🚨 AI is NO LONGER a Tool… It’s an INDEPENDENT WORKER Taking Over!

Well, hello everybody. It is Thursday

where we do the roundup of all the

craziest AI news of the week. And

there's no video like this because, you

know, a lot of people do videos and

they'll spend 25 minutes talking about

one thing. I'll spend less time than

that talking about 50 different things

that you all need to know. And AI is no

longer a tool. It has changed and

project pro progress is exploding

exponentially like each week there's so

much new stuff cap capabilities are

doubling every few weeks as well and

we're going to break it all down. Thank

you all for coming. Thank you to the mod

in the chat as well. And uh never

financial advice. It doesn't matter. Who

cares? And thank you to Patreon 2 from

Steven H. I've been watching your

YouTube videos for a while now and I'm

ready to take next step and join your

Patreon. Woohoo. Thank you for joining

the team. You'll learn a lot. It's fun

even during these kind of choppy

sideways times. Um we try and stay on

top of everything. And of course, don't

forget there's a whole playlist here uh

once a week cover everything that's

happening in AI. Don't forget you can

subscribe to that if you're not

interested in anything else. It covers

everything. I'm going to begin with AI

and LLMs because there's a lot of news

here today. First of all, Anthropic,

they raised 30 another $30 billion at

$380 billion valuation. Their valuation

keeps going up literally, I don't know

what it was, a year and a half ago. It

was worth a tiny tiny fraction of that.

And that is their biggest funding round

yet. More than double the prior one,

which is only a few months back. And

their focus is on of course scaling and

coding and the enterprise. And these

guys are killing it. Killing it. There's

three players in town. Zanthropic,

there's Google Gemini, and there's, of

course, Grock XAI. That's it. There's

another player. We'll talk about them

later, too. But they're trying different

things. And remember, the world is now

multi- aent. This is another important

thing here from Anthropic 4.6 and even

Grock 4.20. You'll notice the whole

theme is multi- aent where multiple

groups work on the same project at the

same time. and uh the 4.6 from Claude,

if you're into coding with their 1

million token context, it is completely

apparently mindblowing. So, watch for

this. And it's funny to see people on

the world stage and the animosity

between different players. I thought

this photo today was very interesting.

There is an AI summit in India. Shout

out to anybody in India. Drop a comment

if you're in India. I'd love to know. Uh

but this is a whole bunch of leaders.

World leader, the guy in the yellow

shirt is Modi. Uh we had the UN

secretary was there. Uh you had Pichai,

Alman, Amodi, Macron all in New Delhi to

talk about AI and governments and all

that stuff. But the most important thing

here is that's Sam Alultman holding

Mod's hand. And this gets really

confusing. I got to get this straight.

Sam Altman is holding Emod's hand and

he's beside Emodi from Anthropic who

refuses to hold his hand. I thought you

can't make that stuff up. They don't

like each other at all. Animosity is

crazy. Nobody seems to like this Ultiman

guy. Not sure why. And speaking of money

raising, not just anthropic, but OpenAI

is about to raise a hundred billion at

$830 billion valuation. And that 1.7xed

in the last 12 weeks. Again, this is

where it gets a bit nutty. Let's look at

some numbers for a second just to put

things in perspective. This is Meta on

the left and Open AI on the right. Meta

is worth 1.6 trillion. Open AI will be

worth more than half of that 830

billion. Meta has 4 billion daily users

and makes over 200 billion in revenue.

Open AI has about 10 billion in revenue,

shrinking 800 million users because all

the other guys are taking market share

from OpenAI and Chachi because people

realize it's no good compared to the

other ones. And not only are users

shrinking, but they're burning 20

billion cash a year.

A lot a lot of cash. Anyhow,

you know, if you had to bet on one of

these, too. And this is kind of the the

sad thing about IPOs as well. They build

all of this up because they know an IPO

is coming. So, it's easy to fill around

because they know they can exit after

the IPO. Ness retail comes in last to

fill their bags and then it's too late.

If you're not into these things before

the IPO, don't touch them at all ever.

And speaking of Meta, still Nvidia is

planning to sell estimated allegedly

about $50 billion in chips to Meta.

Again, people say, "Oh, Nvidia is a

bubble." No. and they'll be selling

black wells and then maybe later

Reuben's and 50 billion a year maybe

even more I don't know the capex for

Meta I think this year if I recall 160

billion or something so 50 billion of

that is on Nvidia chips maybe estimating

here and back to open AAI 2 they are

beginning their ad campaign because

they're desperate for money so it's

funny you've got the very very disperate

valuations between Meta and Open AI.

