And so there's a lot of groundwork that

is really important that's going into

the development for these robotics

companies that I think is really going

to pay off years down the road. Yeah, I

think the top humanoid companies are

going to be worth trillions if not tens

of trillions in the future, right? I

think yeah, I think there's going to be

multiple big winners and I think Tesla

and Figure are at the top of the pack.

It's just the same way there's a lot of

really great car companies out there,

right? And they all fit different

segments of the market depending on what

you need as a consumer. Some of them are

ultra luxury, some of them are more

economy, but there's a big market for

all these products. And that's right,

you're going to need a robot for the

home, but not everyone's going to have

the most premium products. Some might

have a maybe semi-premium product.

You're going to need a robot in

hospitals. Some you might need a robot

in an industrial facility. And those

robots could be designed differently.

For the home, you might need something

that's soft and compliant and safe,

whereas in an industrial setting, you

might need something that is more

durable and it can handle more weight,

etc. I think the smartest thing Elon

Musk ever did was deciding for Tesla to

enter the humanoid robot market in 2021.

He entered exactly at the right time

when AI and hardware were finally ready.

In that market, it's massive, estimated

to be a $5 trillion market by 2050, and

we're still at the very early innings.

There's no doubt Tesla will be a winner.

They're aiming to build 1 million

Optimus robots per year by 2030. But

Tesla won't be the only winner. More

than 100 companies are already building

humanoid robots and there are thousands

of players building all forms of

intelligent robots. Markets this large

create many winners. The problem is the

most promising companies figure

electronic and others are still private.

Most investors can't access them. Until

now. Last month, Robo Strategy went

public on NASDAQ under the ticker bot B.

Its goal is to give everyday investors

access to private robotics

opportunities. Today we get a chance to

meet Andrew Kang. Andrew is CEO of Robo

Strategy and built his reputation

running Mechanism Capital before

becoming one of the earliest major

backers of the humanoid robot movement

with a $19 million investment in figure

AI, now valued at 39 billion. We'll also

be joined by Dr. Scott Walter, a

two-time robotics founder who joined

Robbo Strategy and helps vet their

investment opportunities. Welcome,

Andrew. Welcome, Scott.

>> Hey, Herbert. Good to see you again,

Andrew.

>> Thanks for having us, Herbert.

Andrew, I've been wanting to talk to you

because you are one of the very first

people very early who had massive

conviction in this humanoid robot

market. Scott and I will tell you we

remember I don't know what it was at two

years ago in January. He and I started

doing these videos.

The kind of skepticism we had at the

time was never in my lifetime. Not going

to happen. Doesn't make sense. you put

in not only 1 million into figure and

then you said I'm going to put 19

million way back then. What gave you

that confidence at that such early

stage?

>> Yeah. Well, it was really a lot of

things. Um I mean one you you think

about how big the market is going to be

for humanoid robotics. I think all of

the the Tesla fans out there understand

that, right? Eventually we're going to

have billions potentially tens of

billions of robots out in the world. And

uh you can do the math, right? We sell

one robot for $50,000.

Uh what if we sell a h 100,000 of those?

It's $5 billion. What if we sell a

million of million of those? It's $50

billion. And so there's a pretty

plausible path where you you know you

scale these up 10 times year-over-year

and you could get eventually get to a

point where a major robot manufacturer

could be generating trillions of dollars

of revenue, which would equate to tens

of trillions of dollars of market cap.

Um, commercializing a humanoid robot

company though is is really really hard.

And so when I made the initial

investment, I really had to make sure

that this was the right team to do it

and that the probability of success was

going to be decently high. Um and the

after doing a lot of research looking

into the background of of Brett the

founder, seeing the success he had had

with his previous companies and the team

that he had brought on from not just uh

Google DeepMind but also from Tesla uh

the company itself uh from Boston

Dynamics, from Apple, really some of the

best roboticists around the world and uh

physical AI experts around the world. it

really seemed like this was one of the

few entrepreneurs and and teams that

were able to be up up for that

challenge. Um, and you know, as as

always, it's it's better to invest

earlier than later. Uh, when I initially

invested, most most of the venture

capital colleagues I I talked to, uh,

you know, they they weren't so

optimistic on the robotic space.

Previously, the space hadn't produced a

lot of venture scale outcomes. And so,

because of that, there was a lot of

hesitancy to invest. But I think when

there's a lot of fear, you know, it's

not always the best time to invest, but

sometimes it provides a pretty good

price. And and and so even though at the

time figure was worth $2.6 billion uh

post money, we saw the opportunity to be

in the trillions. And and so there was a

tremendous amount of upside still

available if they were able to execute.

And uh you know, that was that was one

of of many reasons why we made the bet.

I'm happy to kind of go more into

detail.

>> Yeah, let's do that. But uh to to set

the stage a little bit um I remember

those days and uh you know human or

robot comes in Elon says you know it's

time it's you have to have the right

actuators you have to have AI. Uh just

for disclosure uh Scott and I I think

we've always have told people that we

are such believers in human or robot

companies. So early on we ourselves are

investors in multiple humanoid robot

companies and you know and and the only

company that if you're if you're an

investor out there the only one you can

really invest in is the only public

company is Tesla. And clearly Tesla's

going to win um because of their

manufacturing chops and their AI chops

and so forth. But there's this is such a

huge market. We think that we should

invest to as many but regular retail

couldn't invest. But before we talk

about that, uh Scott, can you just give

us a lay of the land of what's happening

today from where it was before where we

are today? What's new information you

can share right now?

>> Okay. Well, glad to do that, but you

know, first what I want to do is I think

you said that, you know, Elon started,

you know, to ride the wave or catch

catch the wave in 2020 of the humanoids

because he saw the convergence. Really

should give him a bit more credit. He

more or less created the wave. So yes,

there was like this convergence of

technology that made it possible. Then

uh for the same reason that the

smartphone didn't come out in 2000, it

came around in 2008 because a

convergence of technologies, but you

still had Steve Jobs was like the thing

that kind of brought it all together

that allowed it to happen. And you know,

the humanit didn't start in 2020. They

they've been going on for a while with

little projects here and there. And you

know, robotics in general goes back to

like 1961 when the first industrial

robot is. But the thing is that they're

very kind of separate and the people who

are working in the humanoid spe were

always considered a little bit cookie.

So it was very hard to really get any uh

excitement to invest in there. You know

some people were getting pumped up about

Boston dynamics. But even then a lot of

people were just saying ah that's never

going to happen this Jetson's thing. You

know maybe that happened in my my

grandchildren's lifetime but it's not

going to happen in mine. But Elon did

see the convergence and he certainly

popularized it at that point. you were

no longer considered kooky if you were

talking about humanoids. It really

seemed like it was feasible and as a

result it really allowed the space to

open up and not just for capital to come

in there but it also meant that if you

were a robotic engineer you could pivot

and actually go into there and you would

be respected by your colleagues for

being able to do that. And so let's say

that moment kind of allowed me to

finally say, "Oh, okay. I'm no longer

playing with the industrial robots, but

you know, really going into something

was always a passion of anyone who's

been a roboticist." And that is to

someday have humanoids that are general

purpose that'll eventually make in their

homes. And I believe that that's that's

going to happen. So that's kind of

answering I think a little bit of your

question is that part of it is like this

convergence that was missing all along.

