The amount of data they must have at this point running so many of those raptor engines has got to be insane...
at least 300+ engine launches now -- wow.
Five years ago SpaceX reported that they had 30000 seconds of test firing time on the Raptor, over 567 engine starts. Since them the program accelerated dramatically. Well over one thousand engines had been produced, and on an average day at McGregor test facility the Raptors are fired for about 600 seconds. That would give about a million seconds over five years. That's a lot for any engine development program.
> SpaceX Rocket Development and Test Facility, McGregor, Texas
> SpaceX calls the facility the most advanced and active rocket engine test facility in the world, and said that by 2024, over 7,000 tests had been conducted at the facility since it opened, with seven engine test fires on a typical day. Despite its low-profile compared to the company's other facilities, is a critical part of SpaceX's operations, and company president and COO Gwynne Shotwell maintains her primary office in McGregor.
A typical test stand would have maybe a thousand channels of relatively slow data (pressures, temperatures, flow rates, valve states, etc), and maybe up to a few hundred of channels for essentially audio data from vibration sensors. This amounts to sub-gigabit per second data rate overall.
If very high speed video / multiple video cameras are used, this could generate massive data rates, but unless something interesting happens it is not clear how important this data is.
In flight, the telemetry data rate from the entire Falcon-9 used to be measured in megabits per second per stage, plus the video stream. It was not a huge amount of data. Presumably now with Starlink they send a lot more telemetry from Starship, but in flight the engines typically have far, far fewer sensors compared to the ground testing.
Tons of CEOs right now keep saying “young people need to learn how to use AI to be successful” and also “we aren’t planning to hire any new college grads due to AI”.. so which one is it.. seems everybody understands the super pro AI CEOs want to lay off nearly the entire company and run it on skeleton crew with a ton of AI and get ultra rich. While “some other” companies should totally hire lots of young people but not them.. where does that end?
It's worse than having to pick one position or the other: both are hypocritical, and both are deceptive hype mongering.
Layoffs credited, or blamed depend depending on your point of view, on AI are mostly a product of herd mentality. As for the advice to learn how to use AI, that's advice that suffers from internal inconsistency. If AI is so embodying of human expertise, why does one have to learn the correct way to use that expertise?
Totally agree — if AI tools are already or nearly at the point where you can say “write a program to do X” and it does it, that’s like telling people they need to learn skills to order something at McDonald’s. The goal is for the barrier to entry to be basically zero. Oh sure today there are things like “I made a claude.md file that does this and I wrote a really clever prompt!!” But the goal is for that work to be deleted as well — where is the magic skill that is / will be needed?
I use coding agents every day for non-trivial projects. But I can definitely say that a prompt of the form "Write a program that does X" will earn you a git rollback of a mess, unless what you want has been done a bazillion times before.
Peak efficiency in using coding agents is a weird balancing act at this point in the development of coding agents: being too incremental and detailed is inefficient, but if you let it rip on a task with multiple sub tasks you have to be ready for the coding agent to get utterly lost while providing you with only hints at what made it to lose its way. It's like an inexperienced intern with a high opinion of its competency.
LLMs trained for coding are most productive when pushed to their limits, but that's where they start to fall down.
Both? The messaging from the last 30+ years has already been that the only way to be successful is to develop your own capital, not to get a job. He's saying that learning how to use AI will be essential in developing the next generation capital.
> run it on skeleton crew with a ton of AI and get ultra rich...
I don't get people who believe this. Why would an AI company provide a service that someone can sell at 10x the price, mostly unchanged? Why wouldn't the AI company sell it directly?
Wholesale vs retail. I think the highest value, biggest markets will have products straight from the AI labs (ie legal review) but there's a lot of "last mile" type stuff that it's probably not as economical for anthropic to care about but maybe for some other company.
if they were selling crack cocaine, he would say “young people need to learn how to use Crack Cocaine to be successful”.
There's nothing to learn, just some CEOs trying to get you hooked on their product and a bunch of hucksters trying to be the number one "AI thought leader"
And if you want to divide up the world along a couple of axes into job usefulness and how much the worker cares, then the author of the article was in the “useless job and cared”, while you are in the quadrant of caring and have a useful job best I can tell — which is great. Some people have useless jobs and don’t care much about their job but are happy to get paid. If you for example are a teacher of small children who depend upon you and you basically don’t care then that’s problematic while having a BS job that you care about a lot about is probably mostly an organizational / bad management situation.
