Dealing with minors has lots of rules and reglementation, for good reason, and usually the justice system of functional countries don't allow you to bypass those with "don't know, didn't ask", because duh of course not.
Using kids data to train? No no. Talking to kids about certain matters (it's still fairly easy to make those model go into drugs or sex or suicide or whatever). No no.
Usually I would very much agree with you, but those things are not deterministic so if that's an issue for you you're probably not making the right choices.
That's a narrow take. Non-deterministic doesn't mean random; workflows can be reasonably validated and consistent to some known degree.
I work for an education department that serves a chatbot for students, and model changes go through painstaking content safety reviews. I initially assumed it's just a bunch of bureaucratic paranoia. But every other model upgrade has a measurably different adherence to the existing system prompts about not talking to the kids about sex and drugs and mental health issues.
I work in the same field for one of my company, in europe, and if you're not self hosting sorry but your worries are not something I can accept because models are very much not reliable on that front, let alone when you let the host decide HOW to serve a model (ressources allocated, different version of the same model, etc ...).
I'm not being a d**, just saying, the problem you have is something that I have faced EXACTLY, and at least here it's not working until you host in house or remote but on raw hardware. Otherwise it keeps having subtle changes, and you will notice no LLM API providers has guarantees about these.
It's not that you're a d*, it's just that you lack any kind of nuance
There's a whole spectrum between self-hosting open weight models and having a cloud provider swap models from under you
Should you self host a model if want to maximize predictability to the limit? Yes. Does that mean it's wrong for someone hitting a model on API to expect that it won't switch to a completely different model under the hood from one day to another? Probably not.
Sure, but if you're working in a field where that limitation is not "because you like it" but "because you have to" it doesn't matter. If you cannot assume it to be true, then you have to assume it isn't.
If self-hosting is the only solution you can think of to maintain model stability, it's no wonder you missed the point. We already do all those things you mentioned, even with cloud providers. Model versioning exists for exactly this purpose, just like it does with any software dependency. We can specify sonnet:v1.23 and decide if and when to upgrade.
But even if we only asked for sonnet:latest, the last thing we'd expect is opus. Model names should be indicative of breaking changes, and change management doesn't just go in the bin because of non-deterministic tools.
The user cares about the distribution of outputs. That distribution is structurally determined by the distribution of inputs (i.e. prompts), the weights of the model, and (these days) the dynamics of the harness guiding successive generations.
The only way to characterize whether a choice is 'right' is to characterize the output distribution (i.e. evals)! Changing the underlying weights necessarily invalidates whatever characterization may have been done. One may assert that one's harness regularizes outputs back toward the desirable distribution, or one may hope the different weights induce a sufficiently similar output distribution.
But no, one should not be completely agnostic to the choice of weights just because there's some nondeterminism.
They’re nondeterministic at a fine level, but can be “deterministic” at a more general level: e.g. you might know that one model will always return properly formatted json when asked. That might not be true of the replacement, even if it is in general “better” and cheaper.
Just the risk of such a thing means regression testing every time you update the model, and you want to be able to run that testing on your schedule rather than having it forced on you.
> They’re nondeterministic at a fine level, but can be “deterministic” at a more general level: e.g. you might know that one model will always return properly formatted json when asked. That might not be true of the replacement, even if it is in general “better” and cheaper.
This isn't true. Even Sol messes up JSON formatting for me on occasion.
Do not delude yourself into thinking these things are reliable. They are not.
Is nobody using structured outputs? They use constrained decoding at the generation stage to ensure the probability of tokens that would break the format are set to 0. I kinda figured everyone was doing this at this point.
Most serious providers are now supporting structured outputs in a reasonable way for all model configs. But for example on ollama structured outputs are still incompatible with tool calling and with reasoning
If you use a large enough volume, you will know this isn't fully reliable. You might get json, and it might not match what the model actually sent because the last layer cut it up to match what you want. At the end, not json, or json but not really matching what the model wanted, it's sort of the same issue: when you use them you HAVE to assume they can have a brain fart. That's fine, just code around it.
We use structured output and To my knowledge it has never failed (millions of data points). There seem to be two classes of people: those doing productive work with LLMs, and those who only get replies insulting their mothers…
Models are not deterministic, but they do have a flavor. When that flavor changes it can change the nature of output in a way that is undesirable.
