In the Loop AI

SpaceX just caught a spaceship out of the sky. Ship 42 launched from Starbase, circled the Earth 14 times, deployed 26 Starlink V3 satellites and came back 22.5 hours later, where the same tower it launched from caught it in mid-air, the first time a rocket's upper stage has ever been caught after reaching orbit. Ejaaz walks through the flight, from the lost booster to the heat-shield cameras to the 10-second deorbit burn, and why the catch matters.

We also cover why this isn't just a story about one rocket: what cheaper launches mean for Starlink, StarMind's AI satellites and SpaceX's compute business, the Amazon Leo launch that failed the same weekend, how far Starship has come since the 2019 Starhopper, and Elon's bigger plan for robots, energy and resources.

==
FOLLOW EJAAZ ON TWITTER:
https://x.com/cryptopunk7213

==
Timestamps
00:00 SpaceX Caught a Spaceship
01:39 Why This Is a First
02:18 Why Elon Is Doing This
03:10 Liftoff to Orbit
03:58 26 Starlinks and a Heat Shield Check
05:41 10x a Falcon 9
06:10 The Fall and the Catch
07:11 From Starhopper to Starship
07:44 Amazon's Rocket Fails
08:26 Built to Fly Again
10:01 Why It Matters for AI
11:11 StarMind: AI in Orbit
12:13 The Starlink Monopoly
13:25 SpaceX's AI Business
14:56 Space Resources and Robots
16:21 Final Thoughts

#spacex #starship #elonmusk #ai #starlink

==
AlphaSpace by Yahoo Finance: A Professional-Grade Investment Platform Built for Everyday Investors.
👉 https://finance.yahoo.com/about/promos/gold/alphaspace/?ncid=100003571

Creators and Guests

Host
Ejaaz Ahamadeen

What is In the Loop AI?

Exploring the frontiers of Technology and AI

Over the weekend, Elon Musk's SpaceX did something that shocked the entire world.

They made history by launching the largest space rocket ever.

I mean, the top half of this thing is like a building that's 17 stories high,

but that's not even the most impressive part.

You see, after this rocket orbited the Earth 14 times, 22 hours later,

upon landing back on Earth, it was caught mid-air by two giant metal mechanical

pincers called Mechzilla.

This is the equivalent of catching 300,000 pounds of metal and pure rocket fuel in mid-air.

And the craziest part is it landed on the exact same tower that it was launched

from. This is a first for a lot of different things.

And I'm so excited that space travel is no longer this bleak,

weird Star Trek-like fantasy.

It's something that can happen as soon as within our lifetime,

certainly, but also maybe within the next couple of years.

You see, Elon Musk and SpaceX has brought the cost of spaceflight.

Down massively, which means that it can open up new opportunities for AI,

for spaceflight in general, for acquiring resources outside of Earth,

or launch satellites that can beam data connectivity for anywhere around the world.

So in today's episode, I want to unpack why this is such a big deal,

not just for spaceflight, but for AI and for Elon Musk's goal and entire vision

of combining SpaceX with the likes of Tesla and automating things via robotics, space, and AI.

You see, this isn't just a story about a single rocket.

It's about everything else that Elon's trying to do. So let's get into today's episode.

Now first some quick context as to why this is such an impressive event because

elon musk and spacex has caught parts of a rocket ship before but that's mainly

been the booster the bottom half of a ship

they've never actually caught the ship before they've had 15 attempts and 10

of these missions have been successful but this is the first time that they

tried to catch the actual ship

and they nailed it the second cool part about this is it was only their second

orbital flight And this thing flew around the Earth 14 times, roughly,

I think, once every 60 to 90 minutes.

So really quick. And the total distance that it traveled was around one and

a half times the distance between the Earth and the moon. So this thing has durability.

And the third bit of context is, why is Elon even doing this in the first place?

Well, his entire vision around space is that it's incredibly important for humans

to become an extraterrestrial life form. the ability to fly through space,

investigate what's going on around us, acquire resources around us.

That is the next step in civilization, according to his vision for SpaceX.

And so what he's trying to do is the first stage, which is make space travel

as cheap and accessible as possible. And prior to this.

It has kind of been a very tough journey when you compare it to the likes of

NASA or any other space rocket company.

They have faced numerous amounts of challenges, whether that's regulatory or

mainly mechanical engineering, and they've invested hundreds of billions of dollars to do so.