But the other part of it is the

excitement, the fact that we have so

many people interested in the problem of

solving it. People going to school to

learn about robotics, pumped up about

the idea, everyone wanting to be able to

startup. And the more and more you have

of that, the more likely you're going to

start seeing success. And we're right

now at the point that we're going to

start seeing the earliest

um adoption or let's say deployments

going on that are going to be able to

show people that it's not just cool

videos, but we are really getting the

point that the deployment phase is

beginning this year and then next year

is when you're going to start to see a

bit more and there's going to be pretty

obvious is happening and then each year

after that it's going to be more and

more clear and that that by 2030

everyone's going to know that yeah,

humanoids are here. Uh, I don't know if

you know, Andrew, but Scott and I have

done

130 plus shows for two, three years on

robots at this point. And the number one

question I keep getting from folks, I

think people are now convinced now that,

okay, human robots is a real thing. It

took that long for them to understand

this. And they're watching all these

demos from China, every bot demo, demo,

demo, and they're believing it now.

They're moving that direction. problem

is um many of them are just demos and

you don't know which is real. So what is

your like you you have I think if it's

correct you have your vision is two

things you're providing right is one is

access to private these these really hot

ones and then you guys do the vetting to

make sure that it's not just a fancy

demo.

>> Yeah, that's correct. And by the way,

Herbert, uh, two, three years ago when I

was initially doing research on the

humanoid robotics space, I watched the

the show with with you and Scott in it,

>> the first one.

>> Yeah.

>> So, it was one of the ways that I, you

know, initially got to know about Scott

and and connected with them. Um, but

going back to your question is, you

know, how do you know what's real and

what's fake? Is that that's the thing,

right, with a lot of these robotics

companies, they can have a robot try a

task maybe to, you know, package uh an

item and they could try 100 times and

they'll take a video of the just the

best try out of a hundred and that's the

one that they show online. And so you

really need to be able to weed out um

you know like what is just a

cherrypicked clip versus you know what

is a robot that is able to do tasks in

the real world autonomously and doesn't

need a human to fix it or to maintain it

you know every few hours right like

these things are these robots are only

useful if they don't really have much

human uh support at all um and so it

means you have to go out to the facility

You have to see the robots in action.

You have to play around with the robots.

You got to try to mess around with them.

See what happens if you block their

field of vision. See what happens if you

move items around. Um only then can you

can you tell if these robots are able to

work in the real world because the real

world is really messy. Uh this the

environment changes all the time. you

don't always have, you know, the the

items that you're working with in the

same place and you might have to work

with different instructions and and so

that is the beauty of robotics, right?

Is that or this new vision of robotics

is that they can adapt like humans can

and so when we do the dil due diligence,

we need to make sure that these robots

are really adapting and they're really

working without uh any human support.

>> Yeah. Yeah. I'll I'll just kind of

follow up on that is that you know we're

we're kind of fortunate that we don't

just have to to look at people's videos

that we are invited to go on site and go

out there and it's usually a lot of fun

and andrew has just he has read so many

papers on machine learning and how AI

works and you know exactly what causes

these models to break and and we'll go

out there and they'll be showing us like

a perfectly working model and will go

over and it'll just change one thing and

then suddenly you know it breaks. He's

he's really good at at being able to go

out and help us in those kind of

situations of knowing that we know

there's there's a certain distribution

that everyone's training in and

sometimes they have their own blinders

on because they think that's how the

problem is going to be and a lot of

times it takes someone from the outside

to say but have you considered this and

then immediately you see what's wrong

and what's been interesting is that a

lot of the companies we've done that

sometimes have been very quick to be

able to actually fix that problem. So,

uh, it has helped us to really be able

to tell, are we just seeing a demo or

are we actually seeing a team that is

able to react really quickly when they

start to see something that's a little

bit out of a distribution's not working

right, what is they can do to correct.

>> Gotcha. You guys have get get in. And of

course, you got Scott Walter. Um, Andrew

is like human or robots, I think, is the

way Elon positions it as being the

biggest market ever figure as a human or

robot.

you know that that's that gets all the

sexy stuff going on. But I I've been

telling my audience for so long at this

point. I guarantee you Tesla is going to

do many different form factors. Anything

that you that that can move with wheels

or whatever. It doesn't even have to

move. It can be anything that you put a

a chip on it and you put a camera on it,

it's going to become intelligent and

they will sell it. If you listen to

Elon, he has always says C3PO and RTD.

R2-D2.

Scott's title is what is it? Robotics.

>> Robotics research diligence director.

R2-D2 at Rob.

>> So weird, man. I would have gone with

chief diligence officer, but you want to

be R2-D2.

Um, but so you know, but it sounds like

your your uh philosophy and vision as

well, Andrew, is not just human or

robots. We're we're we're investing in

everything that is adjacent to the

robotics vertical. So it could be

humanoid robots and that's a big part of

our portfolio because we really think

that is one of the biggest opportunities

is something that is truly general

purpose. Uh if you look at the biggest

companies in the world, right, Apple and

Nvidia, they make general purpose

devices. For Apple, it's your it's your,

you know, smartphone, your iPhone, and

for Nvidia, right? It's the GPU.

And for all of these uh devices, right,

there's also the application specific

alternative. Uh but at the end of the

day, for general purpose devices,

because they're so adaptable and because

you can do so many things with them, you

can make them in such large quantities

that they become really affordable. And

when they become more affordable, then

well, it opens the market up to more

people. and and so you know humanoids is

something I would say is what we're the

we're the most excited about but there

are also other categories that are very

very exciting and we believe are big

opportunities as well one example is

industrial arms and cobots right

industrial arms they've been around for

a long long time uh I think one of the

things that has really maybe hindered

their adoption to the scale of what they

could be in the scale of you know

hundreds millions or billions is the

fact that traditional programming at

Scott knows very well has been very

cumbersome, very expensive, takes a very

long time. But as these physical AI

models get better, these robots, they

can do more things. They're more

adaptable. The time to teach them or

program them becomes very, very short.