At this point if you have cash or compute credits laying around in the tens of billions, better to hedge your bets than to find out the winner that took all was not you.
Unless none of the current crop of AI companies is “the winner,” either because a newcomer appears or the craze fizzles… in which case have $40B in the bank seems superior.
Do we know who is funding this? is this one of these things where Meta doesn't want the responsibility for this, so they are pushing to have the OS have the responsibility or something like that?
They also added this page since I posted that comment: https://web.archive.org/web/20260411112604/https://tboteproj... where they claim their website is "under surveillance" because it got a few thousand requests from Google Cloud et al, most of them to a single page. This really shows how low their standards are.
I share your wariness of the LLM garbage, but I believe the conclusions are correct. This has Facebook's stink all over it. I worked there and know of what I speak.
So we should believe the hallucinations because they sound like something that could be true? Does the LLM in the middle somehow makes it more trustworthy than if GP had just shared their own pattern-matching conjecture?
No. I think LLMs are garbage. Separately, and unrelated: I think Facebook is behind these bills. The LLM may be garbage and still sometimes produce a correct result.
Yes, it would be nice to know with certainty who is behind these bills. It sucks how much opaque money influences American politics.
Josh Gottheimer's press release[1] on HR8250 mentions the "Meta Parents Network." I don't know what that is, but it does have "Meta" in the name.
Buffy Wick's noise about AB1043 claimed it was passed with the support of tech companies. I have spoken directly to one person close to AB1043 who told me Facebook argued against AB1043. I have doubts. But if true, I suspect they were not arguing in good faith and had ulterior motives.
In the end, no matter who is secretly lobbying for or against age verification bills all over the planet, the bills are terrible, and we should fight them.
I was relieved to hear it was an emotionless mega-corporation catalyzing this, and not a sudden competence of evil bureaucracies in the USA and Europe.
> is this one of these things where Meta doesn't want the responsibility for this
Very likely, given the legal liability they are already facing from the "addictive" court cases that are turning against them. Moving the liability for "age verification" away means they will not also be facing a huge number of court cases accusing them of showing an underage person adult age content provided they followed the law's proscribed "ask the OS for the user's age" requirements.
Also, note that only a few months ago Zuckerberg was in court testifying that the single best place to perform "age verification" was in the operating system of a device. Now, like mushrooms after a long rain, at roughly the same time up pop bills in nearly every statehouse, Congress, even Brazil, that all read nearly identically and that all are so broad as to require "the OS in anything with a CPU do age verification". The nearly identical text in each highly implies a single lobbying entity is behind all of them (it would be quite the coincidence that 50 state houses, plus Congress and Brazil, all write nearly identical bills independently). And the connection back to Zuck's court testimony of "age verification is best done in the OS" highly implies that the single lobbying entity is Meta, or funded by Meta to obtain this outcome.
Facebook. There's a wave of child endangerment lawsuits incoming and they want to head that off at the pass by having governments shift all that liability over to the OS vendors.
Microsoft just force-updated my operating system (despite declining every option and prompt) and the first thing I noticed working differently was it offering, in an OS popup, to "connect" the computer to "Facebook".
These people have root access to all our webcams.
I don't think we can tolerate these entities to continue to exist.
What can we do about it? The major tech firms have nearly all the power here, including quite obviously full capture of government (not just here but other countries as well).
Basically a mass-protest via network packets. Could we argue sending packets to a server is essentially a form of protest protected by speech similar to a public gathering?
How does that help Facebook? They already have plenty of signals to guess their users' age, what would they do with an other one? They are not going to ban children anyway.
The OS should start labeling everybody as a child by default. Forbid Facebook to show ads and any harming content by default. The OS has little less to lose with this approach than FB.
FB etc. may argue "device says this user is an adult", even though device may say that only because the parents don't set up separate user accounts e.g. shared family iPad, or because the kids being more tech savvy in the first place like we all were when I myself was a kid.
Every one of these age assurance laws basically says:
1. The OS vendor must provide an age bucket using the minimum amount of data necessary
2. App vendors (i.e. Facebook) must use the OS vendor's age buckets to determine age
The idea is that the next time Facebook gets hit with a child endangerment lawsuit, they can say "Well, we used the age buckets the government told us to, and they said the plaintiff was 18+, so we're not liable".