Elon Musk came through with SpaceX and figured out a much more novel,

efficient, and innovative way to pull this off. And back then,

I think 10 years ago, he was deemed as a crazy individual to be able to pull this off.

And now we see it in reality. So just to run you through the sequence of events,

this rocket launched around 6 p.m. on Friday, and it took about six and a half

minutes to get up and out of the atmosphere.

Now, at this point, the booster, which is the bottom half of the spaceship, detaches.

And the idea here is, and was demonstrated on many launches before,

was that the booster would land by itself, that it would be kept intact.

And the idea here is replaceability, cheaper space travel, except it failed

to land this time because it was low on fuel. But the actual ship itself continued

out into orbit, and it orbited the Earth around 14 times. I

once every 90 minutes, which is just an insane speed. I think at its top speed,

it was at 17,000 miles per hour. Now, once it was in orbit, the ship itself was on a mission.

It deployed 26 new Starlink satellites.

Now, Starlink is SpaceX's entire business unit for putting out high data internet

connectivity into space. The idea being wherever you are on Earth,

you should have access to high-speed internet.

And so these satellites beam down data towards Earth so that you can pick it

up on your phone or any device and be connected to the internet massively.

And we're seeing the business grow pretty massively on Earth.

SpaceX's entire business revenue originally started to come from Starlink itself

in the billions of dollars worth of revenue and has recently shifted in favor

of AI as SpaceX has pivoted more towards the neocloud, but it's still a very important unit.

So 26 Starlink satellites were deployed.

Now, there's a few cool Easter eggs about this particular deployment.

One, the last four satellites that were deployed had cameras on it to inspect

the heat shield of the rocket ship that it was leaving behind.

So whilst it was being deployed, it had cameras that were looking at the heat

ship, which is the external surface of the rocket ship, so that engineers back

down on Earth could determine whether the ship.

Was suitable enough to return back into Earth for the recapture by Mechzilla,

these mechanical metal chopsticks that were meant to catch it,

to make sure that it was good and safe upon re-entry.

So what I like about this is that there's so much thought that's gone into this process.

I think in previous space launches, This just wasn't a thing,

certainly in the NASA era.

And so to have this amount of tech, thought, and durability go into the entire

sequence of events to make sure, you know,

to push into this mission of reusability for space, like it can be cheap enough

for anyone and everyone to do so, and we should have space launches going every single hour.

That, I thought, was a really cool nugget. Now, the second more practical thing

is that in this one deployment of Starlink satellites, it was around 10x the

capacity of any of the other previous spaceship launches from SpaceX.

You see, Falcon 9, which is the predecessor to this starship,

only had the capacity to do one-tenth of these Starlink satellite launches.

So the fact that we've made that 10x jump in such a short period of time,

it's only been just over a year, maybe two years, is so, so impressive.

So the duration of this entire thing was around 22 and a half hours.

And upon re-entry into Earth, which I think took around 22 minutes,

it only burned around 10 seconds of fuel.

And so it was just free falling. Like, imagine a 17-story tower or apartment

building free-falling through the sky, hurtling down towards Earth,

and then 10 seconds of rocket fuel propped it upright and placed it perfectly

within the pincers of a mechanical maxilla that closed around it and caught it.

Now, this thing weighs, my estimates, my interactions with Grok put it at around

200,000 to 300,000 pounds of weight.

Imagine catching that in midair. And of course, this is Elon Musk we're talking

about. The time that it was caught was 4.20 p.m.

Eastern time. Now, Elon Musk put this entire mission success rate at around

50 to 60 percent, which when I first read that, I thought was incredibly high.

But I guess he has a lot of confidence in the tech and what he's built. Makes sense.

He's built this for over a decade. And I guess the reason why I feel this way

is that I remember the first Starhopper launch, which was back I think in 2019

like just under a decade ago,

this thing looked like a giant metal tin can then I compare that to what launched

on Friday which was this sleek,

huge like double.

Like 15 to 30-story building-type structure that was then caught out of mid-air.

And I think, how have we progressed so far? And then I think,

well, what is this going to look like a decade from now?

And I want to emphasize how hard a thing this is to pull off in the first place.

In fact, there was a rocket launch that I think happened around a similar time

over the last couple of days, which was carrying Amazon's own space satellites.