And and so I think with that change in

mind in the future, you're going to see

a huge proliferation of these industrial

arms and cobots. And you know, if you

think about them compared to humanoids,

they're very similar in the sense that

they can also be general purpose. The

difference is they just can't move

around because they don't have legs. But

when we think about all these uh you

know tasks and all the things that you

would need to do in a factory setting or

in other settings right even in you know

hospitality or FND a lot of tasks you

just stay in the same place you don't

really need to be moving around it's

stationary you're doing the same thing

over and over and over and that could be

machine tending that could be things

like packaging that could be palletizing

etc. And so there's a huge market for

industrial arms and America, right? We

want to bring manufacturing back to

America. We want to re-industrialize.

And if we want to do that, we're going

to need a lot of industrial arms and

cobots. And so that's just one example.

We're also investing up and down the

supply chain. There's some really

interesting actuator companies that

we've been looking at recently, some

really interesting battery companies

that we're looking at as well. And you

know, these are all kind of components

that make up the robot that we believe

can be very important and potentially

large companies in the future.

>> One of my favorite episodes, one of the

episodes that had the most views was

Scott uh teaching us, remember Scott,

you showed us Giga Shanghai and you

would showed a video because there they

allowed a newscaster go in there and

Scott goes, "That job can be replaced.

That job can be replaced." And then he

showed us that some jobs you would use a

an ar a hand and arm and some jobs you

want robots with wheels and some jobs

you want a humanoid and you can pinpoint

exactly which role requires which one

and that was very well explained. Um so

the number one question I get from all

my audience members always is how do I

invest in robotics? uh because Tesla's

the only public one at this point and if

I want to invest in the private ones uh

luckily you know Scott and I were able

to do this because we uh we were people

reached out to us through this SPVS but

it's very hard for a regular investor to

do this and then if you do you want to

pick the ones that are the hottest you

know the ones who had the greatest

chance how does robo strategy how do you

guys get what you call what it's called

access right when when a new hot company

comes

they will pick and choose who gets to

invest. You guys are often one of the

first, not only at least there you don't

want to be like the 10th one in. How

does that happen and why you

>> Yeah, I mean as as I mentioned earlier

when we first initially started

investing in in in robotics, you know,

two to three years ago, I asked my

friends or, you know, our network in the

in the venture capital industry, should

we invest in this space? And they all

told us no because it hadn't produced

winners. And so, you know, we saw that

as basically an opportunity to create

one of the first robotic specific

investment firms. Um, and throughout

that process, we've gotten to know a lot

of really amazing founders, engineers,

researchers, etc. And we we've built

this network within the robotics

community. And, you know, we've gotten

to build a little bit of a brand as well

as investors that are really focused on

the space. Uh, they know their stuff. uh

both from a commercial perspective also

from an engineering and also from a

research perspective and you know while

building this firm it's also important

for us to be as they call it very value

ad uh to the companies that we invest in

in support and so that could be from a

marketing angle that could be from

referring talent to them or introducing

uh in other investors and and so um you

know building the firm over the last two

years we've been able to build a little

bit of a track record for ourselves as

being uh good investors in the space and

it's something that we're we're going to

be able to continue to do in in the

future. And so, you know, in terms of

how the access comes to us, that could

come from other founders that have, you

know, we've had a work relationship

from. It could come from just cold

inbounds. We get cold inbounds every

single day from people messaging us on X

or LinkedIn or other venues and also,

you know, other robotics investors in

the community. It's a very what I would

consider a tight-knit or small

investment community. And so we're also

talking to other investors that share

our views and um you know also believe

in the industry growth and um you know

we we collaborate and and um introduce

each other to other companies.

>> Yeah. And then Scott, you uh now that

you're working for Robbo Strategy, you

get to go fly in and check out all these

factories. You get invited to do watch

the demos. What are you seeing now in

the market? uh anything that's very

exciting or something you discovered

like are are actuators ready now what or

you know where are we in this

>> yeah they're getting better and let's

say um there's a lot of people out there

that are kind of rethinking how to solve

some of these problems um so you know

actuators I think they are not a solid

problem yet but we're getting very close

to it being there's kind of two problems

with the actuators is is one are they

able to produce you know the torque

density that you need so that's a little

bit of a design problem there. But the

other thing we we've always also heard

about is like, well, how do we

massproduce these things because right

now it's a very very slow process. Well,

we're seeing a lot of companies that are

really coming together to try to solve

that particular problem. So when it

comes time to scale, I believe we'll be

able to do that. All the pieces will

kind of come together. Um, so that's

that's very encouraging. But what I

think is the most encouraging is the

fact that we have gone from the fact

that humanoids talking about humanoids

was kind of a stigma. It's like oh you

don't want to talk talk about that to

now you kind of have you you're talking

about it but you think well maybe it's

just us. They're not going to be

laughing at us anymore. But the reality

is they all want to start testing. They

all want to be buying. I keep on hearing

from people. It's like they're no longer

saying how do I invest in these

companies. It's like how can I get my

hands in humanoid? Can you just tell me

anyone that we need to get them in to

start testing? And these are from like

Fortune 500 companies all the way down

to small companies you don't know of. So

a lot of them are figuring out how to

put it in there. It means like that's

really real. And it's not just like in

the US, it's also abroad in Europe and

everywhere else in a lot of countries

where you think they would be so

conservative they wouldn't be doing it.

So there's that incredible interest uh

to be able to move in that direction. So

um that that's telling me that this is

for real. Things are going to move. the

problems are going to get solved and

that the other thing is that you are

also seeing a lot of companies willing

to pick up the slack as far as the

supply chain. So before you might you

know a lot of them are in the auto

automotive space right now so they

already know how to scale complex kind

of things and in the past if you kind of

approach them about this they would be

like no I'm not interested are you

kidding now it's like hey how how do we

get that business they they're going out

and they're trying really hard to

compete for the business amongst all of

them. So those are like the very

positive bullish signs that I see that

everything you hear is for real because

you know it felt a little bit different

five years ago and and one of my

favorite quotes from Arthur C. Clark was

like the space elevator will um will be

will be invented pretty much 50 years

after everyone stops laughing. And it's

it's sort of the same thing with with

the humanoid bots is that you know

basically the humanoids will be deployed

five years after everyone stops

laughing. And I think everyone has kind

of stopped laughing right now and they

aren't taking it serious.

>> Can you guys like I'll give you a fun

exercise. One is can you name a company

that you invested in and why can you

name a company you had looked at and

decided not to? And then maybe you can

explain if you are looking at the

Chinese. It seems to me that it's Tesla,

Figure, Electronic, and then there's all

these Chinese that many people think is

actually way ahead in some cases, which

of course got to look at a closer eye.

You want to you guys want to try that?

>> Yeah, I I think Andrew can answer the

the question about, you know, which um

which companies we can invest in. Um the

question of chi of Chinese bots.