Meta makes all their money from

advertising. And now OpenAI is trying to

do the same thing. So you have Meta

failing at AI, but they're going to try

again and again and again. Then you have

OpenAI trying to become Meta also

failing at AI, but trying to become more

of an advertising vehicle. And I don't

know how this is going to end up. It's

not good. Uh winners take most. This is

the way of the world. There's no room

for four winners.

There might be room for two, maybe

three. That's it. Now, back to AI and

LMS. This is how AI's brain power

doubles every four to seven months. A

cool little chart just showing the

speed. And imagine if you're you if your

intelligence could double two or three

times a year. Well, that's exactly

what's happening with AI. The pace, the

rate of change of learning is insane.

So, in the early days of open AI, etc.,

you know, it could think for about 15

seconds. Um, and then in 2025, maybe it

could think for eight hours, and then in

2026, it could think for maybe a day or

a couple of days, and then by 2027,

they'd be able to think for two weeks,

and just do oneshot projects with a ton

of compute power over a very long period

of time to replicate things like web

browsers and Salesforce.com and other

SAS software. We'll get to SAS software

in a minute, too, because that's still

happening as well. But again,

this asteroid that's hitting this earth

is hitting really fast. The rate of

change, I have to use that term again,

it is phenomenal

and also terrifying at the same time.

And there is a thing called AI burnout,

which is a Harvard business study,

Harvard Business Review, and people

thought, oh, well, with AI, it would be

easier to do more. um yeah sort of but

it was actually creating more work and

more work at a much higher pace and this

is a study of about 200 people at a US

workplace and it found that AI had

consistently intensified work increased

stress loads workloads etc and time

management is out the window in fact I

think it was Peter Diamantis from

Moonshots that said I think he's a MIT

or Harvard guy but he said he's never

worked so hard in his life and he's in

his 60s but he's also excited excited to

work on this. So, be careful of the AI

burnout every day there. This is like if

you are inquisitive, you are like a kid

in a candy store and it'll drive you

crazy. Anyway, other crazy stuff is

happening on the AI front. A lot of AI

news this week. One line, one shot

movies.

It's also a moment of silence for

Hollywood.

We don't care about Hollywood. Hollywood

has just done nothing about producing

trash for years. So, it's time. But now

there's software coming out of different

markets in the world where this one's

called Seed Dance. It's like it's funny

cuz there used to be a I Sundance I

think is a movie festival. They called

it Seed Dance. I don't know if it's

related but I think it's kind of funny.

But Seance 2.0 you can do a oneline

prompt and you can get a Hollywood

quality multi-seene video

in seconds for pennies or maybe even for

free. I think the thing you get free

trial is you want to make your own

little movie. But just think about what

that does to Hollywood and all the

screenwriters and all the people that

operate cameras and sound engineers and

lighting and it's shocking shocking

what's happening. And you think of I I

remember years ago there was a a company

that did um they wanted to create a TV

ad and they spent millions and the ad

was garbage. But now, and it was only

like a 30 second clip. Now you can just

create a million-doll ad and make it

better good in seconds for free. Again,

just an example of maybe the abundance,

but also the disruption that's coming. I

have to stress that. Get ready. It's

already here. And finally, this is this

should be actually nearly a meme dour,

but the United Nations

approves a 40 member scientific AI panel

to oversee AI and focus on AI safety

etc. Uh this the panel includes I think

the US was against it but uh the panel I

think it included somebody from Iran and

somebody from Russia uh and all these

so-called AI experts but no they're more

like control experts not AI experts but

yeah the United Nations is involved in

something you know it's not going to be

a good outcome. Anyhow moving on to

Robbo taxi because a lot of developments

there too. First of all, Tesla FSD,

remember FSD in a Tesla is a robot on

wheels and it's some people call it baby

AGI2. It's phenomenal. It's sentient.

It's incredible. Even in very bad

weather like we're having right now,

snow and stuff. Be careful out there.

Anyway, they've racked up 8 billion

miles on FSD supervised. Now, the crazy

thing about that is just seven weeks

ago, because I track this,

they just hit seven billion miles, which

means they're going up a billion miles

every 7 weeks. But the pace at which FSD

is being used is accelerating because it

gets better and better every day with

every patch, every update that gets, you

know, over the year updated to your car.