They have a competing business line with Starlink, or they're at least trying to compete.

And that space launch did not work at all. In fact, it just burned up.

And so you wasted away all those different satellites. And you have real business

lines that are reliant on these types of satellites.

I think Delta Airlines had a deal with those satellites that were meant to be

deployed from Amazon, and it didn't end up making it out.

So we're not in an era of novel experiments anymore. We're at a point where

space travel is becoming a lot more feasible, and businesses that aren't necessarily

space-affiliated are starting to depend on it.

Now, one of the coolest videos from this entire thing was Elon Musk at the top

of the spaceship after he had landed, inspecting the heat shields.

And his prognosis after inspection was that, yeah, it needs about a couple of

weeks worth of refurbishment. But after that, it should be good to go back out again.

Think about that for a second. In the NASA era, we would spend tens to hundreds

of millions of dollars building a ship that would then have a 50,

maybe less than 50% chance that it was going to launch successfully or burn up entirely.

And then it was a given that that entire thing was written off,

that that spaceship would never be used again, that you would need to invest

another $100 million to build the thing and launch it again.

Fast forward to today with SpaceX, and we've done a whole entire space launch,

which is the largest of its structural kind that was then caught back on the

same tower that it was released on

and that will should technically be ready to go theoretically a couple weeks from now

then you might be wondering what about the rest of the ship well upon inspection

by the team apparently they don't even need to pay that much attention to it

it's just standard procedure to maintain the standard maintenance routine that

they've had even before the ship even launched and it should be good to go.

So when I think about the challenge of reusability in space that seemed nigh on impossible.

About five to 10 years ago, fast forward to where we are today,

it is very feasible that within the next couple of years, we'll see a rocket

of this standard certainly much better launch every single hour or couple of

hours, which is just insane to think about.

Now, putting all the impressive launch stuff aside, I want to talk about why

this technology is so important.

And it's not just necessarily about space. It's about AI, data connectivity,

and the entire big vision that Elon Musk is trying to execute upon.

Number one, the cost of putting things out into space.

You see, this spaceship launch was a milestone event because it was the cheapest it's ever been.

I think it was around $2,700 per kilogram. So roughly for if you're in America,

a 50 pound suitcase piece of luggage that you're taking on holiday would cost you around $62,000.

Very expensive, but trust me, it used to be way, way more expensive.

Now, Elon's entire goal here is to bring that cost down to around a couple So

$230 for your piece of luggage, which again, is still expensive.

But if you consider taking it out into space, that's not a bad deal for now.

And the idea behind this payload is to put useful things out into space.

Now, the main example of this has been satellites, satellites specifically belonging

to Starlink that would beam down high-speed internet data connectivity to Earth.

But of course, Elon Musk is also trying to do other types of satellites.

For example, AI training and inference satellites, known as this network called

StarMind, which will create a constellation around the Earth and around the

Sun, which will harness the energy from the Sun and use that energy to train

and inference AI models.

Think about it. The biggest bottleneck that is primarily on Earth for a reason

is because resources such as electricity is very scarce. Power generation is incredibly scarce.

And so in order to acquire those assets through the regulatory tape.

Pushes prices up to train models.

That's why the CapEx boom is costing hundreds of billions of dollars.

Imagine you had the same training infrastructure out in space,

basically space GPUs that could harness the energy of the sun.

You'd end up having the lowest, cheapest cost of energy to be able to train in inference models.

Over the long term, it seems much more feasible. But the biggest barrier to

this is actually getting the payloads out into space.

That's why Elon's number one

step here is to make space travel or space transport as cheap as possible.

Now, it's important to understand that Elon, or SpaceX specifically,

has the entire monopoly for this thing.

Now, right now, I think SpaceX has around 11,000 satellites out there.

You compare that to their nearest competitor, which is Jeff Bezos' Amazon Leo,

and they have around 400 satellites.

And I remember a few months ago talking about Elon Musk's intention,

I think he filed with the government to send out another million to a million

and a half satellites over the next couple of years.

And because SpaceX is the most frequently flying starship or space travel company,

it makes sense that he will reach that milestone much sooner than Jeff Bezos

or his number two competitor Blue Origin.

Now, if he's able to deliver on both Starlink and Starmind, Elon Musk will end

up having a monopoly on two very different industries. One, fighting with the

telecoms and data connectivity companies back on Earth.

