A lot of great inventions we now take for granted initially came with little motivation other than being able to kill each other more effectively. GPS, radar, jet engines, drones, super glue, microwaves, canned food, computers, even the internet. Contrary to the narrative of the internet being about sharing science, ARPANET was pushed by the DoD as a means of maintaining comms during nuclear war. It was then adopted by universities and research labs and started along the trajectory most are more familiar with.
The tale of computers is even more absurd. The first programmable, electric, and general-purpose digital computer was ENIAC. [1] It was built to... calculate artillery firing tables. I expect in the future that the idea of putting a bunch of solar into space to run GPUs for LLMs will probably seem, at the minimum - quaint, but that doesn't mean the story ends there.
I think the Colossus[1] predated the ENIAC but is still in line with your general theme of doing stuff for the military. In this case it was used for cipher breaking, not firing calculations.
You could argue that it doesn't really count though because it was only turing complete in theory: "A Colossus computer was thus not a fully Turing complete machine. However, University of San Francisco professor Benjamin Wells has shown that if all ten Colossus machines made were rearranged in a specific cluster, then the entire set of computers could have simulated a universal Turing machine, and thus be Turing complete."
That’s not the point of the person you are replying to. They are saying if we somehow come up with the tech that makes harnessing the sun a thing, the best we can still do is put a bunch of GPUs in space? It makes no sense.
It kinda does make sense if you consider that solar panels in space have been used for a very long time (to power satellites). However, getting the electricity they generate down to Earth is very complicated, so you end up having to use it in space, and one of few things that would make sense for that is indeed data centers, because getting the data to Earth is easier (and Elon already handily has a solution for that).
However I'm curious how many solar panels you would need to power a typical data center. Are we talking something like a large satellite, or rather a huge satellite with ISS-size solar arrays bolted on? Getting rid of the copious amounts of heat that data centers generate might also be a challenge (https://en.wikipedia.org/wiki/Spacecraft_thermal_control)...
The plan seems to be for lots and lots of smaller satellites.
For inferencing it can work well. One satellite could contain a handful of CPUs and do batch inferencing of even very large models, perhaps in the beginning at low speeds. Currently most AI workloads are interactive but I can't see that staying true for long, as things improve and they can be trusted to work independently for longer it makes more sense to just queue stuff up and not worry about exactly how high your TTFT is.
For training I don't see it today. In future maybe. But then, most AI workloads in future should be inferencing not training anyway.
>Getting rid of the copious amounts of heat that data centers generate might also be a challenge
at 70 Celsius - normal for GPU - 1.5m2 radiates something like 1KWt (which requires 4m2 of panels to collect), so doesn't look to a be an issue. (some look to ISS which is a bad example - the ISS needs 20 Celsius, and black body radiation is T^4)
oh, we'll sure find a way to weaponize that energy for example - just imagine all those panels simultaneously turning their reflective back in a way to form gigantic mirror to focus reflected solar energy on your enemy, be that enemy in space or on the Earth/Moon/Mars ground. Basically space-scale version of 'death ray scyscrapper' https://www.businessinsider.com/death-ray-skyscraper-is-wrea....
Back in the day the Star Wars program was intending to use nuclear explosions to power the lasers, i guess once all that solar for AI gets deployed in space we wouldn't need the explosions anymore.
Interesting that such space deployment can deny access to space to anybody else, and that means that any competitive superpower has to rush to deploy similar scale system of their own. Space race v2.
Pick any Gundam series and watch the last 5 or 6 episodes, at least through the Gundam SEED/Destiny era. At least part of the plot will invariably include a space-based superweapon being deployed by one side of the war to end all wars and the the plot for a few episodes will include the other side engaging in a series of challenges to keep that from firing again and destroying it if possible.
Yes, but isn't that pretty much the point of the person you replied to? We know that a lot of inventions were motivated by that, and so it is incredibly myopic to not pause and try to think through the likely far broader implications.
Scaling photovoltaic production doesn't seem likely to have many broader implications on its own. At best, it makes it easier to change the grid to renewable power, if you ignore the intermittency problem that still exists even at huge scales. PV fabs aren't really reusable for other purposes though, and PV tech is pretty mature already, so it's not clear what scaling that up will do.
