wumpus

I've heard that's all you hear on ISS: the drone of dozens if not hundreds of fans. Since there is no convection, every heat emitting bit of electronics (i.e. everything that consumes measurable power) needs a fan. They add up.

Depends on the probes. I did a quick search for Venus probes and hit this: https://nssdc.gsfc.nasa.gov/nmc/spacecraft/display.action?id=1978-078A Total cost for both probes: $83 million. Congress only deals with hundred million increments, so there was likely pressure to keep it below that magic number. Also when you are pitching a project, using two probes means a lot less "stuff" being researched, unless you can send them to wildly different targets. Seeing the bottom drop out of launch costs, you'd think they'd go back to making copies of probes. But from what I've heard of pitching a NASA proposal, the guys you are pitching to are from the early Shuttle design era, if not Apollo veterans. While they might have heard that launch costs have dropped, launches are still ungodly expensive in their world: no redundant launch costs!

When Science and politics meet, where is the line between investigation & evidence vs politics and casting blame? When something fails, it becomes a battle of casting blame. If all you wanted was investigation and evidence, you wouldn't invite the politicians. Truth doesn't help them, only what benefits their faction. And don't expect scientists and/or engineers to be all that beholden to evidence either, CYA engineering becomes second nature as engineers get seasoned: anything that has something that appears to be a point of failure (and the opposing faction can conveniently ignore why it that failure almost certainly isn't the cause). I'm not familiar with scientists, but in academics expect the politics to be worse than Congress, and even those in government or only mostly in industry still need to keep those grants coming. And those investigating you may well be competing for those grants. Occasionally the investigation is thorough enough that no amount of casting blame around can provide cover: see Apollo 1 and Challenger. But don't expect human attention to last that long without a strong reason, and make sure your designs not only hold up to physical laws, but look good to committees as well (a smart engineer has prepared for a design review as if it was an investigation for a fatal accident. Nothing for the committee to attack. If worst comes to worst, it may be such an investigation. Let someone else who wasn't as careful in their design take the blame).

Because it is completely atypical (67% is extraordinarily deep for a pumped engine. And the thing also has all the complexity of a full flow engine). But certainly for the mission (lighting at sea level and proceeding nearly to orbital velocity) throttling should help. The really irritating thing about the SSME is that it couldn't relight (and they didn't feel safe lighting it in flight) when it was pretty clear that would make a lot more sense for SLS. Pretty sure they would light the SSME (because they could turn it off, and it wouldn't rip any bolts while on) first, then the SRBs. Hopefully it could survive one SRB burning away on the pad with the other a dud, but I suspect that's the "name elementary schools after you" condition that Chris Hadfield was afraid of.

Is this another of your joke posts? All fusion can do with the heavy stuff is bombard them with neutrons, with semi-predictable results. Exactly like fission. I'd have thought that global warming would at least get people to think that "the answer to pollution is dilution" isn't the one and only answer. And that nuclear [fission] power's ability to concentrate its waste into a small footprint is an advantage. Also, sometime in the 1990s, "glassification" became the ideal end product of nuclear wastes (after being left in a pool long enough to decay a few steps) as you wind up with metamorphic rock diluted to whatever radiation level you wanted (concrete would do the same, but erosion would work faster on it). Personally, I'd recommend keeping the rods fairly deep underground, but where they could be pulled out in a shortage.

They don't care about what you would like or be interested in. They only care about getting you to watch/pay for something, ideally which ever makes the most profit for them. Also it would be very hard to measure how much you "liked" previous titles (unless you regularly rated shows/reviewed them), while it is fairly easy to determine the expected amount of profit from selling you something. I'm sure this is especially lucrative for kids shows, when you not only sell them the show, all the ads spammed on the show, and the inevitable toys/clothes/pillow cases that they will try to sell you based on that IP. I was talking about more like things here : https://www.unite.ai/the-struggle-to-stop-ai-from-cheating-on-tests/

I'd like to think so. But I doubt any reached the US (I remember hearing that Finland "cheated" the Oslo accords by replacing coal use with nuclear and was the only nation to meet the stated goals, presumably these are fairly modern). The most recent nuclear power construction start in the US was 1976. Granted, one was completed in 2016 and four became operational in the 1990s, but a lot of that design had to have been set in stone (or reinforced concrete) in the 1970s. I'd like to see a real push for fission in China and/or India, but I'm not holding my breath.

