21

Space Policy Discussion / Re: Jared Isaacman as NASA's 15th administrator

« Last post by yg1968 on Today at 12:54 am »

youtube.com/watch?v=Wv9-p3Ari-A&t=2s

Here is a partial transcript of some of the questions from that interview:
https://aviationweek.com/space/budget-policy-regulation/fast-five-nasa-administrator-jared-isaacman

22

Space Policy Discussion / Re: Jared Isaacman as NASA's 15th administrator

« Last post by yg1968 on Today at 12:47 am »

This change came recently, so it was certainly done at Jared’s direction. Jared is pretty thoughtful. He’s certainly not right about everything, but I don’t think this was just stochastic Trumpian chaos in this instance.

I think that Isaacman mentioned twice that he spoke to Trump about a Mars robotic mission. I am surprised that Trump even cares. Eric Berger wrote about the Mars Telecommunication Orbiter a few weeks ago but I don't know if that is the one.

Launch more missions of science and discovery. The president called me two weeks ago and asked, “What’s the plan for a robotic mission [to Mars] in 2028?” We had a really great conversation about that as well.

https://aviationweek.com/space/budget-policy-regulation/fast-five-nasa-administrator-jared-isaacman

It's at 8 minutes of the video:
https://forum.nasaspaceflight.com/index.php?topic=64097.msg2760037#msg2760037

23

Space Policy Discussion / Re: Jared Isaacman as NASA's 15th administrator

« Last post by yg1968 on Today at 12:42 am »

What did he say were the "advantages" of working in the public sector

I forget his exact words but I think that he meant the various employment benefits that government employees get that contractors may not have.

24

Advanced Concepts / Re: Moving The Cloud to orbit

« Last post by Vultur on Today at 12:21 am »

In the US, about 7% of electricity use is by datacenters, or about 30-35GW. There’s like 100GW planned or in the construction phase as we speak.

I don’t think SpaceX will be near the limit in demand if they actually do build 10-100GW of datacenters in orbit and do so cheaper than terrestrial datacenters

That's true IF there's not a decrease in demand/glut in supply, the way there was for fiber in the early 2000s.

(I am personally quite skeptical that that 100GW currently planned/in construction capacity will all come to fruition.)

I really do think people should try using the most advanced, multi-agent models to get a feel for what AI can do, ie for software development or whatever.

Capability isn't enough by itself though. You need profitability . And that's a lot harder to establish.

Developers are paying money for this stuff.

Granted. But how much money? Seriously, is it publicly known how many subscriptions to e.g. Claude Code there actually are, at what individual price?

I do not think anybody doubts there is revenue. There is however a lot of doubt that the revenue is enough to be profitable given the very high costs.

Again, you should actually just try this.

I don't see how that would help answer the profitability question...

It’s already deeply integrated into Google, especially the core search functionality. Google (Alphabet) has $400 billion annual revenue. Look it up yourself!

Sure, Google can pay for a lot of data centers! But are they actually making money on AI, or are they just making enough money elsewhere to absorb the losses?

Google's revenue is largely ads. Meta even more heavily so. Tech companies dependent on ads can only grow so much relative to the rest of the economy or there just won't be enough new advertising money to grab.

--

This issue actually has very little to do with scams or the actual potential of the technology IMO. There is very often a major bubble with new technologies, even ones that do turn out to be genuinely transformative.

AI can be a bubble and have great potential; they're not at all mutually exclusive.

But the problem with orbital data centers as a project now is that the potential bubble environment exists now.

This could be a Teledesic vs Starlink situation, where current efforts fail but someone else succeeds in 20 years.

But developer tools are the one market segment with the highest adoption rate, and the clearest ROI. Other market segments are not as easy to both adopt and profit from.

However there is LOTS of experimentation going on, which is where A.I. companies are seeing some of their revenue, on experimentation.

