SpaceX potential IPO

[meekGee]

The only way I see this would work is if Musk had the controlling shares, like 51% of all stock.  That way he could keep the company focused on Mars and Xai. 

Even the xAI thing I think is a problematic distraction.

Indeed, it's certainly a greater leap of faith than Starlink was, but in terms of "distraction" Starlink is worth its weight, so maybe this one will too - if it's successful.

Potentially,  but I don't think it's entirely parallel, for two reasons:
- Starlink had a very clear lead over other LEO satellite internet (OneWeb planned a much smaller constellation, Kuiper started much later). The competition seems more even for AI.

- unlike Internet, a lot of AI services currently aren't really being charged at real cost, and it's not clear much of the current demand would exist at prices sufficient to make up real costs.

On the positive side, though, Artemis needs HLS Starship, and Starlink v3 and any space data center projects need Starship launch. So both Starship in general and some version of crew Starship should be "safe" even if an IPO limits Musk's control of the company eventually, so Mars in some form should still be possible. (Hopefully this wouldn't happen until Starship is more proven, anyway.)

Startlink v3 services a market that's known by now, thanks to Starlink v1 and v2, and it has no real competition.

Orbital AI is a great unknown, and pays off only in very large scale. But at that scale it is clearly larger than Starlink. It might eclipse all other compute and possibly power. It's not every day that a transformational technology like this emerges.  (If it succeeds)

If it  succeeds, however, in the race for power between terrestrial solar, fusion, and orbital, Musk jas a good chance to win on timing.

[meekGee]

The only way I see this would work is if Musk had the controlling shares, like 51% of all stock.  That way he could keep the company focused on Mars and Xai.  I would think if he is going to build the Xai servers in orbit, he will need a fleet of Starships to build it as well as go to Mars and the moon for NASA. 

I would also believe that Musk thinks he will have Starship operational by the end of the 2026, and have another million or so customers for Starlink services.  This would make SpaceX worth more. 

I would also wonder if Bezos would do the same for Blue Origin once New Glenn 9x4 is operational.

Even with the 9x4, the entire rocket is only equivalent to a Starship upper stage (SS) and as a system (considering launch rate per tower and number of towers) does not have even 1% of its capacity.  It can't compete head to head with SH.

[Vultur]

Orbital AI is a great unknown, and pays off only in very large scale. But at that scale it is clearly larger than Starlink. It might eclipse all other compute and possibly power. It's not every day that a transformational technology like this emerges.  (If it succeeds)

If it  succeeds, however, in the race for power between terrestrial solar, fusion, and orbital, Musk jas a good chance to win on timing.

Yes, my concern is what happens if it *doesn't* succeed. (Which I think is the far more likely outcome, if by "succeed" we mean that exceptionally large scale.)

[meekGee]

Orbital AI is a great unknown, and pays off only in very large scale. But at that scale it is clearly larger than Starlink. It might eclipse all other compute and possibly power. It's not every day that a transformational technology like this emerges.  (If it succeeds)

If it  succeeds, however, in the race for power between terrestrial solar, fusion, and orbital, Musk jas a good chance to win on timing.

Yes, my concern is what happens if it *doesn't* succeed. (Which I think is the far more likely outcome, if by "succeed" we mean that exceptionally large scale.)

For xAI to fail, it needs to lose out to openAI, gAI, or some unknown upstart.

For orbital data centers to fail, the entire AI sector's power crisis must prove to have been a false alarm.

For Starship/SpaceX to fail, all of orbitalAI, Starlink, and p2p must be duds.

I think it's a safe a position as you can expect from a transformative venture.

[Vultur]

For xAI to fail, it needs to lose out to openAI, gAInor some unknown upstart.

For orbital data centers to fail, the entire AI sector's power crisis must prove to have been a false alarm.

I think the first is quite possible.

But the second is more or less what I expect, yeah. Or, well, no, it's not a false alarm *today*, it's clearly happening now. What I expect is that exponential growth in power demand will not last very long. 

Because I think that once *full* costs for AI have to be paid for by consumers, demand will drop, since many consumers either simply won't be able to pay those costs (individuals or some small businesses) or will find it cost-ineffective vs offshoring labor (for larger businesses). Unless those costs decrease sharply.