Sort of like shooting a rifle - where the bullets hit is (to some order of magnitude, no philosophizing please) non-deterministic, but different very similar rifles will group differently and need to be appropriately adjusted to hit anything.
Yeah but then it doesn't matter if it's "the model changed to another one" or "the model changed but it's the same name". Point is, you're not hosting it, as far as you know it can change at any moment, build around that idea. Is that great no, is that ideal no, that's why I self host (I include actually renting online the capacity and hosting the model myself on it).
You've added a similar response like 5 times, but I don't see you adding more information.
Yes - model hosts can do nasty things to you aside from changing the underlying model. That doesn't mean it's cool to have them change the model automatically.
Yes, it would be preferable to have complete control over your model serving, and no - not everyone is in a position to do that themselves.
Often via api you can pin to the specific snapshot. Yes the model provider can screw you over - it's physically possible for many vendors to screw you over, that doesn't mean that's not a dick move and that you can't expect/push for better behavior
That flavor profile is known -- it's typical behavioral distribution is somewhat understood (and, often, common failure modes addressed). If JSON breaks about 20% of the time, and that drops for 2% or blows up to 90%, it can drive all sorts of issues (not the least, costs for retries).
Nope, strong disagree. The model is one small part of the process harness; behaviors are usually routable with expected propensities. Unexpected model changes avoiding change management messes with monitoring and observability thresholds. Stochastic controls are a real thing when you have your distributions defined; your workflows on a new model will throw that expected prior out the window.
I think it depends on the workload, which put you in the right (in some case, it's dangerous to do that), but for me even for very clean always correct path I always assume a model can, at time, have a vector brain fart (because it keeps happening).
Also in principle it's similar to Anthropic downgrading.
Personally I use the basis that if I don't self host (I include remote host, but that I pay per hosting nor per model or api), it can change behavior without me asking. But they shouldn't, but it doesn't matter that's what they do.
Crossing the street and Russian roulette both have non-deterministic risks of injury. And yet I would be bothered to find out that that on my way to work, I was playing Russian roulette by surprise.
Yes and if you decide to play not using a game rulebook but a website that call it "game A" you can't be shocked if "game A" switched from one to the other, even though the doc said opposite yesterday.
That would absolutely be a dick move on the part of the website and would confuse users.
Also, in this case, the game name is not “Game A” but something like “Deep Seek v4 Pro”, which they have previously chosen to use to describe Deep Seek v4 Pro, not Deep Seek v4.1 Flash.
In deterministic scenarios, this kind of switch can sometimes actually be safe, because we’ve used various engineering techniques to converge from non-deterministic behavior to deterministic decisions. On the other hand, in scenarios that are inherently non-deterministic, the impact of such a change is much harder to predict, so we need comprehensive evaluations to assess the extent of its impact.
These models have specific behavioral characteristics trained into them from reinforcement learning and prompts optimized for one aren't guaranteed to transfer to the new generation. Think if the difference between gpt 5.4 and 5.5 and then 5.5 to 5.6 for example. 5.5 was "better" than 5.4 for struggled more across compaction boundaries and needed much more precise instructions before 5.6 sol recovered some of 5.4's ergonomics. All from the same lab but each model was trained with specific behavioral patterns that were basically product decisions. I would be quite annoyed to find that a model provider was routing a promt optimized for one model to a different one, especially for a dumber/cheaper non frontier model that's not going to be as good at just figuring out what you meant
It’s like being told one day that one of your teammates will be replaced tomorrow by another teammate who’s more capable (has a higher test score), regardless of how long you’ve already worked together and gotten used to each other.
> Also there was a little outcry that the original artist just received the royalty free platform usual ridiculously small one time payment.
I mean, if I tell you I'm to sell something to you for 5 dollars and you will be allowed to do this or that with it, and you give me the 5 dollars and then do this or that, it's a bit ridiculous to then have an "outcry" that you were successfull at doing this or that so you owe me more.
Not trying to offend but artists willing to make bird logo for cheap are a dime a dozen, which is why those site exists in the first place, and they didn't care about which one exactly which is why they went on that site to buy it; it's ot like there was any vile or unfair action being done there.