Scaling rocketry has several fascinating implications but Elon already covered many of them in his blog post.
Scaling AI - just read the HN front page every day ;)
What are we missing here? Some combinatoric thing?
Bollocks, by your standards we can't discuss the most vile people because 'nobody's perfect' but there is a huge gap between the likes of Musk and ordinary people.
Indeed, at least a $700 billion gap. One is reminded of a great Mark Twain quote, "Whereas principle is a great and noble protection against showy and degrading vanities and vices, poverty is worth six of it."
The problem is that the Venn diagram of 'vile people' and 'billionaires' has a lot of overlap so these people are doing a disproportionate amount of damage.
Not to go heads I win, tails you lose, but even if we go down this path - it's the same story because militaries are investing heavily in LLM stuff, both overtly and covertly. Outside of its obvious uses in modeling, data management, and other such things - there also seems to be a fairly widespread belief, among the powers that be, that if you just say the magic words to somebody, that you can make them believe anything. So hyper-scaling LLM potential has direct military application, same as Starlink and Starship.
You're not considering some important multipliers. In space you're already getting a substantial immediate boost due to greater solar irradiance - no atmosphere or anything getting in the way of those juicy photons. You can also get 24 hour coverage in space. And finally they mention "deep space" - it's unclear what that means but solar irradiance increases on an inverse square law - get half way to the sun and you're getting another 4x boost in power. I'm sure there's other factors I'm not considering as well - space and solar just go quite well together.
Contrary to what most people seem to think about the past, slavery was oft seen as naturally repulsive even thousands of years back. It required regular defense. In Aristotle's Politics [1], written some 2400+ years ago, he felt compelled to lay out just such a defense and it was, by far, his weakest argument. He clearly started at his conclusion and worked backwards from there, instead of working forward from first principles and he did in other topics. The reason it's relevant is that he did accurately predict the end of slavery:
"For if every instrument could accomplish its own work, obeying or anticipating the will of others, like the statues of Daedalus, or the tripods of Hephaestus, which, says the poet, 'Of their own accord entered the assembly of the Gods.' If, in like manner, the shuttle would weave and the plectrum touch the lyre without a hand to guide them, chief workmen would not want servants, nor masters slaves."
There were thousands of years of efforts to end slavery, some countries would occasionally succeed at such only to see it spring right back. Yet following the industrial revolution it began rapidly disappearing everywhere that had gone through industrialization + urbanization. The issue in your mental model is that you're only considering local effects over very immediate time frames. Think about the bigger picture.
Industrialization drove big money away from farms and into factories, away from rural scarcely populated rural areas into densely populated urban areas packed with very poor potential workers. As soon as the necessity argument for slavery became plainly absurd, to say nothing of the issue of industrialization also reducing the need for so many workers even on plantations, slavery wasn't long for this world. This says nothing about actual slave holders who, as you said, did not just go quietly into the night. But as their economic might relatively waned, so did their influence.
With enough effort, anything can be obfuscated. But effort costs money and also state level actors have limited funds and time and want to go home to their families ar some point and if the purpose was to get a message across (don't mess with china, otherwise face the consequences) there is no need to really hide the origin.
People can simply have different perspectives on things. Measels has mortality rate of around 0.1% while small pox has a mortality rate of around 30%. So the individual risk from measels is relatively low leaving plenty of room for individual choice.
There's less room for the same argument with small pox. In fact small pox is where the term vaccine comes from - it was observed that milk maids weren't getting smallpox, which led to the discovery that infection with cow pox (which is relatively harmless) provided immunity to small pox - hence 'vacca cine', vacca = cow in latin.
But there's a long history of people trying to self vaccinate with all sorts of things against small pox including using scrapings off somebody's small pox wounds to hopefully give oneself a light infection. Needless to say that came with well understood side effects up to and including full-on small pox infection. But when the death rate is 30%, people were willing to do some crazy things, because the risk:reward was seen as worth it.