Exactly. Software development alone isn't enough to justify a multiterawatt or whatever AI space cloud.

Much current AI use by companies is experiments which aren't succeeding. An enormous amount of current AI use by individuals is free use which is only possible because of massive investment money subsidization.

It is hard to see a scenario where after the investment rush subsides and real costs have to be charged, AI will be as widely used as it is today. As of a few months ago 95% of ChatGPT users were free.

But I agree that current A.I. is here to stay,

Here to stay - yes, in the sense that the technology won't vanish. Its continuing to be super-widely available for free is less certain, IMO ... and certain very common use cases will likely go away if it's not free.

There is an IMO quite possible scenario where it mostly becomes a tool for the very large tech companies who can afford it (Google, Meta, etc), and for governments, to process huge volumes of data -- it makes the current winners win more, rather than upending things.

25

Spaceflight Entertainment and Hobbies / Re: Star Trek Starfleet Academy TV Show

« Last post by Blackstar on Today at 12:14 am »

And there's this:

https://www.giantfreakinrobot.com/ent/scifi/star-trek-avary.html


He was a writer or co-writer of a number of acclaimed movies and offered to work on Star Trek for scale (basic rate) and was turned down. He says he was turned down because he was a Star Trek fan and that's not what they wanted.

I'd take that with at least a grain of salt. Somebody who writes movies is not necessarily cut out to write television. But if you have a huge talent walk in the door offering to work cheap, you should at least give them a try.

26

Space Policy Discussion / Re: Jared Isaacman as NASA's 15th administrator

« Last post by Blackstar on 02/18/2026 11:45 pm »

Science is not the only reason to go to Mars. I think the refusal to acknowledge that there could be anything OTHER than science reasons for going to Mars might be part of the reason those advisory groups were cut loose. They had become unelected and unaccountable gatekeepers for Mars, serving more as lobbyists against humans to Mars than as actually facilitating human missions.

Which of the advisory group meetings have you attended? Which ones did you participate in?

27

SpaceX General Section / Re: Data Center Satellite Design

« Last post by InterestedEngineer on 02/18/2026 11:39 pm »

Solar panels and radiators are the heaviest parts of satellite and if they don't need to be replaced so often, it would be stupid to just let them burn in atmosphere.

what does heavy have to do with anything?

Launch costs and disposal hassle.

vent the water, the solar panels are easy to burn up the size won't change that.

the cooling loop, OTOH, that'll be interesting to design it to burn up.

the budget is a 5-6t satellite, cooling loop is 1/6 that, and the cost to launch at $100/kg is 0.5M, which is about the same as today's starlink.

But as someone pointed out above, starlink profit margin is higher...

28

SpaceX Starship Program / Re: Starship Future Engine Development

« Last post by redneck on 02/18/2026 11:38 pm »

One of the main problems with new engine concepts is that developing them is expensive and chancy. Few people can afford to drop a few tens of millions on a concept that has a high likelihood of being a blind alley. Humongous pressure feds being a prime example of something that seems good at first glance until one actually understands the downsides.

None of these are Raptor, but there are concepts out there.

All that said, there are a number of concepts I would like to see explored. 
The multiple chambers with a common nozzle mentioned above.
Polygonal nozzles mentioned by Ben Brokert that are compensating.
Nested hex nozzles in a honeycomb layout.
Dual chamber concepts where the turbine exhaust is after burned in a low-pressure chamber while feeding a very                                                                                                                                                                                                                  high-pressure chamber.
Mechanical acceleration (centrifugal) of thrust chambers relative to rocket body pulsed as the vectors aligned. 
Laser enhanced exhaust.
Pellet stream assisted launch                                                                                                                                                                                                                                                                                                                                                                                                        .

29

Missions To The Moon (HSF) / Re: Artemis II : Discussion Thread

« Last post by Mythundare on 02/18/2026 11:25 pm »

Those numbers don't seem reassuring.