I just want to point out that if there was an outcry it wasn't by the actual (recently deceased) artist, who, according to Wikipedia, was happy to see his creation get this kind of visibility.
Personally, when companies (or movies, or albums, or books, or games, etc.) become worth enormous amounts of money I think there should be bonus royalties for everyone involved even if they sold their labor for a flat rate.
No, I don't have a good plan for how you'd calculate the distribution. But it would make windfalls more fair to the people that made them happen. And that kind of mechanism could also help fund many open source projects.
Ah, come to France, maybe that explains my "sure of myself" answer in the parent comment. 0.0*01% + "à-valoir" is very common and cover the situation described here (eg you get paid a paid royalty, in advance, if the revenue ever exceed you are owed it, but if it's always under you never have to pay back).
Sort of a "I pay you 100€ but if this turns out to be a billion euro thing you get you gold ticket too".
This is right, and the second half is what bites. Standard stock licenses generally prohibit use in a trademark or logo outright, and the extended license that permits it still gives you no exclusivity, so the image stays on sale to everyone else.
That is why logo marks get commissioned with an assignment of rights instead of bought off a stock site. You are not paying for the drawing. You are paying for nobody else being able to use it. A company that grew into a global brand on a non-exclusive stock bird got lucky that nobody claimed it first.
If a majority of people think a thing should happen, then by definition that thing should probably usually happen.
If the structure of the moment (societal norms, laws, etc) doesn't provide for it, then the outcry is the initial stage of the conversation and process where we collectively decide whether or not the structure needs to be updated. IE, figure out some new laws or legal principles to apply existing laws such that the thing the majority has decided should happen can happen.
Everything you value for yourself and consider normal today, was only the result of some outcry at some point, and at that time there were people going "oh so dramatic the outcry just ignore it"
And similarly, when an outcry is worthless, it's valuable to tell the people involved that they should shut up. Outcries don't have to be entertained and they also don't have to be tolerated.
The genesis of jellyfin is as a fork of Emby by part of the developper who saw the need for finance of Emby to start driving choices being made in the project rather than user and technical needs and merits.
Part of the team needed it, part of it saw that as something that shouldn't go that way, so they did what you're supposed to do in open source and forked. Ergo they are very much against doing that for that specific project.
The genesis of jellyfin is as a fork of Emby by part of the developper who saw the need for finance of Emby to start driving choices being made in the project rather than user and technical needs and merits.
Part of the team needed it, part of it saw that as something that shouldn't go that way, so they did what you're supposed to do in open source and forked.
So it's not really a surprise they don't go that way.
Congratulation and great news ! For Europe, but also the world, space access is one of those frontiers we can (and I believe will) see truly opened in my lifetime and I find that truly awesome.
Interestingly there still seems to be a difference in approach between "few launch, expected to go well" in europe and "many launch as trial and error" in the US, with other countries falling somewhere in between those two limits, it's often seen as a threat here of "if we don't do more more often we won't make any progress" but ignoring the economical side (and I think it's fair, given than even with massive subsidies SpaceX is not financially viable until it feed itself through Starlink) it's been a remarkable series of years for new European vehicules launch success.
Also one thing I have to point:
> Europe now has sovereign access to space.
I understand what is meant here, and of course said company need to push that forward, but French Guyana is full french territory, it's not even a foreign territory. It's part of the EU (known as the outermost region of the European Union), EU law applies, it's citizens are EU citizens, EU citizens from other countries can freely move and work and live there ... It's technicaly more EU than Norway.
But less Europe than Norway, sure. Still, on the context of "sovereign access", I would argue both side of that coin have their attributes, and for long term sovereignty I would prefer EU territory instead of "inside-europe-Belarus" for exemple.
Another and more immediate concern is that anything which contains subcomponents/subsystems or parts from US origin companies runs into ITAR and MTCR [1] issues with the US federal government. There's a number of very complex European origin final aviation/aerospace industry products which are laden with US-origin components. I wonder about exactly how "non-US" this Isar rocket is.
I would put it as a much greater likelihood that the US would decide to mess with export-control permits for various subsystems than a theoretical scenario where they begin intercepting western European flagged cargo ships on the way from Europe to French Guiana.