---
FWIW I did ultimately decide to vaccinate my children against measels, but it was not an easy decision, because it is in general not that risky a disease whose mortality rate had already precipitously declined (from 13 per 100k to 0.19 per 100k) [1] before a vaccine was first introduced in 1963. Obviously I think it's the right decision, but I also wouldn't really fault anybody for going the other direction either.
A friend's favorite Wiki page is on the micromort. [1] I found it kind of banal, but perhaps that's my more casual attitude towards death. I suspect the average person doesn't realize, or doesn't accept, how brief life truly is. One micromort is something with a 1 in a million chance of death. So with measels, we're talking about 1000 micromorts, if you're infected - which is also extremely unlikely. So if you give yourself a 1% chance of ever being infected with measels (which would be quite high) then not vaccinating would be down to a 1% * 1000 = 10 micromorts of risk.
All non-natural cause of death in the US, excluding suicide, is about 1.3 micromorts per day. So it's the same all non-natural cause risk you'll be exposed to over the next week. The page offers a lot of other comparables - traveling 100 miles by biking, 2500 by car, or 10k by airplane, and so on.
In a world with a low vaccination rate, infection with measles is practically 100%. It spreads more easily than COVID.
You can take your risk with herd immunity if you want. That works well now, with a near universal vaccination rate. But if the rate drops below 90%, then you will get infected.
Immunity rate, not vaccination rate. This is why everybody ended up catching COVID, often multiple times, regardless of how much they injected themselves with. Whereas, measles, with an ostensibly higher spread rate, was already on its way out before vaccines were introduced with fewer than 1 new case per 500 people - far below even basic fertility rates. It's because infection with measles is a nice mixture of relatively low risk and providing lifetime immunity. Beyond that, herd immunity is a hand-wavy measurement about statistically preventing perpetual spread within a population, it says much less about individual susceptibility.
Your child being vaccinated against measles is 1. yes, protecting them, but 2. also protecting all the other children at school they interact with, and vice-versa. It isn’t just an individual choice. You should definitely fault people for going the other direction because they are willingly increasing the chance of your child or your child’s friends getting measles.
The US sees in the ballpark of a hundred million tourists, business travelers, and migrants per year. And these people are vectors for basically everything and are numerically very large. And you're inevitably going to bump into and interact with these people. So I think the idea of domestic herd immunity is increasingly nonsensical because 'domestic' is no longer even remotely close to a closed system. And global herd immunity is nonsensical simply because it's wholly unrealistic, and at that sort of scale any small issue can explode into a huge one: see - source of most modern cases of polio.
One reason the idea of eliminating COVID was nonsensical is that it's carried by many animals, like all coronaviruses, and can be transmitted from humans to animals and vice versa. Measles, by contrast, is thought to be human only. So anytime there is an outbreak it's not from an unknowable cause. It's going to be either from a foreigner or somebody who visited a foreign country and returned with the virus.
Various social decisions have led to countries you might think of as measles free, no longer being measles free. For instance Canada, the UK, Spain, and obviously Mexico are all now considered to have endemic measles, with Canada and Mexico already compromising the majority of visitors and "visitors" to the US. And the vaccines are not complete immunization. Double dose measles is around 97% effective, meaning you can expect at least 3% breakthrough infections, possibly more depending on factors such as age, immuno compromisation, degree of exposure, and so on.
My parents both got measels, and they had no hesitancy getting me vaccinated. I’m significantly taller than my dad, who lost 40 pounds when he got German measles when he was 17. I grew several inches that year.
People who don’t get vaccinated are bad people. I have no qualms saying it.
I don't think this is a great argument. The reason is that Rubella/German measels is in the majority of cases very mild, in many - people will not even know they had caught it. If you take a large enough sample you'll absolutely find people winning a lottery that they really don't want to win, but the exact same logic is used to by people to avoid vaccinating. For instance the MMR vaccine can cause all sorts of nasty things in extreme cases, but in the average case it's believed to be mostly harmless. Though of course we're always discovering new things about medicines. There's even new side effects in e.g. tylenol still regularly being discovered and published about.