So, by summer, we were expecting to 10

billion. Now, it'll be, I think, May

16th, so real soon. And the reason that

10 billion FSD or 10 billion miles of

learning etc is important because Tesla

learns from that of course and they

learn from non-FSD miles too because got

shadow mode operating the cars. But the

point is Elon did say once they get to

10 billion that's it could be when we

launch unsupervised which is actually

what's happening anyway with robo taxis

in Austin. They're unsupervised. So this

is a big milestone and we're going to

get there by May 16th which is call it

12 weeks away. Wild. Wild. wild. Anyway,

FSD as well is now, this is a big safety

report published by

Tesla is now eight times safer than

humans. Um, and if you look at the

statistics, basically with FSC, your

probability of a crash of a million

miles is 0.19 crashes per million miles.

If you are a human, you'll crash twice

of a million miles. But, you know, and

that means serious crash when police are

involved, insurance companies, people

are injured and stuff. And a lot of

people prank their cars all the time. So

you don't they're not even reported. But

anyway, FSD is safe. Um so if you have a

loved one, you want them to be safe,

get them FSD. That's FSD advice right

there. And also FCC, which is the

Federal Communications Council or

something like that. They granted Tesla

a an allowance to use ultra wideband

radio technology because that's what

they use for the wireless charging

system that will be used for the cyber

cab. Woohoo. You probably ask yourself,

"What the hell is charging got to do

with the FCC?" Well, it's a radio wave.

It's a signal, very low power, high

power charging, but very low power

signal. And it only turns on when the

car is parked above it. And it only

works at very short range, so it doesn't

interfere with other signals. And that's

why it's allowed. So, they just got that

approved. But speaking of charging,

again, Uber, this is funny. This this is

what I usually call the fingers in the

pie segment. Shout out to Jeff Lutz. The

Uber Cope continues.

Uber is going to invest $100 million in

autonomous vehicle charging amid their

robo taxi push. Remember Uber don't have

a robo taxi. Their partner is a charging

company called EVgo. If you ever go to

an EVGO, it probably doesn't work. The

terrible, terrible chargers. But they

have 20 partners. Not only partners for

charging partners with Whimo and who

knows Zuks and who else we ride car

makers partners partners partners

partners partners partners partners they

forget the magic of succeeding in the

age of AGI and that is complete vertical

integration that's it if you're not

vertically integrated you're not going

to make it this is all in my opinion a

waste and they might be able to make

some money for it for a few years but

once these puppies start getting spat

out of the Austin Texas garage

factories, not garage. It'll be

mind-blowing. Anyway, let's get back to

even bigger than autonomous

transportation, humanoid robots. Uh we

saw this tweet from Elon Musk

uh a couple of days ago and he said,

"Humanoids,

it will begin to transform things in

2027, be obvious in 2028, and have a

massive impact by 2029."

Massive impact. By the way, thank you

once again to the mods and chats and

Martell DBF and Nick M. Appreciate you

guys. But again, it takes time to ramp

these things. Fremont factory is going

to be able to get to a million units,

humanoid robots, and then Texas will get

to 10 million. But it'll be a while

before they actually start impacting

your life. They will be impacting

Tesla's life first, then Tesla partners

like SpaceX, like suppliers second. Then

maybe you might get one in your home.

We'll talk more about that in a minute,

too. But it's going to be crazy. And a

big surprise, by the way, that I forgot

to mention is Elon did say this week

that cyber cows could be purchased for

under $30,000. And everyone know how do

I get some and how many do I need to

retire? We'll get to that as well in

another video at another time. And also,

this is goes back to the importance of

humanoid robots. Midjourney's founder

calculates that five million humanoids

could build Manhattan in 6 months.

6 months. Think about that. 6 months.

Now, imagine what the world looks like

when we have 10 billion of them by 2045

or 2040 or 2035, somewhere in that time

frame. It could be crazy. Ela Mus did

predict that there would be one humanoid

for every human on Earth. But there's

going to be a lot more different types

of robots as well out there. Now, of

course, people say, I can hear the

contractor saying, "No, you can't

rebuild Manhattan in 6 months because

the roadways and the materials and all

that type of stuff." I know, but anyway,

midjourney quote. Very important. And

remember, it's all about the hands.