Agreed, the moment we allowed MBDA to put some american component in our (French) missiles made with european allies (Scalp, Meteor), the US blocked our sales to Egypt because they wanted to sell their own instead, using ITAR, forcing us to fix it for the french version.
The US has repeatedly proven that ITAR is for them as much of a military thing as it is an economic one, and that they're willing to use it against allies. If we refuse to learn that lesson, it's on us not them.
> The US has repeatedly proven that ITAR is for them as much of a military thing as it is an economic one, and that they're willing to use it against allies. If we refuse to learn that lesson, it's on us not them.
I'm actually surprised anyone ever thought it was not a military thing.
Maybe I phrased it wrong, it's not my native language ? I meant it "it's not just a military restriction thing when they think there is military danger, it's an economic thing for when they can take advantage over a seller, and who the buyer is doesn't even matter".
I mean it's expected behaviour in one sense but in another it's becoming a bigger geopolitical risk to rely on the US for defence components/technology in the current world, the risk always existed, it's just shifted more one way.
There are a whole bunch of military suppliers across Europe who are in the same boat as MBDA.
Simply put, the US as an ally is less predictable than previously and that has knock-on effects across every sphere.
I think Canada might recognize that it needs to be a much faster process than anyone expected. Europe is getting some pretty strong signals and should be accelerating the process. Ukraine is obviously already in a full war footing to accelerate as fast as they can. They are pulling Europe along simply because the U.S. has abandoned them and walked away from almost every promise they ever made.
How utterly stupid any individual action is will be a matter of discussion for decades. But the pace has moved towards urgent for at least half a dozen prominent countries.
ITAR and EAR were built to keep U.S. military and dual-use technology from strengthening countries that could use it against the United States or its allies, like Europe
Regularly, yes. On a lot of these sorts of products, nothing happens without governments making deals first. Mostly because governments have a need to make sure they get weapons too, not just money for the components they make.
Perhaps should be described as "sale is forbidden by default, but deals are common".
This goes 100% for either completed weapons, or components that can be directly used as a weapon, then you have "dual-use", and the less directly it's a weapon, the more free you are to do what you want.
Can you name any examples? European countries and the EU as a whole don't have an equivalent of the US's ITAR 'See-Through' rule, which lets the US control the usage of a system that has any US components.
I'm not aware of any sale of a US weapons system containing European parts being blocked by any European government. It has been done within Europe, but only for major components, such as Germany's blocking at one point of sales to Saudi Arabia.
But this is the one-component-is-the-weapon, with everything going in one direction version of such a rule. An F35 has 300,000 components from 19 countries, so you can imagine what a maze the agreements for that one are.
It's also not just about the weapons and the danger they pose. It's also about jobs. Know-how (both spreading and restricting know-how). And even the traditional "synergy" (you've designed helicopter blah, hence you make radio component so-and-so, so how about you manufacture NATO standard blah radios and antennas) and the occasional monopoly (it's kind of secret but there aren't many companies that can make the obvious alloy needed for the trust vectoring to name one obvious thing)
Our side is NATO, and we make weapons that are designed to push the global integrated worldview and human rights ("America and it's allies") that we espouse as an alliance. And this view permeates all weapons. For NATO guns, one, you'll never get ridiculous ammo stores (that would enable you to eliminate a population for example), we all work on these weapons together, design them together, and if you try to use your own ammo, everything will fail. Most NATO weapons are defensive. You aim a NATO weapon at a (small) target, it finds that target and either kills it or it disables itself and falls to the ground, essentially intact. Russian weapons are the total opposite: they're meant to kill and if they can't kill their target, they'll just kill whatever and whoever's nearby. Sometimes up to the point they kill the weapon's operators. They will fire "cheap ammo" and literally come with instructions on how to make it. They are meant to kill on a very large scale, without precision and they are most definitely not safe to use. This philosophical difference applies to everything from machine guns to nuclear weapons.
A quick Google search will find you plenty of European legislation. And you should look at those laws like the Belgian laws here, like they were a company "Values" document. That is how the EU/Belgium (and other countries) react "by default", but specific treaties explicitly create exceptions to those rules between specific parties.