Sounds like one of the big current difficulties in machine learning. You put in your training data, look at the output data and reinforce any positive outputs. So as long as you get the right output for that data, the system keeps reinforcing that scheme. For modern values of "AI", this is exactly what you should expect.

The SRBs contributed ~80% of the thrust (the diameter of the SRBs were carefully tapered to have a "pre-planned throttle schedule", while I don't think the SSMEs can throttle at all. Kind of the reverse of KSP), so they essentially supported the shuttle until they were jettisoned.

Sometime in the 1990s I began to suspect that the whole cult of fusion was a concession to the anti-nuke cult and a belief that we could get it right this time with a new nuclear process. And not only will it have to deal with the anti-nuke cult, it will also have to deal with the existing nuclear plant construction industry desperately defending their turf. I can't see fusion doing any better than fission. Most of them. But why did Germany turn off their nuclear power plants and go back to the dirtiest coal they could find? Also, the greens aren't remotely as good at preventing new power plant construction as the nuclear power plant construction industry is at preventing nuclear power plants from being constructed (by massive delays and cost overruns).

A *hydrolox* powered bomber? While Reagan* spent billions replacing MAD (Mutual Assured Destruction) with weapons systems also designed with a "Strike First and win Global Thermonuclear War", I'm pretty sure** there weren't many systems that would be absolutely useless (like a hydrolox bomber) against an opposing attack. * the shuttle was mostly designed and budgetted under Nixon/Ford/Carter. But I don't recall any other "shoot first" programs being pushed through at the time. ** there was a bit of a scandal over the "dense pack" strategy used by the Peacekeeper missile (MX) not being remotely as able to survive an attack as initially claimed, but I suspect they could get most of the missiles launched in the 20 minute window an attack would give them.

"Someone programmed the code" implies "somebody came up with the specs". And while programmers tend to be careful about code (assuming they have to fix the bugs), managers/customers/whoever are notoriously careless about specs. And for something sufficiently complex as an AI developed to match/exceed human capacity, expect it to strain the "specs" to a limit and some completely unexpected results. It is one thing to fix bugs. Fixing "broken in the specs" can take a rewrite.

To be honest, that is the real question for this thread. What does an AI consider "acceptable casualties", or otherwise "acceptable results". The movie quoted had an AI that was programmed to learn how to "win" Global Thermonuclear War. Fortunately, not losing appeared to have a higher weight than "winning".

No. He talks about a chess game on a computer that is purposely playing less than its full capability. And then promptly increases its skill a notch and promptly beats him again. From quick googling, Stockfish (with only 4 threads, it can scale to 512) is expected to beat any human player. Should be in these packages: https://www.chessclub.com/downloads This thread also needs this: https://xkcd.com/1002/ The human's only real advantage is penetrating the fog of war. Since this is closely related to modern (and highly effective) machine learning research, expect all such advantages to disappear in the near future. Historically humans have excelled at strategy while computers win at tactics. Also this "overall strategy" is in formal parameters (like a chess game), where there exists a limited state to deal with while a real war would theoretically involve everything in the world (or solar system for your example). Reducing this to a workable subset has been hard for a computer, but recent improvements are startling. Of course the real elephant in the room is logistics, and my guess is that computers (mostly old fashioned software with human generated algorithms) have effectively controlled logistics since at least the start of the 21st century.

1. Venus has a very, very small eccentricity. I think it wasn't measured until the 20th century. 2. Zero doesn't exist in physics, except as weird quantum effects (like superconducting). Orbits are old fashioned physics, so literal zero isn't possible.

Venus approaches zero eccentricity, and googling shows 0.007. My guess for a "real" zero eccentricity would be a bi-star system, or otherwise have a large body that has more effect on the planet than any apparent eccentricity. Unfortunately, if it moved that much it might have an observable eccentricity, at least until batted around again. The real issue in trying to claim something is "zero" vs. "less than a specified epsilon" is that the real world simply doesn't allow perfect measurement. Any solar system with 3 bodies not locked into Lagrange points will be unstable, so there simply won't be an absolute value of eccentricity. Even with a two body problem, there are only so many zeros you can pile on before eventually you hit a non-zero number.