>> Yeah, I guess um you know, China and US

is always a is is is a pretty frequently

asked question. uh you see a lot of

videos of Chinese robots and you know

especially the unitary robots right and

I would say most of the time it's it's

for entertainment purposes uh you know

it's somebody controlling them with with

a remote control uh they're dancing

around but for the most part they're not

really doing real real work and that's

kind of the difference between the

American companies and the Chinese

companies is the Chinese companies

they're more so building a platform uh

for people to build on do research on

develop their own applications and the

American companies eventually are going

to be uh following a similar strategy

where the robot acts as as a bit of a

platform but in the short term I would

say over the next few years they're

really not looking to commercialize

their robot uh unless it's for a

specific application for a specific use

case like you know within uh the factory

or or eventually for for the home and

they want that robot to be perfect when

they actually bring it out and sell it.

Um but in terms of you know the industry

and where where both the industries are

going in each respective country I think

both of them are going to be great.

They're both going to be very massive

industries independently. Uh with China

right you've had a lot of government

support a lot of government funding

going into the industry over the last

three years which is why it's really

kind of gone so big. You have hundreds

of robotics companies that are now out

there. Um and I think eventually the

robots are going to be very good. The

thing is when you look at Chinese

companies, I think you have to

understand that uh great technological

proc uh progress doesn't always

correlate with uh you know large market

cap growth or large investment return.

And uh you know as investors that's kind

of what we're mostly concerned about is

you know how large is the actual

opportunity? Can this become something

Tesla scale or or Apple scale? And

China, for example, they produce really

great electric vehicles, cars. They

produce great phones now. But you look

at the biggest phone company, the

biggest car company, right? It's it's

it's Tesla and Apple. And why? Well, why

is that? Is because these companies,

they can charge or they can produce the

highest margins because they produce a

very premium product. Where you look at

the Chinese market, they can also

produce a premium product. But if you

have say 20 companies competing with

each other, the margins really get

squeezed down to the bone. And if you

look at BYD, they sell more cars than

Tesla, but their market cap is something

like 5% of Tesla's market cap. And so I

think a similar thing is going to

probably play out with the robotic space

where you're going to have a lot of

great technological innovation. Uh it's

going to really benefit society. it's

these robots are going to become really

affordable in China and wherever else

they sell to, but it's not going to

produce, you know, an Apple or Tesla

size outcome. Uh, and for America, you

don't have as many companies out there,

but they can still produce really great

uh products. And because there's lower

competition, it also means as an

investor, we see a little bit lower risk

from somebody that can come out of

nowhere and and really take over. uh and

so it is a little bit more attractive to

us from an investment point of view. And

then when you think about well what if

Chinese robots enter America? That is

something we think is is really not

going to happen in in scale. You look at

what happened with Chinese EVs. We don't

really have them in America. And you

look at why well the Biden

administration they put a 100% tariff on

Chinese EVs. And it's very likely if you

look at some of the current bills that

are being pushed through Congress that a

similar thing is going to happen for uh

Chinese robots. Another aspect is

there's a little bit of a national

security concern uh where you look at

Chinese EVs. Another reason why they're

not prevalent in America is because um

the FTC put a ban on cars from foreign

countries with uh you know foreign

software and foreign telecommunication

devices which Chinese robots would would

be have a lot of right you know they'd

be able to see and hear everything where

uh where they're deployed into whether

it be homes or or companies and that

that is a bit of a concern for the

American government. And so I think the

American government if they want the the

domestic robotics industry to succeed,

they're they're probably going to

prioritize and create legislation to

empower American robotics companies.

>> Yeah, I agree. Everything you said very

wholeheartedly.

So um people can go to NASDAQ, they can

buy bot B robot strategy and you got a

basket of all sorts of different kind of

robot uh companies in there. There are

most of them private almost all of them

private and uh a chronic figure are your

you know some of them and there's

others. Would you guys invest like I

think the third or fourth one I keep

hearing about is Boston Dynamic owned by

Hyundai. Uh I don't agree but many

people are going they think that they're

third. What do you guys think about

them? would you invest in them?

>> Okay. Yeah. First you have to remember

the whole idea behind bot is that we are

investing in private companies. So it's

not to go out and create an ETF. So

we're not an ETF in the sense that you

know we're not buying public companies.

Uh if we were to invest in Chinese

companies that becomes very complicated

because it's very difficult to invest in

private capital in China. It could be

possible once they go public but then

again it's like you can just buy it

yourself. So why we want to put in

there? So we are really focusing on just

the the private market at first. We also

aren't we're really looking at physical

intelligence. So that can run the whole

gamut of course from just getting the

full vertical of a humanoid to uh other

robotic companies that are innovating in

a way that makes things very very

interesting. And so that's what we're

trying to do with with this whole

basket. And then you know look at

sometimes picks and shovels that are

going to be going in there. Um, one of

my favorite companies is uh that that

we've invested in is Path Robotics. And

um, they don't actually build a robot.

They're using Motorman arms. Um, but

what they've done is they've added the

secret sauce to make their application

work. And this is for arc welding. And

arc welding is one of the largest

applications for industrial robots. But

it's really, really hard because it's

not just getting the robot to move along

the path correctly. It's also getting

your process parameters right. That's

very very difficult. And this is exactly

what Path Robotics has done is they've

got like a decade of data that they've

been collecting and figuring out exactly

what they have to do so that when you

turn that arc on, it actually gives you

the seam that you want. So that's really

what they've been able to add to and

that's why we're excited about it

because there's a going to be a shortage

of of welders. Um it's just happening.

You know, there already is like a

shortage of them. It's only going to get

worse over time. So that's one of those

companies can kind of go in there and

it's really this physical intelligence

and that's why we're excited. The other

is that this past week there was another

announcement of the closing of of series

C with uh standard bots and um that was

uh uh we were we led that round. We're

excited about them because uh they are

bringing the actual construction of

robot arms back to America. So back in

1961,

the first industrial arm was a Unimation

arm. It was made in America and then

at one point we didn't have any robots

being built in America anymore. And so

standard is bringing that back. So it's

either been Japanese or German or

Swedish.

>> Um and you know and now of recent years

also uh China because they've been also

building a lot of arms themselves. So uh

we're bringing that back to America

which is why we're very happy to work

with Standard Bots. they we believe they

really are leader in this area no doubt

about it and um I think you know just to

go to the other question we are looking

at a lot of companies uh that's what our

job is we cannot discuss or disclose

anything about them you know course

>> uh the other thing is that you know some

companies we have passed on but we

really don't want to say because I think

it's kind of unfair to them because

every one of those companies that we

look at when we start looking at them in

detail is because there's something

there and we think they they are really

good and And there can be many reasons.

Sometimes it's just like a timing. You

know, it's like they closed before we

were ready because everyone has, you

know, there are different kind of

calendars and schedules when capital

races have. And whenever there's a

situation that we decide that we have to

pass at a company, you know, the first

thing I put down there and the last

thing I write is like prove us wrong,

you know. So that's I feel about is

>> I didn't really want you to name names.