So like this: exporting guns and ammo is certainly not something any country allows their companies to do, but most/all countries make specific exceptions in treaties.
This just happens to have the advantage that it makes it extremely easy for politicians to deceive voters about weapons. "Belgian law says it's illegal to export weapons". 100% technically true, and totally misdirects people into believing Belgium doesn't export weapons, which is absolutely false.
Meanwhile, Belgium exports ~50 million euros worth of weapons. But Belgium is just an example, it's not like the Netherlands or France or (definitely) UK, or the EU as a whole are different.
You missed the point. The example asked for was this:
"I'm not aware of any sale of a US weapons system containing European parts being blocked by any European government."
That would be something like "Belgium says the USA can't sell F-35s to Israel because the wings are Belgian-made" similar to how the USA prevents ASML from selling their machines to China.
Oh that's what you want. In the clusterfuck with the Turkey involuntary withdrawal from the F35 program half the treaties were renegotiated (you see, can't have any Turk produce weapons unless he's a "genius", that would not be fair. In other news: the Turkish definition of genius is mostly being the son in law of Erdogan [1])
I believe this agreement is part of the restructuring after that cluster... Note that the agreement is both about production and buying. [2]
"Finland has required extensive domestic participation, with industrial cooperation valued at a minimum of 30 percent of the contract price of procuring 64 F-35A fighters. Under the agreement, Finnish firms will produce the F-35’s forward fuselage structures, including components intended for export aircraft. The agreement also includes ..."
As to how the EU is involved: https://defence-industry-space.ec.europa.eu/eu-defence-indus... (ie. that's where you go to ask for loans to start a factory for making F35 components, they have a page listing the deals ... somewhere). Note that this is one of the EU institutions that isn't in the budget (well, the loans aren't). They're not spending tax money to make those loans, they're printing money to make those loans.
So the Finnish government is making this deal, including details on export (and this is obviously weapons exports).
How is it an example of European governments preventing American weapons from being sold? Are you saying the Finns unilaterally decided that Turkey can't have F-35s, even if the USA wants to sell F-35s to them?
You misunderstand. Everything is prevented by law, with the purpose to make any change in production involve negotiations with the government, who then negotiates weapons/jobs/local production/... Without those negotiations, and the resulting exceptions to the law nothing can be sold, even dual use items are restricted.
And, of course, the other side of that coin: weapons export regulation is Europe is not something that is meant to actually prevent weapons export. Rather it is used like a company policy: it makes sure everyone regularly has to give the government something they want.
Which is, to put it mildly, not how politicians present those laws.
No, you misunderstand. You keep talking about things irrelevant to the key point.
Look at the political pressure the USA puts on ASML in the Netherlands to prevent them from selling certain products to China because those machines contain some American parts.
The question is explicitly when the reverse has happened, not some general info about European weapon exports or legislation.
Things like Wassenaar Arrangement and MTCR(that applies to a lot of hobby stuffs I'm interested in) are multinational and global. The US implementation is the most painful, but they're like non-Linux unices, sameish in concept and not like first-worlders get free passes.
(And these regs aren't complete nonsense; the technologies regulated by these are useful for guided munitions. It just so happens missile technologies are fully commoditized by now and also happens to be immensely useful for totally peaceful purposes too)
These, again, are export control restrictions at the point of sale and transfer, and the US follows them too. It's not the same thing as what we're discussing regarding the US's changing requirements regarding systems that embed its components.
Note that dual-use items (which Wassenaar regulates) and MTCR are not the same as the US's ITAR 'see through' rule which it uses to block third party sales of systems that contain even small numbers of US components.
For dual-use items, the EU's Directive 2021/821 governs things and broadly authorises exports to the United States through EU General Export Authorisations (EUGEAs) that don't require individual approval.[0]
In theory the monroe doctrine, as it was originally written, requires the USA to remain neutral on existing colonies. It only prohibits the expansion of European colonies in the Americas.
Indeed. The interpretation that served as a justification for US interventionism came almost a century later (the Roosevelt Corollary in 1904).
In its original form, the Monroe doctrine declared attempts by European monarchies to reconquer the newly independent Spanish states to be an unfriendly act toward the US. It did not declare US domination over them, as it's commonly misunderstood nowadays.