There have been 9 flights over 300 days at 0g, with the longest being 437 days (in Mir). So there's quite a bit of data for 0g, none for .38g (unless you are interested the 12.5 days at .166g that Apollo 17 did on the Moon. Of course they also spent a similar amount of time at 0g, so I'm not sure how good that data would be for anything).

Don't forget your hydrazine RCS thrusters. Less fire, more evil.

Not sure how you can call it a "loss" when Starlink currently has (beta) paying customers. They'll have to get the reliability far higher if they want the big money tying financial districts together at a slightly lesser ping than undersea fiber. According to the infallible wiki (although they appear to be quoting NASA), Falcon9 1.0 cost less than $400 million to design (price includes Falcon 1 and Merlin development). NASA's "traditional contracting procedure" expected to spend $4 billion to do the same, and Bezos hired guys used to doing "NASA's traditional procedure" (although at least $1 billion of that is ULA profits, so that would be paying yourself). It also cost roughly a billion dollars to develop reuse capability. Even for Bezos, this hobby is getting expensive, and it isn't clear he is going the right way. Right now, ULA is contracted to launch Kuiper. Expect long lead times and a longer wait before Kuiper can be an internal customer. It also isn't clear how being a second internet ISP is going to be a great advantage when he gets around to launching it. Starlink will compete with existing ISP monopolies, especially in less populated areas (populated as seen by satellite coverage), as an effective monopoly in highly rural areas, and as a slightly lower latency provider between financial districts (although it isn't clear when Starlink will be able to take this role, nor how much Kuiper can learn and compete quickly). I can't imagine what the financials look like for that company, although you can say that for Amazon as a whole for the first 20 years of its existence. Seymour Cray is known for saying "One of the problems of being a pioneer is you always make mistakes and I never, never want to be a pioneer. It's always best to come second when you can look at the mistakes the pioneers made." While this gives Bezos quite a few advantages, he doesn't seem to be set up to take advantage of them.

If you want metals, why don't you grab them from the asteroid belt? And why can't you simply smelt something on Mars? If a solar furnace won't work, try a nuclear one (you'll probably want it for power anyway). Smelting on the belt is much more iffy, although perhaps any low quality rock can be used as an open cycle coolant.

Good question. But the big difference is that while Musk has a paper wealth similar to Bezos *now*, he needed the launch cadence to fund his R&D. That isn't necessarily the case with Bezos, although how happy he is with the cost and speed of old space/military industrial complex (he hired guys from old space and got and old space process) is an open question. I'm not sure that Musk would retain his paper wealth (and doing so would outrage investors far more than the monetary hit) if he tried to sell a few billion dollars worth of Tesla stock, so he's probably stuck with largely revenue (and outside investment) financed R&D for now.

The line "rockets red glare" in the Star Spangled Banner refers to actual rockets being used against Fort McHenry by the British in 1814. I'm reasonably sure piracy was still an issue then, but I'd be surprised if this was used historically. Canister cannon loads seem more appropriate (although it would be quicker to reload your rocket for other boarding planks). https://www.airspacemag.com/history-of-flight/rockets-inspired-francis-scott-key-180952399/ [inevitable Scott Manley link: https://www.youtube.com/watch?v=H0deqePQAD8]

Assuming the "cylindrical Earth" argument is valid, you can sail two ships against the wind (one of the first things a sailor learns is how to tack). Not sure if the "use the wheels to drive the fan" method will work, but presumably it is possible.

They've already designed the raptors (although I'm sure they are also a work in progress). If they can throttle deep enough, SuperHeavy can hover. Assuming you can turn off arbitrary numbers of the 30+ raptor engines, that should be trivial to achieve. The other issue is of course maintaining stability. But compared to a Falcon 9 booster, a SuperHeavy will have considerably more moment of inertia and require far easier timing. It might take more force to keep it upright, but you have much more time to react. Making it hover should largely be a function of having enough propellant left in the tanks to hover until caught.

If the atmosphere will erode over time, why does Venus still have a thick atmosphere after 5 billion or so years? For temperature, you need to lower the greenhouse effect lower than Earth. That should get you significantly lower temps, but I can't compute if it will be enough. Your choices come down to dragging a good candidate into the "Goldilocks zone", building your own planet (perhaps getting Slartibartfast to add some fjords), or resurfacing an already inhabited planet. I'd have to assume that moving a planet requires only one step of building your own planet so should be easier.