 Right now people are using ChatGPT for free, Google is putting AI Overviews on every search even for people who don't want them, Microsoft is pushing AI copilot very hard (though if the recent reports of reduced sales quotas are true that might be changing?), etc. that all uses energy and depreciation of hardware that has to be paid for by *someone*. Right now investment money is flowing in like a firehose, but that's unlikely to last indefinitely.

That doesn't mean AI will "go away" no. But I do think it means that *either* it will move to technologies that use vastly less energy and specialized expensive hardware *or* usage will become largely limited to those who can actually pay the costs for that energy + expensive hardware.

Neither of those scenarios implies exponential growth of power use will continue very long.  And if the first (much more efficient technology) happens, much of the current hardware investment may be bypassed.

Fir Starship/SpaceX to fail, all of orbitalAI, Starlink, and p2p must be duds.

Oh, I agree the Starship program itself is pretty safe - Starlink is clearly already not a dud and there's the Artemis HLS contract. And if v3 early flights go well, a late 2026 IPO would probably be after it's pretty well demonstrated (if they don't go well, any IPO would likely be delayed).

What I'm less sure about is Starship Mars efforts.

[meekGee]

What I'm less sure about is Starship Mars efforts.

I'm pretty sure the beginning of the manned effort won't require more investment: 2026 (pilot maybe), 2028 (robotic general rehearsal), 2030 (manned first crew if we're lucky) and 2032.

I mean, the factories and pads are already getting built, and ship production rate implies they can't be THAT expensive.

[Vultur]

What I'm less sure about is Starship Mars efforts.

I'm pretty sure the beginning of the manned effort won't require more investment: 2026 (pilot maybe), 2028 (robotic general rehearsal), 2030 (manned first crew if we're lucky) and 2032.

I mean, the factories and pads are already getting built, and ship production rate implies they can't be THAT expensive.

Yeah, I'm not concerned about the Starships themselves or the pads/GSE, which are needed for Starlink, Artemis, and Mars. My concerns are development of Mars specific stuff like the fuel ISRU plant and habs (which probably would be significantly different from lunar/late Artemis ones)

[Metalskin]

What I'm less sure about is Starship Mars efforts.

I'm pretty sure the beginning of the manned effort won't require more investment: 2026 (pilot maybe), 2028 (robotic general rehearsal), 2030 (manned first crew if we're lucky) and 2032.

I mean, the factories and pads are already getting built, and ship production rate implies they can't be THAT expensive.

Yeah, I'm not concerned about the Starships themselves or the pads/GSE, which are needed for Starlink, Artemis, and Mars. My concerns are development of Mars specific stuff like the fuel ISRU plant and habs (which probably would be significantly different from lunar/late Artemis ones)

Isn't it possible that some of this work is already happening behind closed doors?

[meekGee]

What I'm less sure about is Starship Mars efforts.

I'm pretty sure the beginning of the manned effort won't require more investment: 2026 (pilot maybe), 2028 (robotic general rehearsal), 2030 (manned first crew if we're lucky) and 2032.

I mean, the factories and pads are already getting built, and ship production rate implies they can't be THAT expensive.

Yeah, I'm not concerned about the Starships themselves or the pads/GSE, which are needed for Starlink, Artemis, and Mars. My concerns are development of Mars specific stuff like the fuel ISRU plant and habs (which probably would be significantly different from lunar/late Artemis ones)

How much can that be?  This is not JPL pricing, and it's not some super sensitive science gear that has to work unsupervised and use only 25 Watts.

This will be robust, simple, man-tended equipment.  Mass and power won't be dirty words.  Cheap and reliable.

Otherwise you're down the same rabbit hole that MSR fell into.

[spacenut]

Grok or Xai is by far the best AI right now.  We have used others, but they can steer you in a wrong direction.  Grok is great.  Anyway, I still think Musk is waiting for Starship/Superheavy to succeed next year with the R3's pushing them.  Once it succeeds then more Starlink 3's can be launched as well as AI servers.  This will help pay the way to Mars and the moon.  Musk will probably want more control than he has had with Tesla to keep things going in the right direction.  Hopefully he will live long enough to get a Mars colony started. 

[Vultur]

What I'm less sure about is Starship Mars efforts.

I'm pretty sure the beginning of the manned effort won't require more investment: 2026 (pilot maybe), 2028 (robotic general rehearsal), 2030 (manned first crew if we're lucky) and 2032.

I mean, the factories and pads are already getting built, and ship production rate implies they can't be THAT expensive.