There's some major historical events here. It's not just random movement. Here is a sort of visualization of key points and the USD share of global reserves with the events attached. The number in percent is the USD share of global reserves.
1970 (85%) - Up until 1971 [1] the USD was backed by gold by the Bretton Woods system. Other countries could trade USD in for gold at a fixed rate and in exchange would peg their currencies to the USD and trade in the USD. The idea was to prevent the US from ever being able to exploit their power in this system because if we printed too many dollars other countries could just hoover those dollars up on the cheap, convert them to gold, and make a bunch of money. Nonetheless we did print excessive amounts of money and abuse the power this system was granting us. So France dutifully hoovered up those dollars and made a gold call. We said 'nah', defaulted on our debts, and withdrew from Bretton Woods. This led to the famous quote from Nixon's Secretary of the Treasury: "The dollar is our currency, but it's your problem."
1990 (47%) - Following those events, the USD began seeing rapid inflation, and other countries were rapidly dropping the dollar. This 'peaked' around 1990. But then in 1991 the USSR collapsed. This not only left the US as the undisputed and sole 'king' of the world, but also led to these former nations beginning to dollarize once the dust had settled. The USD suddenly again starts to see mass adoption.
2001 (72%) - The adoption reaches its peak in 2001. At this point not only was there the dotcom bubble, but the world order begins clearly shifting with China and Russia developing increasingly rapidly and looking to become viable world powers once again. And there's been a gradual process of de-dollarization since then declining to where it is today to less than 57% and very much trending to where we were right before the USSR collapsed.
2025 (57%) - Where we are today. We're very much trending towards 1990, only 10 points away, which is the level when everybody was dumping the USD, the US economy was shaky at best, and there was another major superpower in the world and nobody was quite sure who would 'win.' This is far more significant than 'random noise' you're implying.
---
I am leaving out plenty of relevant happenings including the transition to the petro dollar, South American economic crisis, and so on - but these only contribute to eras I think, while the above are the main drivers.
I only check the US stock markets to see the health of the US Dollar. Every time it reaches new high it means dollar devalues a tiny bit more than the last stock market peak.
It was always called the Department of War [1] from 1789 until 1947. At that point it was repackaged as the Department of Defense when we started framing invading countries half-way around the world as 'defense'. Prior to that rhetoric around war was far more honest. We tried to buy a sizable chunk of Texas from Mexico. They rejected our offer so we invaded and took it, because we wanted it.
It's only in 1947 and later that somehow invading countries half-way around the world and shipping weapons to anybody with a buck began being framed as 'defense' or somehow saving the world from whatever - tyrant, terror, communism, burdens of oil, and so on. So in many ways I think it would be far more apt to say that 'Department of Defense' is the cutesy name. They're not defending anything - nukes and geography take care of that, more or less, on their own.
> It was always called the Department of War [1] from 1789 until 1947.
No, what became the Department of Defense didn't exist from 1789 until 1947. The cabinet level Department of the Navy (current Department of the Navy) and the cabinet-level Department of War (later split into the current Department of the Army and Department of the Air Force) did, as separate, co-equal entities with no single civilian head over them beneath the President.
The National Military Establishment under the cabinet-level Secretary of Defense was created as a unified military structure in 1947 over both the Department of the Navy (which remained a cabinet-level department) and what had been the Department of War (which was split into the cabinet-level Departments of the Army and the Air Force). And in 1949 the three service departments were fully subordinated within the NME instead of being cabinet level, and the NME was renamed the Department of Defense (pribably not entirely because it was really awkward having the combined military organization use an initialism that sounded like “enemy”, but...)
All you are describing is a restructuring of which the Department of War had gone through repeatedly throughout its history. It's not like it had the same structure, or anything remotely like it, in 1942 as in 1789. The choice of the name was, as you observe, a choice. And it coincides exactly with the move away from public honesty in international relations and events.