That fact sheet says the flammability starts at 4% (which explains why STS and Artemis 1 used 4%), and explosions become possible at 18.3%. So 16% is below the explosion limit, though not by a lot. It notes that ignition below 10% is quite difficult, but 16% is quite a bit above that.

Honeycutt says they couldn't get hydrogen to ignite at 16%, which doesn't seem to align with what other sources are saying about the flammability of hydrogen (both that fact sheet and randomly googling the matter, such as the abstract of this paper https://www.sciencedirect.com/science/article/abs/pii/S0950423025000129 which shows empirical numbers are higher than theory). Maybe there's something specific to the TSMU environment that makes it less of an issue.

The fact sheet lists the H2 flammability in air, the TSMU environment is purged with (I think) helium and/or nitrogen. No reason to assume you have 21% O2 in there.

Doh, yeah that's obvious now. Thanks!

30

Advanced Concepts / Re: Moving The Cloud to orbit

« Last post by Coastal Ron on 02/18/2026 11:23 pm »

Again, you should actually just try this. I’ve noted that people are reacting to it as if it’s a scam, like NFTs or scam, whose value is built on speculation.

I could be accused of being part of this group, but I certainly don't see A.I. technologies as a "scam". What I do see is that they are over-hyped by the CEO's of the A.I. technology companies.

Like all the CEO's that are claiming that we are on the verge of Artificial general intelligence (AGI) - with Large Language Models (LLM). Come on now, that is just pure hype.

Think about it, we don't even understand how human intelligence works in our organic brains, but these CEO's think we have it figured out in silicon hardware and software that is fed suspicious quality of data?

So to me, I understand that the current generation of A.I. is just like the last two generations of A.I. - they are tools. But it is the CEO's claiming abilities beyond what they actually can do that create the atmosphere of incredulousness.

People are using this stuff today, and paying for it.

I pay for my developers to use Cursor, and I'd pay more if they raised the prices (don't tell Cursor that  ). But developer tools are the one market segment with the highest adoption rate, and the clearest ROI. Other market segments are not as easy to both adopt and profit from.

However there is LOTS of experimentation going on, which is where A.I. companies are seeing some of their revenue, on experimentation.

It’s already deeply integrated into Google, especially the core search functionality. Google (Alphabet) has $400 billion annual revenue. Look it up yourself!

Remember when the Google search engine first became a thing, the concern was that they were limiting the information that we were getting. Now we are getting A.I. summarized versions of that same information, and since Google has such a market dominance, everyone is forced to use it. So to some degree there isn't a viable alternative to using Google search, but that doesn't mean it is better with A.I. Just sayin'...

Don’t react to AI just because a lot of artists hate visual AI (understandably). Visual AI is a pretty small part of the workload and uses totally different models.

Copyright issues have been around forever, it is just the pace of how A.I. uses content that has changed. Out of curiosity though, after all the hubbub about these content generation tools, who uses them on a daily basis? I don't know of anyone. Not saying they aren't a marvel, but so far they are more demonstrations than actual needed products. More entertainment.

I keep seeing otherwise smart and intelligent people hyper-dismissive of AI because it’s part of the political culture war.

I don't know anyone that views A.I. from a political perspective, and I know a lot of A.I. professionals. So I think this is a false impression you have.

I’m personally somewhat skeptical of the god-like AGI hypothesis (or maybe that’s cope on my part), but there’s clear, actual value here. YOU use it.

We are hundreds of years into the Industrial Revolution, and I view A.I. as part of that. Of course nuclear power and nuclear weapons are also part of that, so the caution here is that we have to make sure we aren't letting any genies out of their bottles that we can't put back. Like using A.I. to design viruses that can kill all humans - that used to be science fiction, but it may no longer be fiction.

But I agree that current A.I. is here to stay, but I am amongst those that think we have not figured out how to bound A.I. correctly. And that, if anything, is the only political part that I'm aware of.