Yeah, I'm not concerned about the Starships themselves or the pads/GSE, which are needed for Starlink, Artemis, and Mars. My concerns are development of Mars specific stuff like the fuel ISRU plant and habs (which probably would be significantly different from lunar/late Artemis ones)

Isn't it possible that some of this work is already happening behind closed doors?

Absolutely possible. Wouldn't surprise me at all. But there's no way it'll be ready for prime time late 2026, so it would still require post -IPO investment (if they really do a late 2026 IPO).

[meekGee]

What I'm less sure about is Starship Mars efforts.

I'm pretty sure the beginning of the manned effort won't require more investment: 2026 (pilot maybe), 2028 (robotic general rehearsal), 2030 (manned first crew if we're lucky) and 2032.

I mean, the factories and pads are already getting built, and ship production rate implies they can't be THAT expensive.

Yeah, I'm not concerned about the Starships themselves or the pads/GSE, which are needed for Starlink, Artemis, and Mars. My concerns are development of Mars specific stuff like the fuel ISRU plant and habs (which probably would be significantly different from lunar/late Artemis ones)

Isn't it possible that some of this work is already happening behind closed doors?

Absolutely possible. Wouldn't surprise me at all. But there's no way it'll be ready for prime time late 2026, so it would still require post -IPO investment (if they really do a late 2026 IPO).

2026 is at best a landing demo.
2028 is the practice run and consumables stash.

The equipment needs to work with high but not absolute confidence when the first crew (2030 or 2032) goes.

This is because the stash should be large enough to survive on for 4 years, and the first crews will bring with them another such stash, so if the ISRU equipment is a full failure, everyone is still plenty safe.

And they'll carry two designs anyway, even just to see what works better.

The combination of lots of mass and power and manned operation will do wonders for the price and yield of the equipment.

[meekGee]

...
What I'm less sure about is Starship Mars efforts.

Let's count towers.  A tower launching 10/day, with 150 payload, 330 days, is 0.5 MTon / yr, or let's say 0.25 to higher orbits.

So very broadly, you need 4 towers for the 1 MTon/year figure.

A much expanded Starlink will keep another tower busy.

Mars, with 1000 ships and 5 refuels is another tower's worth.

Crazy numbers.

But even one tower at 1/day used to be crazy, and after seeing a catch and a reflight it no longer is...  So 5 years from now?

[Tywin]

Yeeesh all the detractors are spinning up scenarios where SpaceX shoots themselves in the foot in order to satisfy them...  Either because SpaceX are too dumb to understand the consequences, or because SpaceX is a big hoax to steal investor money on the false premise or Mars.

We also previously had a wave of "SpaceX can't attempt a 2026 Mars launch because of HLS demo.".

Whatever floats your boat guys.

But...  SpaceX is going for Mars as its main goal, and everything else is subservient to that.

Starlink, HLS, orbitalAI, p2p...

It's just that on the way to Mars, it's gonna eat up a lot of other competing companies, partly BECAUSE it's so focused....

I don't know why you said this...

But for me is a good news, and give a LOT of money to SpaceX...

Your emphasis is about how this prevents (in your opinion) a Mars colony effort, just like your previous posts on the matter.

It's one thing to say "I think a Mars colony is too difficult today".  It's quite another to agitate and celebrate every real and perceived difficulty, which is what you're doing.

Let me ask you this: if SpaceX and Musk are able in fact to both go IPO and use the funds towards a Mars colony, would that be a good thing?

YES!

What are they waiting for?


Because Musk already has hundreds of billions and hasn't even launched a CubeSat for Mars.

Where are all the hundreds of ISRU technologies for Mars and their funding?

[SpaceLizard]

Your emphasis is about how this prevents (in your opinion) a Mars colony effort, just like your previous posts on the matter.

It's one thing to say "I think a Mars colony is too difficult today".  It's quite another to agitate and celebrate every real and perceived difficulty, which is what you're doing.

Let me ask you this: if SpaceX and Musk are able in fact to both go IPO and use the funds towards a Mars colony, would that be a good thing?

YES!

What are they waiting for?


Because Musk already has hundreds of billions and hasn't even launched a CubeSat for Mars.

Where are all the hundreds of ISRU technologies for Mars and their funding?

The same place as Blue's "millions working and living in space" technologies? Unless you want to declare that Blue isn't really working on those?

[Vultur]

What I'm less sure about is Starship Mars efforts.