I just wanted you to walk me through

your thinking. That's right. But a lot

of times it's not because there's

anything wrong with the company or the

people or something like that. Sometimes

there's just a few things. And the other

thing is like there are other companies

also bidding for them,

>> you know, and so someone else puts on a

term sheet, they get better terms than

us. And so it's like, oh, okay.

>> Yeah. You have to pick the winners that

you want to do. Um, can I ask you then,

Andrew, just uh, you know, you you

launched BT bought on NASA.

Congratulations. That what a huge

excitement that must have been for you

last month. And then there are some

people who make comments. They will say,

you know, that yeah, it's interesting.

I've always wanted to invest into

private companies. Here's now a new way

for me to do it, but the management's

fees is too high. 2.5%. Is that right?

What's your response to that?

>> That's correct. It's a 2.5% management

fee. And I think you have to understand

that uh there's a lot of costs and time

and effort that goes into our process.

Uh it's not like, you know, we just kind

of picked this off of the public stock

market. Uh we have a full investment

team. That's not just me and Scott,

right? We also have Jack Walter on the

team, Roland, Mark, and we're continuing

to expand in hiring other robotics

founders and veterans, right? Because

there's a lot that goes into it. We're

we're one, we're looking there, Andrew.

I think you said Jack Walter, you met

Jack Pearson. Jack sometimes they call

him Mini Scott. So I think that's

>> Yeah. Mini Scott is is our internal name

for for Jack.

>> Um, but right, we're we're going out and

we're not just hearing the founder

stories, but we're going out to their

facilities. We're seeing if these robots

actually work or not. We're

investigating their manufacturing

process. We're talking to, you know,

their supply chain vendors

independently, doing our diligence,

seeing if, you know, their claims around

their orders are actually true or not.

We're going out and we're talking to

their customers. uh you know, some

referred by the company, but some we're

finding through, you know, external um

what you'd call uh professional

networks. Uh and we're seeing like, hey,

when you're using these robots out in

the field, how how are they actually

performing? Are there any issues with

them? How would you rate them against

their competitors? Uh we're going out

and we're, you know, diligencing the

entire leadership team. Are these people

with a track record of success? Have

they had any issues in in the past? And

we're, you know, there is probably, you

know, hundreds of hours of work that

goes into each individual investment.

And it takes a lot of um, you know,

effort and capital on our side to do so.

And, you know, we're not just building

the investment team, but we're building

a policy team, a research team. We

recently hired Bill Hughes, who was a

big uh, mover and shaker out in in DC

and uh, influencing a lot of um, policy

out there. he previously was heading of

a 250 person team at the executive

office at the White House. Um, and you

know, America, we don't have a national

robotic strategy yet, but I think it's a

very good idea that we should have one.

And, you know, we're really trying to

put our best effort in to help bring the

insight and the, you know, the leaders

from the best robotics companies to the

table to be able to, you know, influence

domestic policy in a very positive way

so the industry can grow. So there's a

lot that uh the one the fee goes to but

then you look compared to a traditional

venture capital fund they charge 2%

management fee and a 20% performance fee

and so you know you do the math on if an

investment through a venture capital

fund what happens it goes 10x versus

through robo strategy you know I think

you might be able to see that maybe that

fee is not so high imagine if for

example you know you invested in a fund

that held SpaceX 10 years ago and they

charge a 2.5% management fee and um you

know well I guess you can do your math

yourself on whether you would think it's

worth it or not versus the alter

>> well most of them are SPVS and if

they're SPVS they take a 20% cut not

just the management fee

>> yeah let's talk about that 20% cut for

for a while that's like a capital gains

tax so in other words you pay a capital

gains tax before you pay the actual

capital gains tax on top of that so it

yeah that can be a pretty hefty

>> but it's worth it because it's early.

You're making a bet on

>> because it's because it's early.

>> It's early and it's going to jump high.

So, you're going to go, "Okay, fine. I'm

going to make a bet." And, you know, I I

hate investing in SPVS. Honestly, I'm

telling you, cuz like that 20% cut, it's

a little it's painful. And so, not

having to pay that uh but going through

this is is, you know, the 2.5% is higher

than the 2% management fee. That's

great. Can you explain um the NAV? So,

net asset value premium. So you have a

fund, you buy all these private

companies, there's a value. I mean, and

unfortunately that value is private. So

it's not a public market. So you, you

know, it's like it it's it's what the

mark what the last fund investment

valued it at. And so you don't know if

it's high or if it's low. But then on

top of that, the stock price for bot

could go up, could go down, it can go

up. And then how do how would an

investor look at that and go, okay,

that's a fair price.

>> Yeah, you know, that that that is

something that is uh definitely comes up

a lot. And what for to understand NAV

correctly I think is really important

because it it it is a little bit uh

subjective in the sense that we have a

framework for for marking NAV and that

is based on the last round valuation

uh and that is based on the last round

valuation and then we add a period right

where we have to do our admin and audit

process that could add a few weeks could

add a month sometimes a little bit more

and so there is a little bit of a lag

sometimes when NAV is reported compared

to you know what the company itself

might have done its last round at

another way to think about it is private

markets and public markets have a

different subset of investors in them

and sometimes those subset of investors

value assets differently and so an

example I like to give is if you took uh

say a tech worker or a factory worker

from China and you place them in America

uh and they might be paid four times

more, five times more even if they had

the same skill set. And that's the thing

with private markets is that you have a

completely different supply and demand

dynamic where the subset of buyers,

right, demand is this small community of

venture capitalists and you know their

network that might invest in these SPVS

compared to the public markets which is

almost everybody in in the world. And so

it's a much much larger universe. And

oftent times, right, you see when

companies go public, their valuations

are higher. And this is this is not too

different from I guess what you would

see from the whole private e equity

industry complex where there are

companies out there, public companies.

Their whole business model is predicated

on acquiring private companies.

>> Yeah.

>> And not taking them public, but

sometimes taking them public, but just

adding them to their balance sheet.

Right. The public companies themselves

they could be trading at say a 15 to 40x

earnings multiple and in the same

category private companies in the same

vertical could be trading at a 3 to8x

and so that company they could say spend

$100 million to acquire that private

company and as soon as added to the

balance sheet now that's worth 400

billion sorry 400 million based on those

cash flows and so this is like something

that you have to maybe consider it's a

little bit hard to conceptualize but

private markets and public markets, they

have a little bit of a different pricing

mechanism and and we're bridging that

gap.

>> Yeah. Okay. And and uh can you explain

this uh two billion equity facility? So

right after you launched then you made

an announcement that there's I guess you

have a partnership and that you have

access to $2 billion then that you can

do that to buy more uh robot companies.

Buy buy invest more and not buy more but

invest more.

