to 30
humans standing on a factory floor, uh,

wearing apple vision pros, and they're
doing different tasks from, uh, you

know, is placing a cup in on the table
or taking a dish or even drinking water

from a cup cup, those kind of stuff.

So humans are actually
standing eight hours a day.

In the entire shift doing different task.

And when we feed this data into the
AI model, we replace the human arm

with the robotic arm and we say,
Hey robot, you have to follow this.

And this is the way you
have to lift the object.

The pressure applied, even
I lift this bottle up or

interact with this environment.

The pressure, the physics,
everything has to be replicated

from that human video, right?

So, uh, this data is
not, uh, you know enough.

Mm-hmm.

Right?

So that's real data, right?

So, uh, what we are trying to do is.

Expand the data sets that
is using synthetic data.

Mm-hmm.

So the world has a lot of real data,
but we need millions of, of hours.

Hours or more data.

Right.

So we create wall foundation models
that can take this bottle and

change the texture a million times.

So from glass to plastic to wood,
to any other material you can

imagine we will implement on this.

Right.

So the robot actually visually learns it.

It's just like as a human child.

Mm-hmm.

Uh, learns from life experiences.

Right.

Nice.

So if you are.

This is the best example
I give people, right?

Even they don't understand how this works.

So let's say that you grew up
in a remote island, in, uh,

some part of the world, right?

And that person, and you are coming to
visit some, uh, highly advanced city

in the US for the first step, right?

You wouldn't even know how to open
a Coke bottle 'cause you have never

seen Coke bottle in your whole life.

Right?

Or go to a supermarket how
to check out and stuff.

Right?

So it's the same case with the robots.

So if you introduce robot to a new
environment, it'll be super confused

because it has never seen those visual
cues before and they don't know how

to interact with the environment.

So that's why we are feeding a lifetime of
visual experiences and cues into the brain

of the robots, uh, all in a factory floor.

Before the public launch so that
the robot decreases the error rates

and it's easy, uh, and seamless
to operate in the real world.

So that's our objective.

Vance: Where does the
synthetic data come from?

Anto Patrex: Uh, mostly from
the World Foundation models.

So we take real data, print
the models on that, and the AI

generates more data based on that.

So alternate the texture of the content,
the materials ads, like, you know,

we get, we have physics engine as
well, so the way it operates is that

we have a digital twin of the robot.

We do have real world data.

Uh, uh, and we also have a physics engine.

So, uh, the physics can actually, uh,
replicate the friction, uh, coefficient

of the water, slippery surfaces.

If it's a terrain that's different,
like a forest floor, if you want to

walk on grass, the grip is different.

If you want to hold a cup, the,
the, you know, force applied from

the arms is different, right?

So we combine a physics engine with the
AI models to actually replicate all this.

Vance: The physics engine
has to be rooted in some.

In the AI model principle, right?

Yes.

Yeah.

The, the

Anto Patrex: principles are,
uh, every physics that you

Vance: observe in the
real world because 'cause.

What if you just try to
start bending physics?

'cause it's synthetic data anyways.

Yeah.

So we, we, the

Anto Patrex: thing is that we
don't want the chair to fly away.

Right, right.

Yeah.

We want the chair to obey
the loads of gra uh, gravity.

The table needs to be
positioned in the right place.

And when you, like if, if I click
this object like that, snap the

object, like the, it won't fly away.

Right.

Rather move this much.

So the displacement, uh, the lift,
everything has to be correct.

So physics has to be a big, so that's
why we need the physics engine as well.

Well,

Vance: what if, let, let's just uh,
pull that thread for a second though.

What, what if you mess with it to
where it, it doesn't make sense?

Uhhuh, uh, what happens?

What's the consequence?

Anto Patrex: Yeah, so the, the, I
think there was this viral video of,

uh, uh, how like robots in China, uh,
there was this, uh, concert, right?

So, uh, the robot thought there was,
so this was a concert happening, right?

There was a, in the stage there
was this musicians performing and

there was like robots standing
near the security guards, right?

So, and there was a big
crowd, uh, in the backend.

So the robot thought there's no actual
physical barrier and it's tried to

push the people forward because, uh,
it did not understand the physics

of the barrier in front of it.

So a lot of harmful, uh, you
know, after effects can happen

because of lack of physics.

'cause uh, uh, let's say that the
robot does not understand the sense of

touch if I touch you like this, right?

But if I like.

Pull you like harder, that's
hurting the human beings, right?

So it's super important, uh, for
the sensory inputs of the robots

to understand physics and how much
pressure you need to apply, where

there is barriers, where there is
guardrails and all those kind of stuff

to, uh, to prevent harmful, uh, stuff
from happening or reduce error rates.

So that, that's my point.

Yeah.

What,

Vance: what's the.

Guidance, boundaries,
regulations, and who's governing?

Anto Patrex: So that's
another question, right?

So, uh, I think as we are evolving,
so there, there has been a lot of God

rails put into lms, but we are still
in, in infant states when it comes

to artificial intelligence, right?

Mm-hmm.

Because this is, uh, the super
intelligence that we have never had

in the entire history of human beings.

We, for the first time, humans have.

Are dealing with a higher
form of intelligence.

And right now we are starting to think,
Hey, this can actually get dangerous.

Mm-hmm.

So we need to put guardrails, and
it should come from a global scale.

It should, it should, it should not be
just constrained to the US or a couple

of people from the uh, Silicon Valley,
but the whole world has to come in

together to actually put guardrails
and ethics on top of this so that the

robots do not do anything that harms
the human beings or other robots.

Or cause like political conflicts
in a, in a geopolitical environment.

So yeah, we are in the
process of doing that.

Vance: What does that process look like?

Anto Patrex: Well, uh, the process,
so there are a couple of companies

or a couple of people who have
come ahead as of today, uh, and

working in the governmental scale.

Uh, people are, uh, talking about
this in DC people are talking

about this in, uh, France.

In the last AI conference,
JD Vance was there.

So, uh, I think.

As more brilliant people come together
and put more importance into this,

we are creating a framework of
ethics from that has to be obeyed by

mo most of the major AI companies.

So yeah.

Yeah, that's the

Arun: way I think what's really cool
about what you guys are doing is, for

example, like just training cars to
do self-driving has been, has to go

through a lot of red tra like, because
you need real world like use cases

and test cases to be able to do this,
but you guys are simulating all this.

So now you can, like you said, you
can generate all of these different

scenarios and it reminds me of kinda
like if you could just, you guys

obviously seen the matrix, right?

When you're trying to learn a new skill,
there's like a whole library of things

and just downloads that new skill into
you and so you're creating this library.

But the question I have is kind of
similar to what, uh, Vance is asking.

If you're using AI to train
ai, how do you prevent mm-hmm.

The kind of hallucinations
that we see right now?

Yeah.

With generative ai.

For the data set that you're kind of
using your models to get trained off?

Anto Patrex: That's a,
that's a great question.

So, because ramifications

Vance: in the physical world,

Anto Patrex: right?

Yeah.

You can see and you're
a little bit greater.