I'm pretty sure the beginning of the manned effort won't require more investment: 2026 (pilot maybe), 2028 (robotic general rehearsal), 2030 (manned first crew if we're lucky) and 2032.

I mean, the factories and pads are already getting built, and ship production rate implies they can't be THAT expensive.

Yeah, I'm not concerned about the Starships themselves or the pads/GSE, which are needed for Starlink, Artemis, and Mars. My concerns are development of Mars specific stuff like the fuel ISRU plant and habs (which probably would be significantly different from lunar/late Artemis ones)

Isn't it possible that some of this work is already happening behind closed doors?

Absolutely possible. Wouldn't surprise me at all. But there's no way it'll be ready for prime time late 2026, so it would still require post -IPO investment (if they really do a late 2026 IPO).

2026 is at best a landing demo.
2028 is the practice run and consumables stash.

The equipment needs to work with high but not absolute confidence when the first crew (2030 or 2032) goes.

This is because the stash should be large enough to survive on for 4 years, and the first crews will bring with them another such stash, so if the ISRU equipment is a full failure, everyone is still plenty safe.

And they'll carry two designs anyway, even just to see what works better.

The combination of lots of mass and power and manned operation will do wonders for the price and yield of the equipment.

Right. I agree they wouldn't bring the ISRU equipment in 2026.

My concern was about an IPO making SpaceX less free to invest in Mars surface hardware, vs the Starlink/space AI/Artemis stuff that has obvious economic return. If the Mars surface hardware development isn't already paid for at the point of IPO...

OTOH, if (these are made up numbers not to be taken seriously) if Musk could sell say $50 billion of stock and still retain control, then he could put that money into a *nonprofit* "Mars Settlement Foundation" or something which could (at say 3% interest rate) have $1.5B/year to pay SpaceX *as a customer* to do Mars-specific stuff. That would avoid any "stockholder money being wasted on blue-sky Mars efforts" concerns.

[Tywin]

Your emphasis is about how this prevents (in your opinion) a Mars colony effort, just like your previous posts on the matter.

It's one thing to say "I think a Mars colony is too difficult today".  It's quite another to agitate and celebrate every real and perceived difficulty, which is what you're doing.

Let me ask you this: if SpaceX and Musk are able in fact to both go IPO and use the funds towards a Mars colony, would that be a good thing?

YES!

What are they waiting for?


Because Musk already has hundreds of billions and hasn't even launched a CubeSat for Mars.

Where are all the hundreds of ISRU technologies for Mars and their funding?

The same place as Blue's "millions working and living in space" technologies? Unless you want to declare that Blue isn't really working on those?

Blue always said this a generational goal, not in our lives.

[meekGee]

YES!

What are they waiting for?


Because Musk already has hundreds of billions and hasn't even launched a CubeSat for Mars.

Where are all the hundreds of ISRU technologies for Mars and their funding?

LOL the irony is strong today

[spacenut]

The only way I see this would work is if Musk had the controlling shares, like 51% of all stock.  That way he could keep the company focused on Mars and Xai.  I would think if he is going to build the Xai servers in orbit, he will need a fleet of Starships to build it as well as go to Mars and the moon for NASA. 

I would also believe that Musk thinks he will have Starship operational by the end of the 2026, and have another million or so customers for Starlink services.  This would make SpaceX worth more. 

I would also wonder if Bezos would do the same for Blue Origin once New Glenn 9x4 is operational.

Even with the 9x4, the entire rocket is only equivalent to a Starship upper stage (SS) and as a system (considering launch rate per tower and number of towers) does not have even 1% of its capacity.  It can't compete head to head with SH.

Yes, the New Glenn is a direct competitor to Falcon Heavy which has very few launch customers.  Where New Glenn might be competitive is because it is a single stick rocket and Blue can use it for their possible developments like Orbital Reef or lunar aspirations.  Competing for 3rd party launches might make it loose out.  They do have project Leo (Kuiper renamed) to do, but it is way behind Starlink already.  I think Blue spent too much time on New Sheppard, loosing time developing New Glenn which is late to the party. 

[Tywin]

YES!

What are they waiting for?


Because Musk already has hundreds of billions and hasn't even launched a CubeSat for Mars.

Where are all the hundreds of ISRU technologies for Mars and their funding?

LOL the irony is strong today

Unless you think Musk is inmortal, we are waiting for Mars stuff since 2018, and at the moment, no ISRU tech, no even a cubesats for study Mars.

I don't see the "irony".