Yeah but although the crazies in charge have been very negative toward our continent and my country (Frace), they've also steered very clear away from any seizing claim about any of our territory, despite going against anyone from Canada to Denmark to Spain to UK, to ...
Deep down, we're an annoyingly arrogant little country with delusions of grandeur salivating at the thought to prove to our EU neighbors how strong we meant it when we said we would accept to oppose the US in case of military attack and I think they knew we may actually mean it and not back down when it start (see also: Denmark leak about how France was essentially willing to provide whatever to help defend Greenland).
I'm not saying this in a "we're awesome" way, more like a hard case of the TACO, bullies avoid the ones who punch back. Nothing would be happening in Iran right now if they had not believed in advance Iran would lie down like Venezuela did after losing its leader, though most would say confusing the two situation makes no sense.
Ultimately this is so much lost time and resources for nothing of value, it's so utterly disappointing...
> Deep down, we're an annoyingly arrogant little country with delusions of grandeur salivating at the thought to prove to our EU neighbors how strong we meant it when we said we would accept to oppose the US in case of military attack and I think they knew we may actually mean it and not back down when it start (see also: Denmark leak about how France was essentially willing to provide whatever to help defend Greenland).
And the rest of Europe loves you for that! But we're also very worried if that'll all remain so if Le Pen or Mélenchon win next year.
Believe me, I'm even more worried than you. I think Melanchon would actually do more damage, as crazy as it seems, Le Pen is incompetent and a sellout, Melanchon actually believes the BS he vomits.
Macron finished the traditionnal left and right wing parties, which were too large and rotten to the core, but it's left us with a large center and two extremes, it's high time we have the normies left and right come back.
If the extremes are unelectable, the center party should split eventually. They'll find something to disagree about.
As long as they are more terrified of losing to the extremists they'll try to keep it together as long as they can. I hope they don't become too corrupt in the process.
> salivating at the thought to prove to our EU neighbors how strong we meant it when we said we would accept to oppose the US in case of military attack and I think they knew we may actually mean it and not back down when it start
No offense to France, I don't really care one way or another, but didn't Macron also say he would deploy French troops in Ukraine? Where are those troops? Surely Ukraine is an "EU neighbor". They're currently suffering from an actual aggressor (Putin), not merely theoretical (Trump).
No, it’s a very strong indicator that a peace deal will never be signed until unconditional surrender. Because concerns about NATO troops in Ukraine is main reason why Russia started this.
So I would recommend making sure to inform yourself properly rather than assuming what you read (or thought you read) is correct
> Surely Ukraine is an "EU neighbor". They're currently suffering from an actual aggressor (Putin), not merely theoretical (Trump).
But as you said, they're NOT EU, not in an official military alliance with France. If you cannot see the difference between the two situation, I can't make it any clearer for you.
If you're saying "we should do the same for those that are out that for those that are in", what you are saying is "being in has no purpose, why take the commitment". I believe your point of view is as wrong as can be.
> The situation in Ukraine, which has been under invasion by Russian troops for the past two years, was deteriorating. The war was getting bogged down. "In any case, in the coming year, I'm going to have to send guys to Odesa," said the president before a handful of guests. [1]
This was 2+ years ago and in the years since Russia has captured even more territory and is now bombarding Kyiv. And still no French troops.
Well, you'll notice how Russia just ignored what Macron said and proceeded with their attempted conquest of Ukraine anyway. Because it's just "game theory". It's just "tit for tat".
You have events completly backward, Russia was already invading when Macron said it. If you can't even get years in the right order you're not really arguing in good faith.
The point is that Russia didn't just turn around and go home when Macron threatened to send French troops to Ukraine. Because everyone knew they weren't actually going to do it. And they didn't.
> Deep down, we're an annoyingly arrogant little country with delusions of grandeur ...
The problem is France used to be the world's superpower, with an economy bigger than the one of the UK. Literally number one.
Then the monarchy fell and all it took was a mere 15 years for that little arrogant man that Napoleon was to get its arse (and France's arse) handed to him in Waterloo in 1815.