That's

Vance: than a chat bot.

Anto Patrex: Yeah.

I think, uh, the best thing we can
do is human reinforcement learning.

Okay.

So humans actually teased
ai, what's wrong and right?

Mm-hmm.

So, uh, it's like a one, AI is a kid.

That a thousand different humans and
staying together to train one kid.

Right?

So let's say that they, for Lums, the
way it works is that, uh, we give them

like a million different questions.

And out of that, if it answers like,
probably like 40% of is wrong, we

take those 40% of data, we teach
the ai, say, Hey, this is actually

a wrong answer and this is the right
answer, uh, uh, the correct answer.

Mm-hmm.

So we feed the, uh, the right
data back to the AI models

and it learns again and again.

So that's called reinforcement learning.

Right.

The same thing applies
with physical AI as well.

So if the robot does a task that's,
uh, it's not meant to do, or if it

hallucinates in the brain, right?

Mm-hmm.

We teach the robot, Hey, this is not
the way to do, and the humans intervene

and say, this is the right way to
do, this is the right path to go.

So human reinforcement learning is
the most effective, uh, uh, you know,

uh, correction parameter there is to
actually, you know, making AI better.

Arun: Okay, cool.

Yes.

So there's a lot of jobs then I guess.

That, you know, people are always
worried about AI taking jobs away,

but you're always gonna need, at least
for now, human intervention to make

sure things are going as expected.

So like there's a lot of
this happening right now.

And is this something that internally
at your company you guys build or

you guys use, like external kind of.

I think

Anto Patrex: as my company grows
right now, we do have like good

models that we have trained on.

But as we get more data sets,
uh, and as we grow as a company,

we will definitely need to have
a team just to teach ai Correct.

AI and how, uh, you know,
prevent it from Howing.

Vance: And when you said, uh, people
are wearing Apple Vision Pros,

like literally Apple Vision Pros.

Anto Patrex: Yeah.

Uh, it can be any VR headset.

Yeah.

But mostly have evolution Pros.

Vance: Why, why the choice
of Apple Vision Pro The best.

Anto Patrex: Uh, I don't,
uh, they, they're pretty

Vance: good.

Gotcha.

Yeah.

That, that makes sense.

So let's take a step back.

What's the big, bold vision?

Right.

I mean, you know, I get so
intoxicated by conversating about

ai 'cause it's just so cool.

It really is.

It really is.

Yeah.

And it just kind of blows
your mind on the capabilities.

'cause it feels like we went from
zero to a hundred real quick.

Yeah.

Mm-hmm.

Right.

But I feel like, just like crypto
sometimes you get so intoxicated

by the technology, you forget,
wait, what are we trying to, what

are we trying to accomplish here?

How do you contextualize it?

What's your, what's your big,
bold vision for the future?

What does it look like in two years,
five years, five plus years, anto?

How does the world fundamentally change?

Mm-hmm.

And how do we achieve more human
flourishing because of this?

Mm-hmm.

Anto Patrex: Yeah.

Uh, uh, that's, that's another great
question I would love to answer, right?

For sure.

So, uh.

As because, because of the, because
AI is there in the world right now,

human advancements have been twofold.

Like it's two x three X every year, right?

Because we have, we are seeing,
uh, advancements in a single

year that the humanity have never
seen in 50 years together, right?

Mm-hmm.

So AI has scaled the advancement
and progress of humanity

at a rapid, rapid scale.

Right.

So, uh, it comes from one, the
hardware, the, the GPUs, the

compute, the silicon chips, right?

And the silicon chips
are greatly improving.

Nvidia is at the forefront of it.

The GTC Jensen like launched like
hyper uh, powerful, uh, you know,

uh, compute systems out there.

I set my team there.

They were just

Vance: amped up.

It's crazy.

It's crazy.

Crazy.

Yeah,

Anto Patrex: crazy.

So that hardware will definitely
enable a larger industry for software

industries to come out, especially
being from AI to physical ai.

Uh, so there we are creating a,
uh, industry for humanoid robotics.

So there are industries like, hey, like.

Human labor.

So there's labor.

One of the biggest problems right now
in a lot of developing countries and

also in developed countries, is there
is a huge shortage for human labor.

Mm.

Right.

Humans do not wanna do all the hard
tasks that we used to do 50 years back.

Right.

And that's where, give examples,
uh, for like, you know,

uh, factory labor like Yep.

Picking up like a hundred, a hundred,
you know, a hundred pounds of, uh,

you know, load and stacking it in the
shelves or even agriculture, right.

Uh, going in the fields, working
like 24 hours out in the sun.

In the sun.

Yeah.

So those kind of tasks needs
to be replaced by, uh, robots.

Right?

Wow.

So the advantage that we get
there is that humans get tired.

Uh, they need breaks, they
need, uh, uh, sleep, right?

So if we replace that with
robots, 24 hour construction, so.

Uh, cities can double in size.

Mm.

So the, the, uh, the
skyscrapers can grow overnight.

So we can introduce like thousands
of robots into these cities

and, uh, construct like things
that we have never seen before.

Vance: Robots.

And you also kind of use
humanoids for, uh, construction.

Construction for example.

Um, are you saying that humanoids are
always the best form factor or No, like in

construction, robotics could be like, you
don't need it, it shouldn't be humanoid.

Yeah.

Yeah.

Anto Patrex: Uh, it's, uh,
I, I would say humanoid is.

A basic factor, but it also depends on
bio mimicking so we can actually make

robots of any kind of animals out there.

So humans, uh, we are, companies are
actually mimicking ostriches, uh,

giraffes for different, but, but why,

Vance: why, um, any sort of
human or animal versus like.

You know, more automated
cranes and things like that.

Mm-hmm.

Like talk, talk to us about that.

Anto Patrex: Depends on
the functionality, right?

So, uh, let's say that I
wanna lift up heavy objects.

Yeah.

So there might be a, uh, a better
physiological structure, uh, in the

organical world, in, in the animal
world that can better perform that

more than a human structure does.

So why do, why not replicate that?

Yeah.

So even for right now, there are Robert,
a lot of roboticist companies out there,

which have, in order for faster maneuvers,
they have replaced ostrich flex.

So the, the way the legs have bend and the
way the, uh, robots have to run and stuff.

Right.

Vance: But it's so interesting
that the base design is god's.

Right, right.

Creations.

Yeah, absolutely.

Yeah.

It is very much interesting.

Yes.

That

Arun: is like kind of like I
would actually agree with him.

I would've thought like why not?

Like a would a fancy crane
that's like automated, right?

Yeah.

That's like we, because we
have robotics now, right?

Yeah.

It's not humanoids.

It's not like some creature that we
can kind of like create a robotic

version, but it's more of like a
crane that doesn't have an operator.

It's just a computer running it.

But it's interesting to think that
that's not the most efficient thing.

Vance: Yeah.

And it seems like in a lot of cases.

Uh, you know, AI design
is inspired mm-hmm.

By organic,

Anto Patrex: yeah.

It's called Bio Mimicing, so, yeah.

