Hacker News Reader

> I’m looking through the list of the heads of nasa ... But why did none of them try what spacex is doing?

There’s no scenario where NASA or a NASA administrator could do what SpaceX is doing, well or poorly. That’s not their role.

That said. Charles Bolden pushed back on congress with regards to SLS as much as was practical. (I imagine other administrators did too, in their way) Congress wanted a jobs program and is getting a jobs program, and in that situation the NASA administrator can just say “this is not what we need” and be polite when congress ignores his recommendations. It’s more complicated than that, but not by much…

Technologically it very well might have been possible far, far earlier. The computer in Shuttle are very complex, more complex then what is needed for a simple Falcon 9 70s style.

NASA shackeled itself to Shuttle, a expensive failure of a project and from them on the basically didn't develop anything new for 3 decades.

The millitary if anything is where such thinking should have been happening. The went threw many iterations of rockets.

A few reasons why they never managed to go reusable. Reliance on solid boosters, once you have them reusability is not really practical.

The millitary also simply doesnt care much about cost. What they care about is reliability above all else.

And you would be suprised how much engineers buy into constraints. The reality is that after a while the status quo of the industry you work on is simply an assumed constraint. Nobody asked engineers the question 'how do you get 100 tons to orbit per week' or whatever.

As Musk puts it, the most difficult part is asking the right question. Nothing is as harmful as smart engineers spending lots of money and time answering the wrong question.

It certainly makes you wonder, what other innovations could we have made decades ago, but nobody was willing to take the risk? Currently most of our innovation is directed towards computer systems for keeping track of people and data, which in many cases can be broadly characterized as systems of control. And that appears to be what our society prioritizes, whereas a company like SpaceX only comes along once in a blue moon. And Elon himself is sort of a fluke as he's a person who is both a child of wealth, and also has a curious mind and a knack for understanding engineering problems. There are lots of curious people with a knack for understanding engineering problems but most of them weren't born into money and of the ones that were, only Elon has emerged as willing to make SpaceX level risky investments. It makes sense; if you're that rich, you have no desire to genuinely innovate, since you're already set for life and by definition are one of the people who benefited from society being exactly as it is.

Maybe that's the real takeaway here: the people at the top of society by definition benefited from things being as they are, therefore those people are the least likely to want innovations that might cause real change. Therefore to the extend that society is unequal, it is also unable to innovate. I wonder if you could put that hypothesis to the test with data.

The reliability factor is always understated. When you have a project that didn't only cost more than a billion but also took more than 10 years to make you really the rocket thing to work the first time.

Its more than just "can we do it", its also "does it make sense".

If you are doing a limited number of missions, it doesnt particularly make sense to go reusable, because you're going to have a hell of a time getting suppliers. If you are going to do 6 launches a year, and you can do that with 2 rockets, then you might think to build 4 rockets total. If youre developing your own engines, you now have to spend all that money to build maybe 4-16 total engines, depending on the design, which means your per engine cost is going to be massive. Same with core stages and everything else. You end up paying massive costs per component, and now you have to figure out what to do with those production lines you just started. Do you keep working on making more engines? Do you shut down the line because you've made enough for your 4 rockets?

That, combined with a lot of reliance on solid rocket boosters, which are less important to make reusable, makes the idea sketchy. If we needed to put tons of stuff into LEO 30 years ago, the projects probably wouldve been attempted before. We just didnt really have the launch cadence to support it

A huge problem I think is that just entering the market is terribly difficult.

It's not a setting where you can get a million or two and see if it works out. Trying this requires pouring lots and lots of cash into it, it can all go up in flames, and even if it works you're then competing for a small market. If you just want to make money there are probably much easier and less risky ways if you have that kind of budget. You need the rare combination of somebody who is really into space tech and can get the funding for it.

And even if the thing flies, it's easier said than done. NASA tried to make the shuttle quickly and cheaply reusable. It didn't quite go to plan.

And even after Musk proved it can be done... not much happened elsewhere. Whoever else also wants to do it either needs to have some reasonable expectation they can compete with SpaceX, or needs to have basically unlimited money to spend on a passion project like Blue Origin.

> I’m absolutely certain that Elon musk isn’t the first person to come up with the idea of reusable rockets, and the people who actually did the work of designing those rockets were already around before Elon musk hired them. Tom Mueller, who was the first rocket engineer Musk hired, was building rocket engines in his garage out of frustration that nobody was taking him seriously at work. So the question is, how deep does institutional rot have to go to stifle innovation on this level industry wide? How does an organization dedicated to exploration become so utterly unable to take risks?

I don't think this is a NASA problem, it's really a problem across the US Federal government. You see it on the military side of government-funded technology programs as well. (If you want to see a really terrible example, look at the ongoing failure to produce a replacement for the M109 howitzer.)

My personal theory is that it's mainly driven by funding mechanisms, especially those covered by FAR.

It's not, as some people allege, that the government is particularly bad at things; the government is roughly as competent as any large corporation when it comes down to task execution. I've consulted both to private sector corps and the government, and the fundamental challenges are largely the same. Yes, government compensation structures (heavy on pensions) do tend to attract people in different phases of their careers and lives than, say, equity-heavy packages at a startup. But there's not much evidence linking this to performance.

Rather, the difference is that the US government has an institutional attention-deficit disorder. What gets funded one year might get cancelled the next. No plan can possibly be accurate beyond the end of the current budget cycle, no priority fixed beyond the next election. Everyone is going for "quick wins", because they want to be able to justify the continuation of the project. This is not conducive to risk-taking, at all. Quite the opposite. What government PMs want are projects that can set a ~24 month goal, and then show monotonic progress towards that goal every month, like clockwork.

Projects that can't show those kind of clean metrics are often managed to death, if they are even started at all.

Private-sector organizations, particularly those run by single individuals without regard for short-term profitability, can take on risks that the government won't—basically can't—deal with. Elon Musk made some pretty big technical bets that paid off... but the government isn't allowed to play at that table with tax money.

> It's not, as some people allege, that the government is particularly bad at things; the government is roughly as competent as any large corporation when it comes down to task execution. I've consulted both to private sector corps and the government, and the fundamental challenges are largely the same. Yes, government compensation structures (heavy on pensions) do tend to attract people in different phases of their careers and lives than, say, equity-heavy packages at a startup. But there's not much evidence linking this to performance.

I generally agree with the overall gist that large organizations are basically incapable of achieving excellent results.

However, having worked as an employee for both large corps and government, I see two or three attributes that cut against government being effective.

1. Private industry is mission oriented whereas government is process-oriented. Government bureaucracies value adherence to process above all, regardless of what they create or impede. It's all process all the way down. Even when large, ossified firms make process a centerpiece of management, they still ultimately are attempting to pursue a mission.

2. The government's culture is homogenous and based on a mythos about "serving something bigger than yourself," whereas each firm in industry tends to have some more concrete aspirational statement.

3. Government always has many more people on the critical path who can seriously impede work, because they're hard to get rid of or move out of the way, whereas private companies would have no problem firing such people.

On the whole, I think even the largest monopoly firms would tend to be at least a few degrees more efficient than government just for missional focus and the ability to remove bad performers. If a competitive firm were to be as (in)efficient as government, it would die.

However, communication, internal politics and management are major challenges for any organization at scale. It really is a difference of degrees, not orders of magnitude.

I think society just got scared by the pace of progress and change, then decided to tap the breaks.