You have things like WW1 being framed (at the time) as 'The War to End All Wars' but I think that was probably naivete whereas after we started calling war 'defense' we entered into the era of 'police actions' instead of wars, like the Korean War, and outright false flags such as the Gulf of Tonkin Incident for Vietnam. All the while the CIA was running around acting like a rabid chimp all across the world. It was entering into an era where deceiving the public became standard operating procedure, of which framing war as defense was but one typical aspect.
I believe we are now leaving that era, and I think that is a good thing for everybody.
> All you are describing is a restructuring of which the Department of War had gone through repeatedly throughout its history
No, from 1789 to 1947 there were two separate cabinet-level departments, War and Navy.
> It's not like it had the same structure, or anything remotely like it, in 1942 as in 1789
Internal to the two cabinet-level departments? Probably not.
At the cabinet level? There was exactly the same structure: the Department of War with the Army underneath it and the Department of the Navy with the Navy and Marine Corps.
The War Department did not become the Defense Department.
In 1947 War was split into Army and Air Force, and a fourth cabinet secretary, the Secretary of Defense was added, heading the combined National Military Establishment that was created over both what had been the War Department and what still was the Navy Department (all still cabinet level departments). In 1949, the three service secretaries (two of which headed parts of what had been the War Department) were formally subordinated to the Secretary of Defense and the NME was renamed the Department of Defense. The Department of War was direct predecessor to the Departments of the Army and Air Force, not the Department of Defense, which was a new level of coordination interposed between the President and the formerly organizationally-separated services.
This is inaccurate. The Department of War initially had oversight over the Navy as well. The separation of the Navy into a separate department (which did not even exist at the time when the War Department was created) was one of those many restructurings it went through, without ever having a name change until we entered the era of deception.
You're omitting the context provided by the article. This wasn't just a random scenario. Not only was this by an elementary school, but during school drop off hours, with both children and doubled parked cars in the vicinity. If somebody doesn't know what double parking is - it's when cars parallel park beside one another, implicitly on the road, making it difficult to see what's beyond them.
So you are around young children with visibility significantly impaired because of double parking. I'd love to see video of the incident because driving 17mph (27kph for metric types) in this context is reckless and not something human would typically do, because a kid popping out from behind one of those cars is not only unsurprising but completely expected.
Another reason you also slow way down in this scenario is one of those cars suddenly swinging open their door which, again, would not be particularly surprising in this sort of context.
That's my thinking as well. Taken in some abstract scenario, all those steps seems very reasonable, and in that abstract scenario we can even say it would do better than an average human would. But that is missing the overall context that this was an elementary school during drop-off hours. That's when you crawl at 3 mph expecting kids to jump behind any car, and not going at 17mph.
> But that is missing the overall context that this was an elementary school during drop-off hours. That's when you crawl at 3 mph expecting kids to jump behind any car, and not going at 17mph.
Indeed. Sure the car knows the limit, it knows it is a school zone, it can precisely track people within the reach of its sensors (but not behind blockages it can't see through).
But it is missing the human understanding of the situation. Does it know that tiny humans behave far more erratically then the big ones? Obvious to us humans, but does the car take that into account? Does it consider that in such a situation, it is likely that a kid that its sensors can't possibly detect has a high probability to suddenly dart out from behind an obstacle? Again obvious to us humans because we understand kids, but does the car know?
Urm, ime people frequently drive significantly over the speed limit in all these places, at all times of the day.
Blows my mind how you guys confidently state this with authority as if that's the normal behavior, when the reality is that it probably should be - but isn't actually.
So you're confidently stating it's not the normal behaviour... Can you tell me what the average speed is for human drivers outside elementary schools at drop-off times?
> Urm, ime people frequently drive significantly over the speed limit in all these places, at all times of the day.
The focus on speed limit as some truth is not the best way to think about it.
It might be 15/20/25 (varies, but those are the most common values I've seen).
But in terms of what is safe, it varies far more.
There will be circumstances where driving double that limit is 100% safe in front of the school. (For example, small numbers of high-school kids standing around but far from the street, so even if they did a mad dash to the street (which that age kids will not do), they still couldn't get in a spot to be hit by the car.)
And there will be circumstances where even a one-tenth of that speed will be far too dangerous to consider. (For example, high density of elementary school kids on a narrow sidewalk with many visual obstructions.)