Wow.

It's, uh, in inspired from the nature
we replicate it, make tasks much easier.

Yeah.

So that's the way to go.

And to answer the second part of your
question, what is the future like?

Right?

Yes.

What is the future in, uh, let's
say the, in the next, uh, 10 years?

Mm-hmm.

Excuse me.

Yeah.

So we are entering into a space
of like advanced computing.

Yeah.

And that's where Quantum comes into play.

So I do run a non-profit
quantum society in the Bay Area.

Mm-hmm.

Where I try to, uh, bridge the gap between
quantum computers and AI companies.

So, uh, quantum is an.

In the realm of multiverse.

So it, uh, it thinks in the terms
of atoms, not like binary systems.

Most of our computers
works with zeros and ones.

Quantums is in qubit states.

So when we have an advanced computer,
like quantum computers introduced into ai.

That will actually bring
in a truly conscious, uh,

intelligence into life, right?

So we can think in terms.

So the way our human brain
works, it's not zeros and wants.

We are thinking in terms of qubits.

Like we, every neuron that's
fired is an atos, right?

So we are bringing this into
the field of ai and that will

create new set of materials.

So we are entering into a world where.

All the periodic tables, all
the elements that we see until

now can actually be improved.

So even the TEFL coating of your plant
frying pan can actually be made more

healthier using quantum and ai, even the
rocket engine, the propulsion systems.

So right now we are actually boarding
fuel to actually proper as to space

that can actually be replaced by.

So the periodic table

Vance: is not complete.

Anto Patrex: We can make it better.

Yeah.

So yeah, material, material science
is one of the greatest events.

So what,

Vance: what happens with chemistry
courses in America, in the world?

Anto Patrex: Well, humans are
meant to change and adapt, right?

Yeah.

So we need, we should not hold onto
the past, rather strive to improve

and learn, uh, for the future.

Okay.

But.

Vance: Again, you know,
it's very interesting.

This is exciting technology.

Yeah.

What does that mean?

So, I, I got it in the first part of
your answer, which is, Hey look, you

know, we're gonna have 24 7 workers.

Cool.

Uh, they don't get tired.

Mm-hmm.

They're more efficient.

Uh, they're more accurate.

So we're gonna build buildings faster.

We're going to see human flourishing
advance because we're gonna be able

to develop the world, especially
in places that maybe didn't have

the economics to be able to do so.

Mm-hmm.

Um, at a rate.

Of lower cost and more
effectiveness, right?

So I get that.

Um, quantum computing, you know,
we're not just building cities

faster, but what's happening now?

We are creating

Anto Patrex: an intelligence that
can actually 'cause the, uh, right

now, when we think of a construction,
when we think about materials,

when we think about drugs, right?

We are thinking in a very
trial and error kind of phase.

We are trying different outputs.

Uh, we fail different times.

That's why the length of like a
kept, uh, let's say I have an idea.

So it takes like five
years for it to come out.

Especially drug discovery, right?

Or creating vaccines.

It takes like years and
years for it to come out.

One of the, uh, problems
being trial and error.

So how can we make it short?

So that's why Google came up with vow
chips and Microsoft with, uh, majora.

Mm-hmm.

So it can, uh, do tasks that within
seconds that take conventional

computers, like tens of trillions

Arun: to do.

One really cool way I've heard it
described is like, you know, like

those maze puzzles that you're doing,
like a traditional computer would

have to do each one sequentially.

It'll take down each path until
it finds the end goal that you're

trying to get to, which is the exit.

Whereas like a quantum computer can do
every single path at the same exact time.

And they'll find the, the
solution, which is really cool.

Right.

So it's even

Vance: more than a human.

'cause you said we think in quantum.

Arun: Yes.

Vance: But you're actually
talking about in that analogy.

Yeah.

Yeah.

Than human.

This is surpass what we can do more,

Anto Patrex: more than a human.

So it's like.

All the knowledge in the entire world
combined to a super intelligent, so

humans are limited by data as well.

Mm-hmm.

I do not have all this data and all the
knowledge in the world, but with uh,

hardware, with the computers, we can
just feed the entire data that humans

have ever learned in the past into one
system and you can introduce computer.

Okay.

So, but

Vance: you're saying,
'cause you know, obviously.

People hearing that for the first time,
at least half the people would be like,

that sounds terrifying, super intelligent.

You just did that with a smile.

Sounds like a horrible idea.

But I think what I hear you saying is that
no, it doesn't have to be a horrible idea

if we have the right guard rails because
we can think of cures for cancer faster.

Yes.

Mm-hmm.

Mm-hmm.

We can, you know, talk,
talk to me about that.

Anto Patrex: So let's say that
Covid came out today morning, right?

Yeah.

With a quantum computer, you can
actually dial the drugs and the

vaccine by the end of the day.

Right,

Anto Patrex: so you don't have to
wait a couple of months or you know,

scientists to bring a vaccine up.

So if a new disease is discovered in
the morning, you can cure it by night.

So that's the advancement.

You can cure cancer, you can
actually, uh, grow limbs.

Like, uh, there, there is like,
so right now the people are

looking into quantum computers,
into, in different angles, right?

People are actually using
brain cells to actually power

computers, actual brain cells.

To power quantum computers.

So the, the another factor is
that how far away are we from

actually achieving quantum?

Right?

So some people say we are 20 years
ahead, uh, you know, 20 years more to go.

Or, uh, Jensen said that, uh,
right now it's like 10 years ago.

Oh, within 10 years we'll achieve quantum.

So, uh, what do you believe?

Uh, I do believe 10 or less.

Oh, wow.

Or less.

Yeah.

Wow.

More than just this is, this
is not great for crypto.

Exactly.

Condom cryptography is much different.

It's.

Yeah, that's, that's a whole different
topic to talk about for sure.

Yeah.

So, yeah.

Uh, with ai this is possible 'cause
we are able to get, dive into

more data on quantum and drug drug
discovery, different ways of achieving

quantum photonics to qubits, to bio
quantum, to nano quantum, to a lot

of different ways of looking at it.

So, uh, Jensen actually had, uh,
two big panels with all the quantum

CEOs in this world, uh, talking
about quantum and how it can be

integrated with AI and GTC last week.

So Silicon Valley is actually adapting
from AI to, uh, the quantum era.

Vance: So when Jesus prayed on
earth as it is in heaven, do you

think he was thinking about ai?

Anto Patrex: I don't

Vance: know.

You know, because I mean,
he's a great healer, right?

And, um, you know, maybe to scale
that power, uh, it's not just

through Pentecostal, you know,
sensationalist services, right?

Right.

But maybe it's through
practical powerful means.

Mm-hmm.

You know?

'cause when you talk about.

Uh, you know, these rapid
drug developments, what

I hear is rapid healing.

Mm-hmm.

Anto Patrex: Yeah.

So I, I,

Vance: what I hear is less sick people.

Mm-hmm.

Anto Patrex: Yes, it's true.

Yeah.

Uh.