Never forget the hands. There's three

things that matter with a robot. the

ability to manufacture just like rebuild

Manhattan of course the hands of which I

slipped in the ice and smashed my left

hand today so it's not working and the

brain axi SpaceX Tesla brain they have

the three key components but the most

important thing from this article is

they talk about the 50 actuator position

Tesla Optimus Gen 3 hands reveal 50

actuator position leap which is so

important and the Reason that's

important in this little table here is

again it is designed around the human

blueprint for exactly this reason. The

forearm power, the wrist transmission,

maximum strength, hand etc. Uh this is

very very important because all the

other robot companies

built a robot with a pretty little hand.

But if you look at the human autonomy,

anatomy, not autonomy, getting my

anatomy and my autonomy all confused,

all the power in your hand is in your

forearm. Very important. And this is why

it's key. I'll get to the point I'm

trying to make here. But there'll be 25

actuators per hand, 50 in total. And the

engineering complexity is massive. In

fact, I think it was Kathy Wood said the

humanoid robot is like hundreds of

thousands of times, 200,000 times more

complex than an autonomous vehicle,

which is in itself extremely difficult

to do. But the reason I talk about the

hand is because a lot of people are very

excited about, oh, everybody has robots.

Well, this is figure AI. Now, if you

look carefully, that's the CEO of Figure

AI. If you look carefully at the hand on

this thing, there is no connection to

the wrist. Okay? And the power belongs

in the wrist and the forearm. Okay, the

hand is the business end that needs to

be light and precise and the strength

has to be the physics and the biology

etc has to be in the forearm. And this

image perfectly spotlights in my opinion

why they are focusing on robots to empty

dishwashers in the home. It's a retail

solution. It even says it on their

website. It says, let me see if I can

read that up. The future of home help is

here. So they're skipping the entire

need to do commercial robots with more

domestic robots, which is very clever of

them to be focused. So not knocking them

at all, but it just shows you the

importance of all these little robot

companies and you see all the Chinese

robots, etc. Same deal. Hand is a hand

when really or this in this case it's

just a like one of those Lego. You see

the little Lego things and the hand is

just like a little round thing. Well,

that's what they have. They got Lego

hands. They're never going to be able to

perform surgery or work in a factory and

stuff. However, these are the Chinese

human, humanoids, and they did a spring

festival thing, New Year, Year of the

Horse, whatever, from Uni Tree, etc. And

they were very impressive in terms of

their comedy skits and their ability to

dance and move and do flips and karate

and everything else. So, these things

could be extremely good war machines. In

fact, one of our teammates here was

worried. He said, "Could those robots

working on this today?" She's like,

"Could those robots go to war?" And I

said, "Yes, unfortunately, yes." Anyway,

on to the Mag 7. A lot of ground to

cover, but really done. Hang on. Music,

not just movies. Music

or else is also being in it doesn't

matter because no good music has been

produced in my opinion since the 80s or

'9s, you know. That's me. Yeah, I kind

of like the older stuff, even the 70s.

Um, but so we're not missing much. But

the point is a huge percentage of music

today is already built by AI. In fact,

somewhere some Spotify or something

else, 40% of new track uploads are AI

generated. But this is from LIA 3 from

Google Gemini. And now you can just

create music and again one shot. You can

write a sentence, you know, or mash up

an album or whatever. Uh, and you can

generate 30 tracks from a text prompt or

even an image. You can now build a song

around an image. Say, "Make me a love

song around this picture of this

beautiful flower with this beautiful

girl or boy, whatever you want in the

countryside." Boom. It's done. And it'll

do it in eight languages. Hindi,

Spanish, English, Japanese, German, I

think, Spanish, French, I'm not sure.

Anyway, if you like music and you want

to you're looking for good music because

you haven't got any in decades, go build

your own. There you go from a picture.

And then the SAS apocalypse continues.

People are worried, why is the market so

sucky? Why is everything down? Well,

when the market falls down, everything

falls down. But there is a trillion

dollar plus selloff in software and data

stocks. And it continues. This is the

software index compared to the QQQ. And

basically, it's over the last 5 years.

It's down 43% over the last 5 years, but

nearly 25% of that dip happened just

this year. It been absolutely trillions

of dollars boom eviscerated because of

anthropic and AI. So if you think it's

not affecting things or it's a bubble,

that's another data point where it's

not. And Microsoft, shout out to

Microsoft. This is a chart of Microsoft

over the last six years and it has

actually underperformed

the S&P 500 if you look at those two

data points. I know you could always

cherrypick data points, but the point is

it hasn't gone nowhere for a long time

and it did have some very nice big

spikes way back in what was that? July

2024. Another nice spike in July 2025.

Interesting pattern there. Let's go back

to July 2022.