I have a hard time believing that you, or anyone, would drive the same speed in both scenarios without any consideration to the circumstances of the moment.
I know I most certainly would never. If there is obviously zero chance of an accident I'll drive the limit or above. If there are tons of tiny kids in brownian motion, I'll slow down to a crawl or even stop if I sense risk (like a kid disappearing behind a parked car and now I don't know where they'll pop up).
But if you’re plan on building a fleet of cars operating all over the country or the world, do you want to model them after the careful driver, who has awareness about the situation (school, drop off/pickup hours, etc) or say “what the heck, some drivers are not paying attention so neither will my robots, it’s fine”
"I'd love to see video of the incident because driving 17mph (27kph for metric types) in this context is reckless and not something human would typically do"
I am not sure what your definition of typical is. The reason we have lower speed zones in school districts at specific times is because humans typically drive fast. The reason police officers frequently target these areas and write a plethora of tickets is because humans typically ignore speed limits.
Your claim seems to be that a human would drive much slower than the posted speed limit considering the conditions, but the laws and the court room suggest otherwise.
There is probably a multimodal distribution of behaviors for people in these situations. Any robotaxi ought to behave in the more cautious mode no? I drive at near idle speeds in these situations.
Driving is based so much off of feel so my numbers may be off, but in the scenario you are talking about 5mph seems reasonable, 10mph already seems like to much.
The want to be E but really armchair engineer in me for this context says there's far too little Engineering safety of the situation.
That school should not be on a busy roadway at all, it should also not have a child dropoff area anywhere near one but instead, ideally, a slow loop where the parents do drop off children, and then proceed forward in a safe direction away from the school in a flow.
It's funny because now you're sounding like you're blaming the school/the city for the situation.
Things are what they are. Driving situations are never perfect and that's why we adapt. The Waymo was speeding in a school zone. Did a dangerously fast overtake of a double parked car. It's engineering safety failure over engineering safety failure from Waymo's part, on nobody else.
Source? The article doesn't list a speed limit, but highways.dot.gov suggests to me that the speed limit would be 25mph in the school zone, in which case the waymo was going significantly under the speed limit.
> If somebody doesn't know what double parking is - it's when cars parallel park beside one another, implicitly on the road, making it difficult to see what's beyond them
I don't know about you but here in the UK, close parallel parking is normal and expected, and I was taught to avoid trying to cross in the gap between two parked cars, and to be extra careful if I was going to. In this scenario some blame might lie with the driver for going too fast, but I would certainly also blame the child (or their parent) for stepping far into the road without looking.
While I completely agree with your premise, the software can be improved. It can be programmed to drop down to 5 miles an hour or less depending on street size, pedestrian proximity, school zone, etc.
If only the same could be said for the other parents in the school zone. I’ve seen people roar by in similar scenarios at 30+ miles an hour.
There's a great analog with this in chess as well.
~1200 - omg chess is so amazing and hard. this is great.
~1500 - i'm really starting to get it! i can beat most people i know easily. i love studying this complex game!
~1800 - this game really isn't that hard. i can beat most people at the club without trying. really I think the only thing separating me from Kasparov is just a lot of opening prep and study
~2300 - omg this game is so friggin hard. 2600s are on an entirely different plane, let alone a Kasparov or a Carlsen.
Magnus Carlsen - "Wow, I really have no understanding of chess." - Said without irony after playing some game and going over it with a computer on stream. A fairly frequent happening.
It's the senescence that makes the fall, more or less, inevitable. Warren Buffet wrote about it in his final letter to shareholders [1] : "When balance, sight, hearing and memory are all on a persistently downward slope, you know Father Time is in the neighborhood."
The tale of computers is even more absurd. The first programmable, electric, and general-purpose digital computer was ENIAC. [1] It was built to... calculate artillery firing tables. I expect in the future that the idea of putting a bunch of solar into space to run GPUs for LLMs will probably seem, at the minimum - quaint, but that doesn't mean the story ends there.
[1] - https://en.wikipedia.org/wiki/ENIAC
reply