I think as I, uh, learn more about
technology, AI and quantum, especially

in the field of quantum, quantum is
something that will bring us closer

and closer to God and spirituality.

'cause everything that happens in the
quantum world is same like spirituality,

same like how God designed us to be.

It's like quantum is
all everywhere, all in.

Once like you can, using quantum
entanglement, you can actually

transfer data from one corner
of the universe to another con.

Teleportation is possible.

So ev or well teleportation is in

Vance: the Bible.

Anto Patrex: That's crazy, right?

Oh, really?

Oh, that's crazy.

Yeah, yeah.

Yeah.

Philip,

Vance: Philip baptized the
ethe eunuch and re reread it.

Okay.

I, this you put it in show, show notes.

Yeah.

Uh, the scripture.

Yeah.

Um, and the Bible says
that Philip transported.

Anto Patrex: Wow.

Wow.

That's crazy.

So

Vance: teleportation, so quantum
obviously exists, uh, in the totality

of, of the supernatural, right?

Mm-hmm.

Um, but you're saying
that science can reveal.

Maybe what's already
happening in the supernatural.

Anto Patrex: Yes.

So that, that's our mission.

I, I, I guess that's our mission.

To understand what, wait, wait,

Arun: wait, wait.

I just wanna back up.

Did you just say we could also
develop teleportation on earth?

Anto Patrex: Yeah.

Quantum teleportation already
achieved in terms of particles really.

So we can actually teleport particles
from transferring information.

So right now we require
Bluetooth, wifi Yeah, yeah, yeah.

Wire and all that to transfer information.

Right.

So using quantum teleportation,
you can act, actually transfer,

replicate the information in one atom.

To another atom across the world.

Vance: Ha.

Have they done that already?

Yes.

Anto Patrex: I guess

Vance: when was the
first time that happened?

I don't remember the date, but
yeah, years Or like five years ago?

Or 10 years ago or, no?

Anto Patrex: Very recently.

Recently?

Vance: Oh very.

This happened very recently.

Very recently.

Very recently.

Oh.

Like within the last few years.

Uh, not a few years or
like this past year.

Past year, I guess.

Yes.

Oh, really?

Yes.

Okay.

So then, wait, wait.

I must have missed this.

Yeah.

This feels like big news.

Yeah.

This is, this feels like, I feel like
I, I had to have a seen on Twitter.

We're like breaking it on the hot.

Yeah.

Yeah.

I don't think we're breaking it, but
like, but, but if you could travel, there

should have been a bigger deal in there.

There should, yeah.

That should have been a

Arun: big deal.

But if you could travel through space,
I've always considered that that

means you can travel through time.

Mm-hmm.

Is that.

True then too?

Anto Patrex: Uh,

Arun: potentially.

Anto Patrex: Well, maybe, uh, the
thing is that in quantum everything

that we know is real can be rerouted.

Interesting.

So, uh, any, the loss of physics
can change in the quantum world.

Wow.

So everything that we know and
understand of time, space, and reality.

It's different in quantum world.

Vance: So you said something
interesting, right?

Yeah.

You said, uh, hey, this is
gonna get us closer to God.

Mm-hmm.

Yeah.

This is the concern.

Yeah.

Right?

Is this Sounds like the Tara Babel.

Arun: Yeah, that's what I was thinking.

Yeah.

Vance: This sounds like the Tara Bab.

Mm-hmm.

How is it not,

Anto Patrex: um, that that's
a, that's a good question.

So I, I don't know.

Uh uh, so the.

The way an engineer thinks Yeah.

Is definitely, uh, I need
to, uh, solve problems.

Yeah.

Yeah.

I need to build, build, build.

That's the only way that humans, uh,
as an engineer, that's the way I think.

Right?

I don't think about, oh my God,
this, do I wanna take over the world?

Or anything like that.

Of course, guardrails are important.

Uh, on the other hand.

How

Vance: guardrails and guardrails.

Ooh, it's called 'em guardrails.

Exactly.

Anto Patrex: Uh, on the other hand,
uh, how I ca started thinking about

guard was when I actually, uh, with,
when I actually was dealing with wealth

foundation models and robots and ai,
uh, in order to prevent, uh, from robots

to do something, harm to the humans.

I was, I started thinking about 10
commandments for the robots, right?

Mm-hmm.

So that's when I started thinking, oh
my God, now I understand the reason

why God gave humans 10 commandments.

Guard rails, right?

Guard rails.

So if we disobey that, a lot of
things can go wrong, like space.

Same for robots.

If the robots harm another robot,
the robot harm another human.

Mm.

A lot of things can go wrong.

It's, it's not right.

So that's why if they disobey
the uh, uh, commandment.

It has to shut down.

Mm-hmm.

Anto Patrex: Right.

So, uh, now we are getting into a space
where we as humans are dealing with

intelligence and, uh, we are trying
to think like god's, uh, like God, but

I don't know if you'll get there, so.

Well, it's

Vance: interesting, right?

Build, build built sounds like Babel.

Yeah.

It does sound like, right?

Yeah.

So how, how, but it's

Arun: like, I think build,
build, build is okay.

The issue with Babel was
they, their target was they

wanted to reach heaven, right?

Like they wanted to.

Become like God, right?

What if you can build here and you
just set the right future goal?

It's, it's okay.

Vance: The gospel is kind of
the reverse polarity, right?

Mm-hmm.

Because, um, what do I
have to do to get to God?

Mm.

Vance: Right.

The gospel is actually
what God did to get to us.

Mm-hmm.

Yes, exactly right.

And so interesting that as they
were trying to do that right,

uh, their languages mm-hmm.

Um, you know, start diverting.

Right.

Which is an interesting.

Little barrier.

Mm-hmm.

Because I feel like they could have
just learned each other's languages.

Right, right, right.

You know what I mean?

Like, like, like people learn Spanish.

Right, right.

Arun: You

Vance: know, but if you

Arun: can't, if I can't communicate
with you, then it's hard.

No.

But over time, yeah,
it's gonna be possible.

Like through, but it slows it down enough.

Right.

People,

Vance: well, well, they,
it just didn't happen.

Arun: Right.

Vance: Right.

Right.

It just didn't happen.

Like the Tara Babel stopped.

Right.

Which means that the
heart of those people.

Mm.

Was never actually to do it in unity.

Interesting.

Was to try to do it in conformity.

Mm-hmm.

Right.

So you have to conform to my thing.

Mm-hmm.

You gotta conform to my name, you
gotta conform to what I'm doing.

So it's actually pride.

Mm-hmm.

Is what actually stopped the Tara Babel.

Mm-hmm.

Vance: Because it's actually different
in the New Testament in the upper room.

Mm-hmm.

The same thing happened.

Their language is.

Diverted.

Mm-hmm.

Right in the upper room it says they
started speaking in tongues, but it wasn't

like a tongues that was nonsensical.

Mm-hmm.

They started speaking, the Bible says
literally in the New Testament that,

you know, acts, acts two, uh, that they
started speaking in other languages.

Mm-hmm.

Actual like known languages.

Yeah.

Right.

But what happened was that people were
able to actually interpret that language.