>> Yeah. To explain that that that is not

um you know a credit line. we're not

taking leverage on that $2 billion. So,

what that means to us is we can

essentially raise more capital by

issuing new shares and Roth, our capital

partner, right? They would facilitate

that process for us. We're not going to

be issuing $2 billion of shares, you

know, in one day or even one month.

We're going to look to do it in a way

that is value accreative to

shareholders. Right? A lot of people

they see a big dilution number and of

course if dilution and share issuance is

not managed well and the proceeds of

share issuance and capital raises are

not me uh managed well then it could be

bad for a company. On the other hand,

companies can raise cash by issuing more

equity and reinvest that and generate a

return that is good for shareholders and

that that is our our intention. We we

don't want to sell stock to to really

negatively affect the share price. We're

only doing it in a way where we you know

if the value of the stock is above the

NAV it is accreative to shareholders

because you know we take basically the

public market cost of capital and the

public market prices and and um you know

we take that and put it into the private

markets at the private market prices

which which are lower and in that way

benefit shareholders.

>> Yeah. I mean the whole game isn't it?

that you guys are looking for these hot

private robot companies. You want to be

chosen first, get there, you vet it and

then you go, "Okay, I like it. Now you

need to have money to invest." Uh then

you invest in it and then it likely

should hopefully go up. Some will be

losers, many will be losers, but many

could be huge winners. That's the whole

point. And then when they go big, um

that's what you you need the money to

invest and that's what this is. Okay.

>> Correct.

>> Yeah.

>> Oh, Scott. So, you know, what is the

latest uh that you're most excited about

in the robotics industry?

>> I wish I could say

uh we're excited uh about a lot of

developments out there, a lot of new

companies. Um I guess you just have to

stay

>> You're not going to share it. Okay.

Gotcha. Gotcha.

>> I mean, I think we should we should talk

about Bigger. Um, if we're talking about

one company, I I don't know how much you

guys have talked about it on previous

episodes, but I mean, one of the reasons

I I really like them outside of, you

know, their competency in manufacturing

and hardware design and, you know,

physical AI development is they have a

very unique taste. Um, and you can see

this by the videos that they put out

that they're creating this this image,

right? They're designing the robot so it

feels like a premium product. It's

something that you want in your home.

And if you know you had visitors in your

home, you'd be proud that hey, look, I

own I own a figure. In the same way that

you'd want to own, you know, an iPhone

versus, you know, maybe like a cheap

generic alternative. Uh, and people

people are willing to pay for that that

premium experience because they know

it's reliable. Not just it looks good,

but it's reliable. And when I open the

box, right, the Apple box, it's like

this is really welldesigned. There's

been a lot of thought put into it. And

um this is one of the few robotics

companies out there that I feel like

really embodies that Apple spirit. And

so, you know, I I I like to call them

this like the Apple of our of our

generation. Um you know, they still have

a lot of work to do. They still got to

pull it off, but it's very few companies

out there that I I feel like that

attention of detail to product and and

design is is really embodied by by the

team. Um, and so, you know, if I want a

robot, right, to take care of my

parents, for example, like I I would

want to buy them the most premium

product, I wouldn't want to, you know,

rely on some, you know, a comp, you

know, a generic product that might not

have every single detail thought

through, right? Because if their design

is is off, right? Like, well, what do I

expect from its reliability?

if it's, you know, doing things around

the house, like do I want to risk it

falling apart or, you know, breaking

things or right because these things,

these robots can be dangerous if if if

they don't work well. So, I want to make

sure it's reliable and it works

perfectly.

>> So, three of us are all investors in

Figure and now people if you want to,

you can invest in BT to get access to

that as well because they're still a

private company. Um, but Scott, how are

they doing in terms of manufacturing?

So, I agree with you completely, Andrew,

that you know, you know, you see all

these bots from China and they're not

they don't they're first of all, they're

tiny like they're three feet tall.

People think that they're really big or

something like that. They're tiny or

they're just clunky and they clearly are

there for demo purposes, not for real,

you know, real working. Where are they

though? The the the number one most

important thing of there's three, right?

Isn't it? Intelligence,

hands, and manufacturing.

>> Where are they with

>> scale? So, um, they've been working on

on that quite diligently and as we've

seen since January they've been ramping

up in each month they've been able to

increase their numbers more and more and

I believe uh last month it was about 250

bots uh that were produced. So, I think

they're getting pretty close to like one

bot every 30 minutes something like

that. So, um, that's impressive. And

what that means is that by the the end

of the year, we're going to be talking

with the produce thousands of of bots.

This that's for figure three. Um, but

Brett has already announced the that

figure four, the design has been locked

on that and they're getting ready for

that. So my overall feeling is that the

the right way you do something is that

if you get this next generation coming

out, you don't take the previous one and

scale it up to like hundreds of

thousands of bots unnecessarily when

you've got another one that's like so

close behind. And they are learning

everything that they need to right now

about how do you scale up? So they're

already setting up bot Q. They're

getting their manufacturing processes

together. They're working with their

supply chain. They're finding out what

the kinks are in the supply chain, what

they've got to kind of work out and get

that better. So when it gets ready for

the next generation one, they're not

starting from scratch. They already know

what's going on. There's probably going

to be a lot of very similar components

that going to be going in there. So um

they are definitely well positioned to

be able to take care of the scaling

problem. Now we've seen that in China,

they've also been able to scale rather

large numbers, which sound pretty

impressive. So I think did about 5,000

last year and everyone's talking about,

you know, maybe to have similar numbers

this year, you know, maybe the order of

10,000. But the way they're doing it is

um not a way that I would say is really

long-term scalable because the bots are

all the components are being made from

aluminum billets that are just being

machined in CNC machines. And we know

this because I was at Monroe and

Associates about a week and a half ago

and there's going to be a podcast coming

out about that where it was very clear

that everything that makes up the unitry

bot all the the entire skeleton that you

look at that thing you can go yep that

whole thing was machined in a CNC

machine. And

>> they are able to do that in China

because they literally have football

fields just full of CNC machines. And

what that does allows you get to a scale

of about a thousand and then suddenly

you hit a dead end. And what you have to

start think about is if we're going to

go an order of magnitude above that, you

start to need these other production

techniques, which means you start going

more to casting and stamping stuff like

that. Well, we've already seen that

figure start to do the the casting a lot

of the components because when we

visited there for figure two, which is

like a very very early prototype, they

were doing exactly that. And we joke

that um the the one thing that Figure

produces more than anything else is

aluminum chips. and they get literally

like 95% of the aluminum that comes in

the door ends up exiting the door a few

weeks later as everything gets processed

down to get everything. So that's what

you can do like an early prototyping

phase, but you don't try to do that to

take it to full scale. And so this is

where I think you're going to be seeing,

you know, both figure and Tesla and

others that are thinking of scaling are

thinking more on that because um we

can't rely on this idea of football

fields full of CNC machines. That's only

for prototyping. And so that might be

something that actually is like to a

detriment of China is they think, well,

we could just crank it up a little bit

more, right? It's like, well, you

squeeze a bit more, but that order of

magnitude is going to require a

completely different process.