This, of course, assumes that most social media and Internet regulation will continue being targeted at children only, both because courts are reluctant to enforce 1A on laws that censor children[0] and because the current political class actually benefits from the harms Facebook does to adults. Like, a good chunk of government surveillance is just buying data from Google and Facebook.
[0] The root password to the US constitution is "th1nk0fth3cHIldren!!1" after all
with age requirements for use of social media, Meta faces tremendous liability in many countries if they cannot do the verification correctly.
they don't want to do it, nor face the risks, so they'll push it to the OS.
they also know that banning under-16s means a huge market will be gone, so they want an easy-to-bypass OS fix. if their tween market gets around the hardware and OS it's not their problem, but Meta can't it if it's on them.
in other words lets annihilate the free internet and maybe democracy so we can lower our risk profile
If something is codified in law, they can comply with the law fully, and yet not have any real backlash from users. This can also shield them from many lawsuits. Conversely, if they start ratcheting down age-verification on their own, users will become quite upset. If they don't ratchet it down, then... as you can see, potential lawsuit.
And this isn't just about LLMs, once the concept of "a platform is liable for harm" happens, it's about everything. Including content other people slap into an app store. And the US has been talking about section 230 removal, countries around the world are reducing such exclusions, so the wind is blowing towards even more liability for platforms.
If you look at Google's recent moves to identify all developers prior to install on Android, there may even be some of this in that. How can they ban someone from publishing illegal material, or material Google will be liable for, if they don't even know who the publisher is? They'll just slide into a new account.
(Note, I said "some" not "all", there is often not just one reason for an action)
So I suspect that the push is from all online platforms of any size or scope. It will shield them, protect them from liability, whist at the same time redirecting user ire at the legislation, not them. HN types might still brood, but the average person won't have insight. "Protect the children" as a reason works for the average person, it works very very well, and really, that's what a lot of these lawsuits are about.
So I point back to such lawsuits as the start of all of this. And I see it as why there is a push from Apple, Google, Meta and so on. And simply because I'm saying "big corp wants this, not just Meta", doesn't mean I'm saying "Meta isn't doing anything".
Meta can be pushing this, hard, whilst at the same time every other large corp can be working towards the same outcome.
It must be OS responsibility because that’s the only place that allows the next step.
Everyone is so concerned with kids pretending to be adults, what about adults pretending to be kids? Any service that has any kind of private chat or picture sharing option will be a playground for “verified” kids.
Next step, “we must go further with the verifications until everyone is verified everywhere”. This is where the OS part comes in. Wish it was sarcasm.
MS is ratcheting up the 'mandatory Microsoft account' on Windows, probably for this reason. The 'identity strongly bound with the device' stuff on corporate devices is being tested and secured in that environment, and it is almost certainly one step from being forced onto non-corporate devices, once they 'have to' by law.
The "children's version" has to be engineered to assume some adult users anyway-- since you're going to have some types of helicopter parents logging into the same platforms the kids are on to make sure they're all right. So the threat model of "what if a paedophile gets a Club Penguin account" is already wargamed out.
In many cases, this consists of dramatically limiting user-to-user comms, hyper-aggressive filtering, sometimes even to restricting to pre-canned messages only. (I'm sure someone is already encoding morse code ethnic slurs into patterns of friendly gestures, but that's another story).
... which is entirely consistent with his statements.
Here are more of his own words from the same letter:
> And what country can preserve it’s liberties if their rulers are not warned from time to time that their people preserve the spirit of resistance? Let them take arms. The remedy is to set them right as to facts, pardon and pacify them. What signify a few lives lost in a century or two? The tree of liberty must be refreshed from time to time with the blood of patriots and tyrants. It is it’s natural manure.
The blood of patriots and tyrants. He never expected rebellion to go unchallenged, he was advocating that we should maintain the spirit of rebellion as a guard against tyranny.
How long before there is a "we've detected your account has been used multiple times to re-setup a phone.. we've re-enabled the Google Nanny Safety mode.. also we've locked your google account just in case.. "
I mean other than hackers, who has needed to factory reset their phone more than once in a year you must be doing something shady... right right?
To be accurate, this is originally from Hastings 1955, Princeton
"APPROXIMATIONS FOR DIGITAL COMPUTERS BY CECIL HASTINGS", page 159-163, there are actually multiple versions of the approximation with different constants used. So the original work was done with the goal of being performant for computers of the 1950's.