'cause like, oh, that's Italian.

Right,

right.

Oh, and that,

Vance: that's Hebrew.

Oh no, that's.

You know, um, my language.

Right?

And all of a sudden, because
they could interpret each other's

languages, it actually brought unity.

So there was a diversity
in what they were doing.

Mm-hmm.

But because their intent was
of the heart of God mm-hmm.

They were able to still
move forward in unity.

And the Bible says that
the church grew rapidly.

Right.

And so I think what the Tar
Babel, you know, is a symbol of.

Is what is actually the intent.

Yes,

Vance: that's true.

I think maybe, you know, the build,
build, build in the AI world, we need

people to hopefully have the right intent.

Yeah.

Vance: Because if people are build, build,
build without the right intent mm-hmm.

Then maybe it's another destructive
narrative like the tarot Babel.

Arun: Yeah.

And I think that's why you even went
and approached a pastor to chat about

how do I build my foundational model?

With these God rails, right?

Mm-hmm.

Like you as a kingdom builder, like
think about these things when you

guys are building your company.

Exactly, yeah.

The

Anto Patrex: way I think about it is
not to become God or overtake God,

rather to understand God more and more.

There you go, right?

To know, understand in depth what
spirituality actually means, what

God meant for humans to actually be.

Why are we on this planet?

Why are we on this universe?

Right?

And we, even quantum is super
basic when it comes to God.

Like we are just starting
to explore particles.

We are just starting to explore
atoms and trying to, to tweak it to,

uh, you know, uh, meet our needs.

Right?

Right now we are, uh, humans are
just constrained to this tiny

planet called Earth in our solar
system, I think God is omnipotent.

He, he is like all around the universe,
the billions and billions of galaxies.

So we are nowhere close to it.

So no matter how much quantum is
there, no matter how much AI is

there, we are never gonna reach that
scale in the next billion years.

Wow.

So, yeah.

So yeah, that's not gonna happen.

Why, why are we here

Vance: Anto?

Anto Patrex: Uh, I think
that's, uh, I think it depends

on the, the interpretation
depends on person to person.

Mm.

But for me, I think it's definitely, I,
I know that I have no, the pa, my passion

and my goal is to build things and make.

Solve problems for humanity.

So I have a deep sense of, uh, will
that, hey, I can actually identify

this problem and using technology, I
can solve this problem for the people

who are suffering in this world.

So that's my, uh, uh, God's
purpose or something like that.

I don't know.

Yeah,

Vance: yeah, yeah, yeah.

Jesus, like we said earlier,
prayed on Earth as it is in heaven.

Mm-hmm.

I mean, revelation describes it as
there's gonna be a new heaven mm-hmm.

That actually overlays with the new Earth.

I feel like people don't actually
realize that Heaven looks a lot more

like Mountain view than they think

Anto Patrex: Interesting mountain view.

Vance: Interesting.

Anto Patrex: Yeah.

That

Vance: heaven looks a lot more
like San Francisco than they think,

oh, that heaven looks like a lot
more like Birmingham or mm-hmm.

Miami, Florida than they think.

Mm-hmm.

I think sometimes we think that heaven
is like, you know, this place where

there's, you know, chubby babies with
wings and you know they're playing harps.

Right.

You know, I, I don't know if
I ever read that in the Bible.

Mm-hmm.

But I did read that there's gonna be a
new heaven overlaying with the new Earth.

Right.

And so I just wonder if all this
technological advancement is actually part

of God's plan to see practical redemption.

Mm-hmm.

Yeah.

Uh, you know, restoration, why I say
that I think heaven might look more

like mountain view than people think
it's because it's not mountain view.

Exactly.

Right.

But it's a completely redeemed one.

Mm-hmm.

It's a completely renewed one.

Mm-hmm.

Without sickness.

Yes, exactly.

With full healing.

Mm-hmm.

With full restoration.

And when I hear things like, hey, like
we're 10 years away, advances from

creating, you know, vaccines in a day
that can heal thousands of people that

might be getting this one sickness.

Mm-hmm.

And we can heal cancer
and things like that.

I, I hear healing, right?

Yes.

I heal, I hear a redeemed state.

Mm-hmm.

You know, so interesting.

Really, it's really good.

Arun: I wanna pivot really quickly
because while we have you here, just

talk about just AI wars in general.

Like we have a lot of these
different big, big companies that

are playing the game right now.

Um, you, it's like, it feels like
every week you have a model that

comes out that beats the next model
and then the next, and it's just like

this race to see where it's going.

I'm curious what you think, um, 'cause
you're in the space where it's going.

First off, you have
Nvidia, which seems like.

You know, like 10 and heels above
every other hardware kind of play.

But on the software side of
things, where do you see it going

with the big tech companies?

You've got the Google's
apples meta mm-hmm.

Versus these, you know, other ones
opening ai, deep seek and stuff.

Anto Patrex: Yeah.

Uh, I think, uh, the race is happening
almost every day, not in months or weeks.

Right.

So what are you doing here?

Yeah, go man.

Like after this forecast, I'll
go check the Benchmark and some

other companies on top, right?

Yeah.

Arun: I think recently
Gemini came up with 2.5.

That just said, I don't know what
benchmark you use, but they claim to

be the best right now model out there.

And that was like yesterday maybe.

So, yeah.

So that's the thing.

Anto Patrex: One, the Gemini will
come up then next day open ai.

Yeah.

Another day.

GR three.

Yeah.

So GR three was one of the first models to
actually, uh, go in the 1400 scale mm-hmm.

In the chatbot arena.

So that, that's one of the
standard benchmarks out there.

Yeah.

So growth three came out to be top.

And right now there are other, uh, even
deep seek and open AI are all fighting.

Like it's one, one, uh, one
step ahead or one step down.

It's like.

Uh, the fight is always happening, right?

Which is good.

So, uh, when companies strive for
co be, become competitive and strive

for the to be the leader, uh, we
are seeing great improvements in ai.

Mm-hmm.

So right this week, open AI
released the new image generation

Ima image editing tool.

Right.

Last week, rock released
the image editing tool.

Right.

You don't need Photoshop anymore.

Mm-hmm.

In order to create, uh, you know,
posters and stuff, you just ask

AI and it will do it for you.

Arun: Have you tried this, by the way?

Vance four.

Oh.

New.

Uh, model to generate images.

Yeah.

Yeah.

GRS is really good, by the way.

Yeah.

Um, but the one that, it's like
taking over Twitter by storm,

like everybody's talking about.

Yeah.

Right now, the image editing, if you
haven't tried it out, definitely go.

So is Adobe stock just crashing?

It must be like you, like you
said, you really don't need those

tools anymore to do anything.

It's crazy.

Exactly.

Yeah.

Anto Patrex: And the other
fascinating factor was that.

If you had just two weeks back, if
you asked, uh, AI to actually make

a poster, none of the words Yeah.

The text.

And that was eligible, like
readable, like it will jule up the

words the English was not right.

But right now, in just two weeks, the
AI corrected all the words is readable.