>> Yeah. There's going to be a lot of

growing pains, right? When scaling up

manufacturing, one of the things that we

hear constantly is that uh when when

these companies are getting the

actuators in from from their suppliers,

there's a huge variation in quality. And

some of them can be good and then some

of them can be really bad. And you need

to be able to have a process to be able

to differentiate a good one from a bad

one because sometimes it's not so

obvious. And if you don't have a good

process in place, well, you could have a

bad actuator, you know, in in your

robot, that that's not a very good

outcome. And um that's something that

just takes time, right? It takes months,

years to really iron out your your

manufacturing process, your QA process.

And that's what our portfolio companies

are are doing, right? You you can't just

start a robot company tomorrow and have

all that ready. Yeah. Um, and so there's

a lot of groundwork that is really

important that's going into the

development um, for these robotics

companies that I think is really going

to pay off years down the road.

>> Yeah. Yeah. They're getting the reps in

now with the supply chain so they have

the muscle memory that when they're

serious about it, they know exactly how

to manage the whole thing.

>> Andrew, can you give me a sense of what

you're thinking about timelines? Um, so

we know that Figure is one of the

leading companies. We don't, you know,

obviously Tesla's, you know, they've got

two factories that they're building, but

figures the other private company, the

private company is leading the most and

they're valued at 39 billion. Where do

you see them going to? Electronic is

worth 5 billion. The last both of these

are like previous uh funding market caps

or caps based on last funding. What what

what do you see happening to these? Some

people will say, "Hey, I missed I missed

the boat."

Yeah, I I think the top humanoid

companies are going to be worth

trillions if not tens of trillions.

>> Some more. Okay, we know Tesla could be

worth trillions market cap, but you're

thinking that even Figure could be worth

trillions, not just 39 billion.

>> Yeah, I think there's going to be

multiple big winners and I I think Tesla

and and um you know, Figure are at the

top of the pack. It's just the same way

there's a lot of um really great car

companies out there, right? And they all

fit different segments of the market

depending on what you need as a

consumer. Some of them are ultra luxury.

Uh some of them are are more economy,

but there's a big market for all these

products. And it's right, you you're

going to need a robot for the home, but

not everyone's going to have the most

premium product. Some might have a, you

know, maybe semi-premium product. You're

going to need a robot in hospitals. Some

you might need a robot in an industrial

facility. And like those robots could be

designed differently. For the home, you

might need something that's soft and

compliant and safe. for as an industrial

setting. You might need something that

is more durable and it can handle more

weight uh etc. And so right I think

there's going to be multiple big winners

which is why we're taking uh you know a

multi multiple company investment

approach. Um and you know I think can be

up there as well. Um in terms of

timeline right like I think the

intelligence is actually getting there

pretty quickly. I think we're at the

GPT3 level of physical intelligence for

robots. It's a little bit hard to see

because, you know, we we sometimes we

forget how bad GPT3 was, right? It was

giving us gibberish answers. It was

wrong most of the time, but it was it

was useful sometimes. The thing with

robots is that you need them to be right

or accurate 100% of the time for them to

be really useful. And that's kind of

where we're getting with some of these

more chatbot or LLM models now, uh, you

know, two, three years later. I think

that Excel that development timeline is

going to be really compressed because

there are a lot of learnings that we

went through uh when getting from GPT3

to five right that don't need to be

exactly reproduced all of the

innovations in say mid-training uh post-

training RHF uh right there's a

different mechanism for RL for for

robotics and right data filtering to

make sure that we have the right data uh

getting the right annotation

infrastructure in place a lot of these

techniques and research kind of

innovations we we can also bring over to

physical AI. And so I don't think it's

going to be three years to get to the

GBT5 level equivalent. I think it's

going to be maybe like something along

the lines of one to two years. And then

once we're there, well then it doesn't

mean we're going to have billions of

robots everywhere. We're still going to

need to scale manufacturing. That's

going to be a bottleneck for us. Supply

chain is going to be a bottleneck. And

so we're going to have really great

robots, I think, in two years time, but

maybe we're going to have maybe hundreds

of thousands. Maybe we'll have millions

of them. And then, you know, when when

we're four years out in the future, five

years out, I think then you you start

seeing, you know, hundreds of millions,

maybe maybe even billions in the in the

in the 2030s. But the market, right, it

always prices things sooner than later.

Uh it it it looks into the future and

that's why we want to invest now as

opposed to later.

>> Okay. uh you know what about the angle

that some people will say that I think I

believe it which is you need to first

get to call a number 20,000 30,000

robots get it out there in the live have

a huge supercomput with simulation and

then have all those 20,000 be trained

and then learning all sorts of skills

and then only could you once you have

that then you get to the you know

generalized GPT5 kind of level of

physical AI to be able to do everything

really really well. One shot uh being

trained it to do anything that you want

it to do. Uh do you agree with that?

>> Well, you know, instead of simulation is

one approach, right? There's a lot of

different methods to collect data for

these robot foundation models. Um the

thing with simulation is that it needs

to simulate everything to a very very

high degree of fidelity and that can

work for things like locomotion, right?

um a robot walking around where the

contact forces can be modeled very

easily. But for things like you know

you're working with deformable objects

or um you know fluids for example those

are really hard to model with

traditional you know physics or

simulation engines. And so this is where

what they call world models come into

place which you can kind of think about

as some people call it neural

simulation. It's it's kind of how humans

think, right, and act in the real world

is well, we don't exactly imagine every

single frame ahead of us, but we have a

sense of how the world may change given

certain actions that we may take. And

these world models are getting really,

really good. The backbone of these world

models is essentially these video

generation models uh like Sora for

example, right? And these video

generation models as they get better and

better uh at understanding the world

then the robot foundation models get

better and better as well because

they're built on top of these video

generation or world models. Scott, I

mean happy a little bit more, but that

is kind of the high level detail.

>> Yeah. Yeah.

>> Yeah. But you need to have 2,000 to

10,000 robots physically out there. And

so my point is that whichever companies

can get it out there, then you get the

data then you can learn faster.