There's a little pop there, too. July

2023. Yeah. So, funny. For some reason,

Microsoft likes to pop in July. I just

noticed that right now. So, maybe

there's a little trade there. Get into

it maybe March, April, wait till July.

Sell it. Get out before August. Anyhow,

moving on. Apple, we're in the Mag 7

category still. And these these guys are

going all in on wearables.

They've outsourced their AI to, I don't

know, Google.

uh to do their Siri, which never worked

for years. Although it you can say, "Hey

Siri, remind me to let the dog out in an

hour." It'll do that. Kind of like the

Alexa thing. Alexa is even better than

Siri, but the Google Assistant is the

best in my opinion from the for the

home. But now they're going all in on

wearables.

And this means, imagine walking around

with a little necklace that has a camera

on it spying at everybody. Imagine if

you saw somebody wearing one of those,

you just run a million miles away. don't

want to be on video. Okay, first of all,

glasses with them. We've seen that

before. Matt has been doing that for

years. And now I think they're putting

cameras in the AirPods, too, so you can

still

in a stealthy way spy on people. James

Bond would be proud. Anyway, moving on

to space. Let's see what's going on in

the space world. Not a lot of space news

this week after the crazy space news

over the last 6 weeks, but this is a

tech crunch deep dive and they're

looking at the economics of orbital AI

are brutal and they estimate it would

take about 42 billion of investment just

to put a gigawatt of data center

capability in space. A gigawatt of

orbital data center again would be at

least three to four times the

terrestrial equivalent.

Difficult. However, Elon Musk and SpaceX

have proven that they can put 10,000

satellites in space that have no

electricity except for solar power and

they can handle all the radiation

shielding and thermal management and the

vacuum etc. So Elon seems to think it's

possible. I think it is possible but I

don't think it's coming next year or the

year after or the year after that. I

think it's four to 5 years away, maybe

even later and by then it'll be cheaper

to do. But it is an important thing that

that everybody thinks, oh, we're going

to we don't need to build data centers

on Earth anymore. We're going to put

them all in space. It's years out. And

then the other impact of AI is good

impact is longevity and cures for

illnesses and other stuff like that. But

a fun piece of news is there's some guy,

his name is Brian Johnson. He said he

tweeted, he said, "We're launching our

immortals program. It's $1 million per

year and he has only three spots.

available. H interesting again

where they can I guess test your blood

and make you do some exercises but you

can do that all yourself and save a

million dollars and invest it in the

markets in winning assets. Remember

being in the winning assets that was

kind of funny but there's a guy called

Jeff Jeff Tang who's hilarious. He uh he

mocked him. I just signed up for

Blueprint Immortal for $1 million a

year. Because I use CLA, I can make one

payment per month for the next thousand

years interest free. Again, he's mocking

that what is it? Buy now pay later

stuff, which is very dangerous. Every

don't do that. If you can't afford

something, don't buy it. Do not put it

on the card or do not put on the CLA or

the buy now pay later platform. But the

other hilarious thing is that in that

time I'll be able to launch millions of

OpenClaw rappers. Each one making 10,000

a month and then I can exit for 2

million after just 24 hours

and still have lots of money without

putting anything down on the Brian

Johnson program. I just thought that was

very funny. Anyway, speaking of very

funny meme, shout out to Ocean JJ who

creates the best memes in the world. He

said, "In the year of the horse, be on

the fastest horses." This is investment

advice, but not necessarily to buy

Tesla, but very important. The world is

changing very, very, very, very quickly,

very quickly. You got to be on the

bleeding edge. Do not be in something

legacy that will be disrupted and wiped

out. And do not be in loser companies.

Only be in the winning company. Stay

smart, my friends. And thank you as well

to Adam, Tesla, and Anjo and Piper. The

whole gang is here today. Sean D, TND,

Cipher, Sniper, KH, and everybody else.

You're all amazing. I hope you enjoyed

that AI impact. And remember, I do want

to stress one thing. AI is terrifying.

But in Switzerland, we used to say, hept

the which means grip the worry, grip the

fear, grip it. Don't be afraid of it.

Lean into it. Embrace it. Use it. That's

how you survive. Thanks all. Good night.