Like you can create like
a whole passage Yeah.

Image.

And it's, you can read the entire text.

Arun: That's crazy.

I, I was wondering, just curious
what you think, how it's actually

happening because when I, when I
was playing around with it today.

It does a initial pass of creating the
image, and then it looks like it's almost

doing what you would do with Photoshop,
which is it's doing another pass on

the image and correcting the text.

Is it actually the AI model got
better or their feature of image

generation just got better?

Anto Patrex: Uh, it's,
uh, the AI got better.

Okay.

That's why I said because AI
got better because, uh, we

haven't fine tuning the models.

Yeah.

So it's more about fine tuning,
Hey, like the English is wrong,

if the letters are not readable.

If we ask the AI to repeatedly.

Corrected multiple times.

That's where the human reinforcement
element comes into play as well, right?

Mm-hmm.

So over the past couple of months, I
think image generation has improved

mostly because of fine tuning the models
and, uh, human reinforcement learning.

That's the two main parameters for it.

Okay, cool.

Yeah.

Arun: And so, um, actually I
wanted to ask you about the

deep seek stuff that happened.

So deep seek, you know, came out with
their really super efficient model.

They didn't require as much hardware.

There was some like, you know,
back and forth of how much hardware

they actually ended up using.

Um, and one thing I heard, like the
discourse that came about was because

they didn't have as such good hardware,
they were forced to innovate and

create a better model, like there.

Whereas here we have all this incredible
hardware and it led to like, maybe.

You know, like the common thing
that you hear at Google was

just throw more hardware at it.

Like you don't actually have to
create a more efficient model

or a more efficient system.

We just have so much money
and so much hardware that we

can just scale using that.

Mm-hmm.

I'm curious where you
see the dynamic between.

Where we are with hardware and actually
being able to create efficient models.

Yeah.

Anto Patrex: So deeps seek
is, uh, there's no deniability

that deeps seek is a bad model.

It's deeps seek is definitely
a really good model.

Mm-hmm.

But I'm not, I, I'm not fully aware of
how much hardware they used, how much

money they raised, or those kind of stuff.

Right.

So hardware definitely plays a
very crucial role when it comes

to the, uh, the power of ai.

So that's something that
was evident in X ai.

So Xai built the largest super
clusters in the world, a hundred

thousand GPUs, and they're probably
doubling it right now, right?

Mm-hmm.

So the, there's a technique that,
uh, that's used in, when it comes to

large scale infra, is that we use data
panelization and model penalization.

So we kind of distribute the load of
AI paddle into a hundred thousand GPUs.

So one GPUs, uh, one single G is never
exhausted, so the thinking is faster.

The processing is faster and
the results are more accurate.

So, uh, that's the step that
we took at, uh, XAI, right?

Um, and as the hardware improves,
we will see more powerful ai.

So hardware is definitely imp important.

And another issue when it
comes to hardware is the power.

So GPUs are the one of
the most power hungry.

Uh, hardwares in the entire world.

Mm-hmm.

And in the scale.

So the future of AI infra is that single
company will own millions of GPUs.

Mm-hmm.

1 million GPUs will require at
least 24 nuclear power plants.

Right.

Anto Patrex: That's crazy.

Right?

Yeah.

So AI is the most power hungry
animal the world has ever seen.

So how are we going to match
up with the power demands?

Right.

So, uh, in X ai, how they did that
was with Tesla Mega, mega packs.

Mm-hmm.

Anto Patrex: So, uh, because
there was like fluctuations

in voltage and frequency.

And they had to like, um, you
know, turn back to dc It's like

a war between Tesla and medicine
happening right now in the world.

So like they use DC power to actually
maintain the fluctuations and frequency

that the transformer did not burn out.

And, you know, subsystems
do not like, uh, flash out.

Yeah.

Anto Patrex: So, yeah.

So hardware is improving.

Another, the biggest concern
there is not hardware.

Uh, however that's a concern as well,
but the biggest concern is power.

Mm-hmm.

So where will we get the power from?

Are we investing more
in nuclear power plants?

Uh, renewable energy will never meet the
standards, so how are we going about it?

So the

Vance: deeps seek stuff was misleading.

'cause it's not like $6
million to build this.

Anto Patrex: So I cannot talk, I
don't know how to talk about it.

'cause I don't know in depth of how
they built it or the hardware they used.

So, but the malls are pretty good.

Yeah, that's what I can say.

Yeah.

Yeah.

Arun: Yeah.

And on that power, uh, conversation
about power, what's really

cool about innovation and like.

Uh, the competition we're seeing now
is it's actually driving a lot of other

sectors outside of just like tech, right?

Where now you actually have to see power
innovate to be able to keep up with this.

And you're seeing a lot of these
people who were slowly moving in

like the nuclear power space now
realizing, oh, we need to like figure

this out sooner rather than later.

'cause AI's coming one way or the other.

Mm-hmm.

Okay.

So let, let's do that.

Vance: Yeah.

Big winners, big losers.

Okay.

Yeah.

In the next three to five years,
um, you know, obviously anybody

that is in the forefront of.

Power.

Yeah.

Energy.

Yeah.

Big winners.

Mm-hmm.

Yeah.

I don't know.

What's big losers SaaS like this
is, is SaaS like a big, like what?

What's your, what's your take?

Anto?

I don't know.

Like big, you know, big three winners.

Big three losers because of ai.

Anto Patrex: I think energy depart,
energy sector is gonna be a big winner.

Big.

Big, for sure.

Big winner.

Batteries are gonna blow up, like mm-hmm.

A lot of battery companies
are gonna come out.

Robotics is the biggest winner, so.

Last week, uh, the figure ai Yeah.

Uh, private stock mm-hmm.

Is more in demand than Xai stock.

Xai was the highest in demand until now,
then it took over by robotics company.

Right.

Yeah.

Arun: Can you talk a little bit
about what figure just came out with,

because I know they, they made some
big advancements in the research.

Yeah.

Anto Patrex: So even I know the
public stuff they released, right.

So yeah, the humans are, the human
robots are able to walk like a human now.

Really cool stuff.

If you haven't seen this
man, it's really cool.

Do different tasks like
Boston Dynamics, actually.

B broke the silence last week
with, uh, the robot actually

running like a human being.

Yeah.

Crazy.

Taking a sprint.

They have a good

Vance: marketing team too.

They always like their
robots are always dancing.

Yeah.

They got swaggy robots.

I love the wonder where they're just like

Arun: pushing 'em around and it's just
like, that's the ones you always see.

It's like, why do you guys mad at them?

Well, the

Anto Patrex: biggest
concern is China though.

So China is unitary robots is
outperforming every other robot out there.

Interesting.

So.

I don't know.

China is like blowing up
on robotics right now.

Yeah.

They're super bullish on robots.

Their robots are doing back flips,
uh, all kinds of sort sorts of things.

So how, how many

Vance: years away till, um, I
have a humanoid net at my house.

You think?

Two years, five years?

Well, Elon

Anto Patrex: says three years.

Three.

Three years, okay.