>> Yeah. So from the post-training

standpoint, you would need to do a

little bit of that because of that

infamous sim to real gap. So again, it's

like what Andrew was talking about is

that in a lot of these physics

simulators, you can do walking pretty

well, but even then, when you bring it

to the real one, there's like a little

bit of adjustment you have to do to get

it kind of work the same, same way. When

you start doing manipulation, which is a

much tougher task, you would need a

little bit more. But these world models

are getting kind of better and better at

that if that they start seeing enough of

it. um it's almost like you don't need

the underlying physics. It just they

kind of get the idea of how it works

enough that things will start to work

successfully in the real world. So it's

a question of how many do we actually

need in the real world? Do we need

thousands, tens of thousands, or

hundreds of thousands? That's kind of up

for debate. Um I think everyone agrees

it's probably minimum like a thousand

that you would need to get out there and

you would probably be able to do uh

pretty well, but the question is whether

you need to go up a little bit more. And

that's kind of the race that's going on.

you've seen with everyone that you know

sort of the their second or third

generation robots they want to crank

them out in enough numbers that they can

start doing that and getting those

numbers. So we're seeing that a lot in

China that a lot of the companies that

was really only their second generation

bots they started to scale quite a bit

and then they're putting them in these

gymnasiums where they're getting lots

and lots of training data to try to to

cover that. It could be that um you know

there's going to be other methods that

come on there to do it. Nobody knows and

it it's every month it seems like the

industry kind of shifts directions

because a few months ago we weren't

talking about world models you know we

talking about blas and now suddenly

everything is shifting that way and it

could change uh another and we're seeing

the same thing like how what about data

collection what's the best way to do it

at first like you had to do everything

teleop with a real robot and then

suddenly well wait a minute um is coming

out now so you don't actually need to do

tell with real one so we can put on

these data gloves and go there capture

all the data that way. Well, now there's

like another push. It's like, well,

maybe we don't even need to do that.

Maybe we can just go pure egocentric. We

don't need to have special gloves or

anything like that. We can just go ahead

and put a camera up there and be able to

grab it. So, there's kind of this debate

that's going on within the community,

which is the right approach. I don't

know what the right approach is because

they don't know what the right approach

is. So, how am I going to know? So,

they're all kind of arguing trying to

figure that out. And we'll only sort

that out with time. But I think we all

agree you need a certain scale of your

physical hardware to test it out to make

sure it actually works. But do you you

do not need it to be able to do all

tasks. You just need it to be able to do

enough tasks that it's useful and that

someone is says yes, it's ready to

deploy in this particular setting and is

able to do useful work. And of course

that means it's more likely to be in a

commercial setting where the task

requirements are far more narrow than it

would be for a general purpose robot in

your home. Unless there's really just

one task that you want the robot to do

in your home and then you know maybe

you'd be able to do that.

>> Well, there you go. Yeah, that's I

perfect explanation. Both of you

appreciate it. Uh so Andrew, what's your

what's what do you have forecast to

happen this year? What's what's your

plan? What are you expecting? What are

you excited about? Yeah, like I said, I

I think we're going to get a really what

I consider like jump for the

intelligence of the physical

intelligence models where before you

were seeing more so demos in in a lab

setting, but robots actually deployed

more robots being deployed into the real

world and for them to be actually doing

tasks without any kind of human support

at all. And uh we're going to really

start scaling manufacturing. So, you

know, at last year, you know, you were

seeing some of the American companies

maybe producing hundreds of robots per

year. Now, that's scaling to the

thousands. Towards the end of the year,

I think the pace is going going to

increase to the tens of thousands. And,

you know, every year that goes forward,

we're going to increase that rate uh

10x. Um,

and uh, you know, we're getting to the

point now where you're going to be able

to command a robot with with language.

Before, you know, it was a little bit

more technical. Maybe you you would need

some very I guess special type of

prompting or programming, but it'll get

to the point. It's getting to the point,

right? They call it lang language

steering where I can tell a robot, hey,

uh that cup over there, can you put it

in the dishwasher for me and it'd be

able to do it. And so we're we're seeing

some early uh you know, signs of life

there. And uh another innovation, right?

You and with with the world of LLM, you

saw a lot of what they call emerging

capabilities. One of the emerging

capabilities you you you saw was in

context learning. The ability for you to

just show uh you know an LLM something,

right? You you tell it, hey look uh this

is what I'm dealing with and it'd be

able to remember that and and and use

that context to give you an answer. Uh

that is happening with robotics as well

where some of the more frontier models

you're saying hey robot um maybe you

should uh draw a letter like this or

maybe when you're putting a cup you

should place it upright as opposed to

like this position right and it's

remember that so there's some really

exciting I would consider like emerging

capabilities that are emerging are going

to continue to emerge

>> that is so fun I think Elon even

mentioned one time that you know you

could teach something one shot like

you've never been taught it before But

you go, "This is the way I want you to

make my coffee." Okay. This way. This

way, this way. It's got to be exactly

this. And then

>> Yeah.

>> It's so exciting. I think we're all uh

one of the reasons, you know, Scott and

I started these started doing these

videos is we're witnessing history

happen from like when everybody said it

was nothing. Now we're kind of at that

toddler stage. Okay. We're watching

robots walk, run, uh we're watching them

do these simple factory tasks and then

they're they're on version three, right?

And then there's this leap that you're

talking about, but uh I still don't know

when it will happen, but just like AI,

you can't predict it. It can happen

really quick. It could take two years,

but generally it happens sooner than you

think. Just the way it's happening with

AI at this point, it's just coming

sooner than we anybody's predicting. Um,

and like you said earlier, the AI, the

work that they're doing with AI

translates,

>> right,

>> uh, to the physical AI except for the

actual hardware production and

manufacturing. That's why you guys have

to go and kick actually open it apart,

right? So, you tore apart that uh,

that bot from China. Yeah, you tore it

apart, Scott.

>> Well, yeah, I didn't. Monroe did. So,

when associates had a a unitry G1 and

they disassembled it

>> and we did a bpsy and we had a good

chance to kind of look at the different

components to get an idea how they were

manufactured, how they could probably

improve some of their manufacturing

techniques.

And so, there was actually some

suggestions that I think would be good

not only for unitry but anyone else

thinking of putting together a bot.

>> Good. Okay. Well, you you you got Scott

Walter uh joined the company. He he's

the one that checks out how if this is a

demo or not. He he knows what he's

doing. Tell us uh how do how do people

find you uh Andrew and how do they find

Robo Strategy?

>> Uh you can find us at robo strategy onx.

Uh you can also find us at

robostrategy.co.

That's our website. And we're going to

be putting out a lot of great content

information uh deep dives and interviews

with the founders of our portfolio

companies. So there's a lot of exciting

uh content that we hope to share with

everybody in the future.

>> Great. And you're on NASDAQ BT, but uh

again just general thing, this is not a

financial advice. Every person has a

very different position. You should look

at all sorts of and um you know your own

scenario. We don't know if you should or

shouldn't. It's your call. You decide.

But human robot is exciting and it's fun

and it's the next thing that's going to

go uh skyrocket. Thank you very much

both of you. I appreciate it. Thank you,

Andrew. Thank you, Scott.

>> Yeah. Thank you, Herbert. Keep building.

>> I've created a website that is the most

comprehensive resource for the Tesla

investor. Please check it out. Simply go

to my website at herbalm.com.