Max Fagin

I'm right there with you on the first paragraph. Fill the solar system with telecopes! But I'd also like to fill it with humans.

Stable human populations only occur catastrophically or coersivelly. The best civilization for the people living in it is a growing one, and that requires outer space.

No solutions, only tradeoffs 🤷‍♂️ But given that the glut of LNG from the fracking boom has mostly replaced coal, I think it's a win. LNG may be CO2 intensive, but at least that's *all* that it is (Unlike coal, which is CO2 intensive *and* directly harmful to human health). Plus, cheap helium!

Slight misunderstanding. We were risking a helium shortage, but the fraking boom fixed that (helium is a side product of LNG refining).

But no one *ever* proposed importing regular helium from the moon. The proposal is specifically He3, a rare isotope that could fuel (currently nonexistent) fusion.

YouTube video by Erik Wernquist - Digital Artist Opening Sequence - THE HIGH FRONTIER

I don't know about your future space, but my future space space is full of people in spinning habs with *better* gravity and radiation that Earth offers 😉

youtu.be/pSsWkooeIds?...

Kessel Syndrom is a Star Wars thing 😉

Kess*ler* syndrome is only a problem for irresponsible operators that intentionally generate uncontrolled debris. SpaceX has demonstrated that they can operate a large constellation while the amount of debris actually goes *down*

bsky.app/profile/maxf...

Lol at the scale and lol at the economics if you wish. Those are fine things to be skeptical of right now.

But stop lol'ing at the heat rejection. It's a silly objection.

bsky.app/profile/maxf...

No, but you would with ~1 million of them with optical laser interlinks, which is what SpaceX plans to do.

bsky.app/profile/maxf...

bsky.app/profile/maxf...

Same way they cool the ~10,000 of these that are in space right now.

SpaceX has already launched a constellation of satellites with ~1000 times the power of the ISS.

O'Neill got that in the 70's, and SpaceX recently gestured in that direction: At some point in this build out (~mega to giga tonnes) ISRU+ISMA in orbit beats Earth launch for the simple/dumb mass. That logistics chain is not easily automated. It requires ~civilizations on the moon and in space.

That is a huge part of it, agreed. But I don't want people to come away from this with the impression that Space NIMBYs aren't also a thing.

I solve this problem with kids by pronouncing it "Kuroo-an-us" (slightly rolled r). Then if any kid corrects me, I get to tell them how "Uranus" is our mispronounciation of the god we now call "Kronos", and by the time a 12 year old untangles that 🤯, they've forgotten the butt joke.

I'm sorry. The proximate cause that made me finally post this was Kyle Hill's YouTube video on the subject, and the term fit. He was scoffing at something Starlink is doing *right now* as "breaking physics", all while lampshading the fact that he isn't a physicist, as if that made it better 🙄

SpaceX has already proven they can launch/operate thousands of satellites.

And based on their plan for this future constellation, they want to do ~1 million satellites.

I won't say SpaceX can do it. But I will say if anyone can do it will be SpaceX.

Well, no, we aren't bound by municipal zoning laws (why would we be? Satellites aren't buildings). But we still have laws and regulations in space that ensure people's safety. And US based satelilte operators are still subject to US federal and state law.

People will always strive for that. It's part of being human.

But we don't get from the DaVinci flyer to a Boeing 777 because airplanes are cool (even though they are). At some point, someone needed to invent commercial air travel and make it profitable.

Comrade, have I got some news for you:

For All Mankind wasn't a documentary

🦅🦅🦅🦅🦅🦅

The full answer is complicated, but the short/simple version is "square-cube law". One ~100 kW space station needs massive deployable radiator assemblies to avoid overheating. But ten thousand 10W CubeSats don't need any radiators

(I'm not saying SpaceX is using CubeSats; just an extreme example).

Don't know what aerospace industry you're a part of, but we are still bound by the same laws everyone else is. Arguably *more* since we have to contend with various levels of international law too.

That's why the end goal for these plans is civilizations in space to build/support this infrastructure. Giant O'Neill habitats, offering greater comfort than planets can provide.

A megaconstallation like this would (maybe) be step 5 of ~100. But that's great when we previously were stuck on step 4!

Apollo was that. But you can't make an ecosystem out of proof of concepts. It needs to self-sustain and pay for itself at *some* point or the whole thing fizzles out.

Good question! Why people in space?

In simple terms, because:

1) People good

2) More people more good

3) Maximum people maximum good

4) Maximum people need people space.

In much more elegant terms 😉:

youtu.be/XH8Dn_d5mIs?...

"Because it's there/cool/meaningful/good" is great for Apollo style projects! But for building something that can sustain itself over multiple generations, I think profitability helps long-term viability.

That is not correct. Each V2 Starlink launched in the last few years has an installed capacity of at least 30 kW_e PP (and a typical 5-13 kW_e OAP). Thermal is slightly less, but still enough for *several* GPUs, not just one. V3 will be several times higher than that, especially in SSO.

It's definatelly a problem! But it's a problem mostly to be solved with a good old-fashioned orbit selection trade. The kind every satellite constellation operator knows how to do.

A kW of heat rejection doesn't care if that kW is coming from a radio amplifier or a GPU. These latest generation Starlinks may not have a lot of compute on them (why would they?), but they do have the ability to reject heat for a ~rack's worth of compute. More even, if they were launched into SSO.

Yeah. Rather than having to build ~100 datacenters and navigate land use and zoning in ~100 municipalities, they only need to get approval for ~one thing (launch a satellite into SSO on Starship) and then "only" do it ~1 million times. Better economies of scale.

And that's a totally fair skepticism to have! It's even easy to defend, especially if you are skeptical about grandious extrapolations of AI trends.

I just wish the discourse around space based compute was more that, and less of "lol, Earth's #1 satellite operator doesn't understand radiators" 🙄.

An advantage of orbital compute over SBSP is you *don't* have to build a giant centralized solar farm. No massive transmitter. No massive receiver. The mass launched to space in is still ~millions of tonnes, but it's all atomizable and identical copies. No in-space assembly required.