Yeah.

So Elon's rights optimist timeline.

So in the cost of a car,
you can have a human.

What you think optimist is

Vance: gonna do?

Is he you doing my laundry
and my dishes or like.

Anto Patrex: Yeah.

I think the purpose of human robots
right now, uh, as per definition is to

replicate all the tedious tasks done
by humans from laundries to cleaning

up dishes, to, you know, cleaning
the apartment, mowing the lawn stuff.

Mm-hmm.

Yeah.

Mowing the lawn and all
those kind of stuff.

Yeah.

But that requests training, right?

So that aspect, which is

Vance: interesting that that's an
example of how, instead of making

all these different form factors.

The human form factor is pretty elegant.

Yeah, yeah.

Oh, for sure.

Oh, and actually a lot of the things
that we've built is for the human

form factor, it's Yeah, yeah, yeah.

Exactly, right.

Like so that actually makes sense.

Yeah.

Even

Anto Patrex: the, even this is designed
a way for the human that he can grab it.

Right.

So, yeah.

So that's other thing.

Vance: Yeah.

So all, all the world's
infrastructure is oriented, right?

Uh, you know, against
our organic structure.

Mm-hmm.

Um, okay.

So big winners is energy robotics.

One more,

Anto Patrex: and AI models slate.

Yeah.

Vance: Foundation models.

Yeah, the foundation models.

Yeah.

Who, who just top of
mind do you, like what?

What industries are you scared for?

You're like, oh man, that's
gonna be tough in an AI world.

I

Anto Patrex: don't know.

So it's something that one of the VCs
actually asked me a couple of weeks back.

Uh, someone asked me, Hey, why didn't
you actually build just an app and

make it, you know, release it, right?

Yeah.

So we are out of the face of apps.

You can just use it.

Like AI operator?

Yeah.

Like my news or like other
operator from philanthropic

just to create an app in a day.

Yeah.

So if you're creating an app right now,
I don't think that's the best way to go.

App builders, uhoh, I
think every, no, no, but

Vance: like you.

So there's a, there's
a distinction, right?

Mm-hmm.

Because, uh, Airbnb and
Uber is more than an app.

Ah, yeah.

Right?

So, so what, what, what's an example
of more so what you're saying, like,

Anto Patrex: so let's say that
there's an app to, uh, what

was it, app to actually, uh.

Uh, controlled appliances in your,
in your house or something like that.

Right?

So if I want, I can just ask him
a news, Hey, can you actually,

uh, build an app to actually
control my appliances in the house?

So take over Alexa and stuff.

Another thing is that, hey,
I wanna send bulk emails.

Mm-hmm.

So if you create an app for that, yeah,
I can just dial up an app in day for it.

I don't want to use your service if
there is an app to automate Excel

sheets, financial tools, right?

Yeah.

So we are getting into the
field of AI agents, but, uh.

AI agents should not be,
uh, limited to an app.

It should be more than that.

Yeah.

A, a true AI agent is something that
actually takes over that a hundred

percent of the task of a human being.

Yeah.

Be able to control different
apps in the desktop, close,

different, uh, you know, uh, tabs.

Write email for you.

Send it, record that in your
calendar, do different things.

So it's a, it's, it's beyond an app.

It's more, it's office system.

Vance: Mm-hmm.

SaaS is in trouble.

Arun: It is.

But that's why I think, and I
would love your take on this,

what the biggest losers for me.

Are actually gonna end up being all
these AI companies that aren't big tech.

This is my opinion.

This is why I think it's gonna happen.

Ooh.

Because I think the Googles of the
world are the ones to best give me.

That feature that you're talking about,
which is knowing me, they, they can

Vance: just collect up all their
apps to make the super app.

Arun: They make the

Vance: super

Arun: app, right?

Like they already have all my information
across their entire suite and now

Vance: they can have it.

Oh man.

Small tech is a big loser in general.

Huge loser.

Arun: Huge loser.

Vance: Oh man.

So big.

The bit, the Mag seven, they're just Yeah.

Get bigger and bigger.

Arun: Right.

I'm curious, do you agree or do
you have like it's a counter.

It's

Anto Patrex: actually.

Really difficult to
create an Everything app.

'cause that's the vision for X, right?

Mm-hmm.

So X is trying to be the Everything app.

Arun: The WeChat, right?

Of the the us Yeah.

Something like that.

Anto Patrex: I dunno.

Yeah.

So, but it's taking a long time,
so it's not, it's not an, an easy

task to make an Everything app.

Mm-hmm.

So it will take time.

So eventually, well maybe

Vance: not an Everything app, but um,
I just need that model to have access.

Right.

Arun: Say it again.

I just need my assistant
to have access like I need.

Mm-hmm.

I don't need it all combined into
one, but when I go ask Gemini or like

your app, Siri, whatever, it knows it.

So like I don't need it to be
all in one, it's just one device.

I have all of it allowed, and now
when I go ask it, it just gives me the

answer because it knows everything.

Mm-hmm.

Mm-hmm.

I don't need an Everything app.

I just need an everything device,
you know, with all of it loaded on.

Vance: Mm, that's true.

Interesting.

Okay.

What do you think about
form factors though?

Ooh, do you think?

Good question.

This is never going away, like
the smartphone, the phone.

Um, is this just too sticky or, ah,
is is, is VR actually gonna happen?

Is ar actually gonna happen?

Because, I mean, people have been
talking about it for 20 years, but

Uhhuh it hasn't actually happened yet.

Anto Patrex: Well, I think VR are entering
a phase where devices will be no more.

That's when newer link is there, right?

So we are entering probably in the next 10
to 15 years, uh, humans will be cyborgs.

Like, so we're all gonna be ironed 10

Arun: to 15 years.

Yeah.

So, yeah.

Anto Patrex: Is actually doing a
lot of stuff right now from here,

you know, enabling ly patients
to control devices and stuff.

So I think the future, future
is like, we don't, we'll, we'll

never carry a brick in our pocket.

Mm-hmm.

Well, we'll never carry devices.

Vance: Oh.

So that's a hot take.

Anto Patrex: Yes.

Vance: You're, you're saying
the smartphone is, is there's,

it's an existential threat.

Anto Patrex: Uh uh not
immediately, but in the future.

Right.

But in

Vance: 10 to 15 years, that's.

Anto Patrex: Uh, well, as
for my kids, as for the world

leaders, I don't do not know.

I have no, I do not have my
own hypothesis again, again.

Mm-hmm.

On top of it.

But as per the leaders in this
field, they say that 15 to 20 years.

So yeah,

Vance: that's, so you're gonna, um,
just, I guess, read your emails through

your contact lens or through thought?

Oh, just through thought.

Yeah, I can.

Oh, I see what you're saying.

Anto Patrex: So that's what
Neuralink is working on, right?

So, uh, with this implant, either
immersive or non-immersive.

Uh, chips in your brain.

You can actually think and talk to you.

So like tele kind of stuff.

Vance: So we actually,
so like Professor Xavier?

Yeah.

Everybody's like Professor Xavier

Anto Patrex: kind of.