Then the famous Abramowitz and Stegun guys put that in formula 4.4.45 with permission, then the nvidia CG library wrote some code that was based upon the formula, likely with some optimizations.
I ran this down, because I have a particular interest in vectorizable function approximations. Particular those that exploit bit-banging to handle range normalization. (Anyone have a good reference for that?)
Regrettably, this is NOT from Hastings 1955. Hastings provides Taylor series and Chebyshev polynomial approximations. The OP's solution is a Pade approximation, which are not covered at all in Hastings.
When you say "this is NOT from Hastings" I had to double check my post again -- I guess you are saying that the Pade approximation is not from Hastings, but the polynomial approximation that the OP referenced from nvidia from A&S and ultimately from Hastings, definitely is in Hastings on page 159 -- I think you were referring to the Pade approximation not being in Hastings, which appears to be true yes. In the article it is interesting that the OP tried taylor expansion and pade approximation, but not the fairly standard "welp lets just fit a Nth order polynomial to the arcsin" which is what Hastings did back in the day.
I imagine there is a process in place to allow cities / states / communities to place the cameras on polls. If Vegas somehow got around public comment process to put these on poles, then what would stop any random company from requesting to put their own camera there? Like lets say a motivated individual went through some process to put a camera on a pole someplace near somebody that would definitely make the govt official / flock exec etc nervous, what is stopping them? It sounds an awful lot like Flock is basically going to town's and saying "we will put up a bunch of cameras in a bunch of places" probably based upon algorithm's etc. How do they decide where these get put, who gets to decide that? Why can't any random company request to put up a camera on a random power pole? After they give the map to the govt officials, do they get a chance to say "oh this one by my house, can you move that?"
So using Stefan-Boltzmann equation if you have a 1m^2 surface at 100C you can radiate about 1kW from that surface -- assuming both sides radiate that, then lets assume it is double. Assume each blackwell chip + support electronics etc needs about 2kW of power to run. So each 1sq meter of say a copper plate is needed to cool 1 blackwell chip. So if you have some way to make some massive radiators that are basically giant plates spanning thousands of square meters, then you should be good. the Stefan-Boltzmann equation is proportional to the 4th power of T (in kelvin), so if you can somehow manage to use a heat pump for the heat from the GPU's into your heat sink such that you could run your radiators at a much hotter temperature, then the blackbody radiation that they put out dramatically goes up. So cooling is quite challenging but not impossible. (I also neglected importantly that you would need to use the giant solar panels as a sun shade for these radiators otherwise they would be pulling in heat from the sun)
For power, you need to somehow manage to generate all of the power that you would need to cool. So the most logical would be some huge solar panels -- assuming you could use similar tech to the space station, you can get aroudnd 100kW from those solar panels -- assume you can do say 10X better somehow, then now you have 1MW of power.
Unclear what the goal here is -- if the idea was doing this for cost, it sounds super unlikely to pan out -- if they want to put a datacenter in space such that nobody can tell somebody what to do, it would seem just as easy to go hide a datacenter in some random far flung corner of the world in a bunker. Seems just like a great way to light some money on fire.
Note all of this is mass that currently needs to be launched from Earth at significant cost - it is indeed nice this cost if finally going down thanks to partial launcher reusability (and hopefully full reusability soon as well) but I really don't see this making any economical sense unless a lot of this mass eventually comes from in situ resources you don't need to lift to orbit.
Also about the radiators - ideally they should radiate into empty space. If there is something in the way, like parts of your station or other radiators, then it will heat up - reducing effectiveness (you will have to remove this heat again) or even making stuff overheat.
At least I don't think you can realistically overheat the Sun. :-)
But I guess if you make the radiators reflective and hot enough they should still work to a even at Earths orbit in full sunlight ? Well, this is already in a "calculation needed" territor.
probably something like a stirling engine + working fluid going down tubes in the plate, it becomes worth it to develop silicon-on-insulator GPUs and other weird technologies that run at higher temps
Yea I’m super curious if you could build a heat pump to move the heat from the 100C GPUs to concentrate all of the heat into a blazingly hot radiator — and how well that would actually work.
Yea in an ideal world there would be a legal construct around AI agents in the cloud doing something on your behalf that could not be blocked by various stakeholders deciding they don't like the thing you are doing even if totally legal.
Things that would be considered fair use, or maybe annoying to certain companies should not be easy for companies to just wholesale block by leveraging business relationships.
Barring that, then yea, a local AI setup is the way to go.