AT&T, Verizon, both companies which I've used, which I've had arguably terrible

experiences, I would love to just be able to connect to Starlink directly wherever

I am, even if I'm on a flight or in some remote part of the world and have

strong internet quality to be able to do whatever I need to do.

It gives people a lot more freedom, a basic necessary right.

The other industry being AI.

Now, look at what Elon has done on Earth so far. He has been able to scale AI

compute quicker than anyone else. He has the largest cluster of GPUs,

specifically NVIDIA GPUs, the most powerful kind.

And he's been able to direct that compute both towards training his own models,

as well as selling it to other frontier labs. And he charges a very high premium.

I think he's charging Anthropic around 1.5 to $3 billion per month to get access to his GPUs.

It's still a constraint. Can you imagine what the AI landscape looks like if

he's able to acquire all the compute for a much, much lower fraction of the cost out in space?

That's another big deal. And so the vision there is that there'll always be,

an ever-growing exponential demand for compute if AI succeeds on cannibalizing

or growing into every single vertical industry.

So even if you assume that SpaceX just stays as a neocloud provider,

so an AI compute provider, which, in my opinion, isn't as self-fulfilling of

an ambition as what Elon Musk has been going after.

It's still a very lucrative business. You see what Amazon has built with AWS,

you see what Google's built with Google Cloud, Microsoft, Microsoft Azure.

These business lines for them are some of the most highly profitable industries,

especially over the last couple of years as they're selling more and more compute.

If Elon Musk is able to become the number one provider there,

that could end up being a very consistent flow of revenue for SpaceX full stop.

And we've seen it on the last quarterly earnings where it's turned them almost profitable.

And the third business line, I guess, is space travel. And remember,

space travel doesn't necessarily need to be for leisure, although I hope and

pray in my lifetime that I get to go visit the moon or go check out Mars.

I might be waiting for a bit more, but I hope it actually happens.

But the other side of things is space travel for resources. There are resources that are out there,

chunks of gold flying around orbit that would be cool to harness or use for

things back on earth or maybe even an extraterrestrial life.

Elon Musk has made it very clear that he wants to set up a human settlement.

He's talked about hyperdrivers on the moon. It's just all crazy stuff that seems

phantasical until it's not.

And think about it. If Elon is somehow able to pull off his very ambitious vision of creating

robots that scour earth, that outnumber humans, humanoid robots that are his

optimist kind, which are meant to be revealing over the next 12 to 18 months

for mass scale deployment in factories to start off with and then with humans in our homes alone.

If he's continuing to put out Tesla vehicles, whether it's cars or some other

type of machine or computer that is able to kind of ingest data 24-7,

move around autonomously, if you have all these other robotic setups that are going to exist,

you're going to need energy.

You're going to need compute. You're going to need data connectivity.

And you're gonna need materials and resources.

Earth itself may not be able to quench or satisfy. And listen,

I hear myself speaking, and I know that sounds incredibly bullish.

And there are a lot of conditions that could go very wrong that may not allow

this to happen, certainly over the next couple of years, or maybe even over the next decade.

But if you look at Elon's track record and his ability to scale at least physical

atoms, maybe not all the AI stuff for AI models, but at least physical atoms

from zero to one, there's no other better person that's been able to pull it off.

But yeah, I guess that's the recap of what I thought was one of the most impressive

things to watch on my TV at home over the weekend. And if you didn't catch it,

I recommend there's a million clips out there.

There's the entire live stream out there. Just go watch the launch and the recapture at the very least.

It is one of the most, I promise you, one of the most impressive things you've

ever seen, certainly over the last couple of months.

And we're talking about the age of AI over here. So yeah, I hope you enjoyed this episode.

Thank you so much for listening. If you're listening to this on YouTube,

Spotify, or Apple, and you aren't subscribed, or you aren't following us,

or you haven't given us a rating, please do so. It helps us out massively.

Also, please continue to leave us comments, or you can also DM me on X,

my profile is linked below.

And yeah, share this with a friend if you think this would be interesting,

because I certainly find a lot of Frontier Tech interesting.

I know I talk a lot about AI on the show, but there's just so much going on.

I want to try and cover as much as I can.

And we're here five times a week, Monday to Friday. So yeah,

I guess I'll see you on the next one. Thank you so much for listening.