So a month back, uh, I actually
hosted a, did a party called, uh,

in a Gigo party in SF in Temple.

Mm-hmm.

And we had some quantum art installation.

That's the most Silicon Valley
nerdy thing I've ever heard of.

That's awesome.

We are celebrating 25
years off the century.

I also wanna get invited,

Vance: so Yeah.

My invite must have got lost
in Yeah, yeah, yeah, yeah.

I'm gonna need the invite to the next one.

So

Anto Patrex: there we
have this immersive art.

So we, uh, there was this artist who
came in and she had these variable

devices, and once you wear it, you can
actually control the art installations.

You can actually draw images
on the screen and everything.

It was a non-immersive tech
that you wear on your brain.

Mm-hmm.

And use your, using your brain
signals, you can actually read your

thought and control the screen.

That's really cool.

So that's Yeah, like widely commercially
available, which is, is it gonna

Vance: be like luxury to talk?

It's like if you want, if
you want a unique experience.

Yeah, yeah, yeah.

This is a talking restaurant.

Yeah.

I feel like that's
definitely gonna happen.

I don't know.

Yeah.

Arun: There's gonna be retreats
where you just turn all of this

off and then you just like.

Like we're doing now back is
like, remember when we podcasted?

Yeah, yeah, yeah.

Remember

Vance: when we even had to do that?

So juvenile.

Yeah.

Arun: But I love this because,
so old school, I agree with this.

I agree that this is the
next era of where humans go.

And that's where I think we match the,
the like superior intelligence you get

from machines, because now we'll have
that implanted in US and now we can go

toe to toe with severe intelligence.

Vance: Oh, so, so you're
saying, uh, that you know.

The super intelligence will
definitely be smarter than humans.

Anto Patrex: Oh yeah.

Like we,

Vance: so we are at risk of Terminator.

Anto Patrex: Not really.

If you have a good guard rail.

So that's what the leader says.

Yeah.

But if, if, if

Vance: you are creating
super intelligence, wouldn't

they develop a free will

Anto Patrex: if you
have enough guardrails?

And we, if we don't teach,
teach it to be free.

Yeah, I guess it Movies,
code Us comes down to,

Vance: well, it all comes
down to intent, right?

Because it's in the creator.

Yeah.

Right now, ai, so people might
definitely build this without guardrails.

Yeah.

Arun: Or not know how to build all of
the exact guardrails that you need.

Right?

Yeah, yeah, yeah.

Vance: Either intentionally
or an unintentionally.

It is definitely happening.

Yeah.

That if you are saying that it's
possible for super intelligence to

be smarter than us, we're at risk.

But I guess we're at risk of like a media
hitting us too, so Yeah, that's true.

But I think that's why it's so

Arun: important to get this
digital twin thing right.

Uh, uh, because you, if you can copy
the human psyche in some digital

form, it'll have a morality, you'll,
you can figure out what that is.

Like, what is a digital
consciousness, digital morality,

what does that look like?

And then create the rule
book based off that.

Mm-hmm.

Because I think there's still
a lot that's, unless there's

a spiritual realm, right?

Like there's like still so
much undiscovered about.

How we work and operate as humans.

If you can map that into a digital
twin by like doing this like,

like reinforcement learning and
continuously building these models.

Mm-hmm.

Then I think you can get it right.

And I think that's probably, you
know, we, we think about movies when

we think about how this has been
done in the past, that's probably

something they haven't thought about.

Right.

It's like, how does that look and how
do you train and build these models?

There is a world in which this works

Anto Patrex: well.

Yeah.

I'm, I'm not a fan of Hollywood.

They always think, they always make
robotics and AI seem to be bad.

Yeah.

But it's not always the case.

So Yeah.

Vance: You feel like you have like a
optimistic lens, like of course I have to.

It doesn't seem like you're as
scared of them taking over the world.

Anto Patrex: Because there's greater
chance that we can advance as a

humanity compared to the risk.

Okay.

The

Vance: reward outweighs the risk.

The reward is

Anto Patrex: greater
than the downsides of it.

And if there's downsides there, we, that's
why guardrails are important, right?

Yeah.

So we need to implement
God rails to shut it down.

And we are at this phase in humanity
that all, all that we are speaking

right now, it's not fantasy anymore.

Mm-hmm.

It used to be fantasy two years back,
but right now this is actually real.

So crazy.

It's so true.

So yeah,

Vance: that's the title of the
podcast, by the way, guard rails.

God rails.

Dude, I love this guy.

Has the best of, that's locked.

That's locked.

I'm a, I'm a dad of four.

That's locked.

Is really, uh, no, that's,

Arun: I think there's, I think there's
been a really good conversation about ai.

I feel like we've been grilling
you, but I wanted to kind of move

away from AI a little bit and
just talk about you as a person.

Kingdom Builder, you're part
of the HYPE Network as well.

Why is it so important for
you to be a part of this?

And I know you have like something
that you're really trying to push

in this space, but like, yeah.

Why is this so important to you?

Anto Patrex: I think, uh, every
human being has to be grounded

in a sense of truth, right?

And, and something that
you can come back to.

'cause everything that
you build can be replaced.

Any company that you build, any empire
that you build can be replaced any second.

So, uh, at the end of the day, you need
to come back to God, the creator, right?

And that's what church is bringing
me, is the community, is the

spirituality that I'm grounded in.

And I need that.

Uh, more than anything, and that's
why I'm a super bullish hype as well.

Mm-hmm.

Like I'm, I work, uh, for free, uh,
any, any help that hype needs, so, no.

Yeah.

I don't, I don't care about the
monetary benefits or anything like that.

Yeah.

So it's, I think it's a mission for me.

It's rather than an obligation or work.

Arun: Yeah.

For people out there that haven't joined
the HYPE Network, talk about the value

that you've seen with, with HIV network.

Anto Patrex: Uh, there's
definitely and potential too.

Yeah.

Future values.

Yeah.

I think there is definitely a lot
of kingdom builders people who are

outside the realm of monetary benefits.

So one thing when it comes to, like a lot
of founders I've seen is that hey, they

wanna make millions of dollars mm-hmm.

And get away with it.

Right.

Just sell their company or quick benefits.

Right.

The way, uh, I think if you're a
true Christian, if you're a true

believer, that's not the way you think.

It's more about solving people's problems.

Good.

More than the money.

Right?

Yeah.

And that's the way I think.

And, uh, I, I think this is the
greatest place to meet people like that.

Mm-hmm.

And that, that's a value difference.

Arun: So.

Cool.

Yeah.

Fantastic.

Well, it's been a great podcast.

Um, I'm really excited.

50 episodes come on.

50.

We got high.

Great.

Dubai coming up.

Um, really glad that you're part of this,
uh, you know, space that we're, we're

building the Pipe Pod, the Hype Network.

Um, and if you want more information,
go check out hype network.org.

Um, get all the
information you want there.

Follow us on our socials and
stay tuned for the next episode.

Cheers guys.

Absolutely.

Cheers.

Let's

Vance: go.