15 years to throw away centuries of greatness: what an amazing success the revolutionaries managed to pull. They fucked up so bad they realized it and tried to, peacefully, re-instate monarchy, but it was too late. Charles X was king of France after the Waterloo SNAFU (1824 to 1830) but it was too late to fix things.
It's nice and all, on paper, to want no king. But France is considered to have lost, forever, its place as the world's n 1 superpower when Napoleon lost in Waterloo.
I'm not saying monarchy is better: what I'm saying is that the french revolution coincides with France losing its status as the world's 1st super power.
And France has been arrogant ever since but it went from 2nd to 3rd to 4th to 5th to 6th to now 7th place.
It's a never ending fall.
With the biggest public deficit and the biggest public debt in the entire EU, and the level of primary education going down the drain (university is still doing OK), I fear it's only ever more pain ahead for France.
The country is literally crumbling left and right and communists, siding with islamists (the party of Melenchon are downright commies, who understood that if they were cozy with islamists, they'd get a chance a winning the elections), are planning to destroy the little that's left of France's past grandeur.
If you think those in charge now are negative towards their own country, you haven't seen anything should either the communists (Melenchon) or the national-socialists (Le Pen / Bardella: which are for a national preference with a deeply socialist program) win the next elections.
Besides an aching aircraft carrier France still has the atom bomb and that's it.
No more money. No more education.
France is toast.
P.S: the biggest success, from their leftist point of view, of french intellectuals is managing to present national-socialists as "far right" (as if they were fascists)
Am no expert at French history, but this feels like cherry-picking to me of what happened to France.
From what I read, one of the biggest reasons (besides the defeat at Waterloo) for the downfall of France to Britain was that in the 1700’s, France had wild speculation in the colonies in the Americas, specifically, in the Mississippi Company. When the bubble started collapsing, the Prime Minister at the time (John Law), put a lot of the French treasury into trying to support that bubble, but it collapsed anyway. Now the French government was relatively cash poor, which eventually led to losing a lot of its colonies to Britain, a great deal of unrest in France and the French Revolution.
And I am no history expert, so am unsure of this myself, but in my humble opinion, the idea that monarchy is why France lost its place in the world doesn’t seem like a major factor.
I have no idea what's your point but I have to say I find "The country is [...] siding with islamists" when we were literally the only country willing to defend the freedom to speech and caricature religion a couple years ago against islamism, with Quebec the outlier that sided with us (even though Canada did not).
Overall: I don't think any country should be on top. I'm glad France is not number one, and I hope it remains strong but never get in that position again. Part of why I love the EU is that no one, not even us or germany, is strong enough to fully impose its will without concession. Some would disagree with that, but they clearly don't know their history. So ... I'm not sure what your point is, but I think I disagree with it.
> P.S: the biggest success, from their leftist point of view, of french intellectuals is managing to present national-socialists as "far right" (as if they were fascists)
I wander how "national-socialist" sounds in other European languages, and whether it might have some history of referring to fascist movements. Say, what would the Germans call someone like that? Nationalsozialistische? I don't remember if there is any history about someone who called themselves that being a fascist...
Yeah, "when have national-socialists ever been fascist!" is all I read in GP's post after I got to that part. It's just a brain crash for me (I'm German)
I think it's ridiculous to write that much without realizing the most obvious thing that happened, that France had its birth rate drop first and very quickly, meaning that they lost their demographic advantage against the rest of Europe.
Good thing is that that Trump will never manage to figure out where French Guyana is on a map. Or worst case there is a 50% chance he’ll invade (non French) Guyana.
French Guyana is occupied territory. That's pretty clear. And so it should be deoccupied and the locals will decide whether to use their land as "sovereign' launch port by Europeans.
It's a new world. He who has the biggest guns makes the rules.
I don't like it, but in the Caribbean [probably throughout the world] these days, you really are dealing with the raw realities of power in a way you didn't have to previously.
I doubt the EU has the stomach to confront Trump. And French Guiana is not Africa, there will be no China coming to save them.
On the same day, more soldiers together with the Chief of the Royal Danish Army, General Peter Harling Boysen, started arriving in Greenland, and Denmark announced that a "substantial contribution" of its armed forces would be sent there
On 14 January 2026, Denmark started to send military reinforcements to Greenland[186] to strengthen its military presence.[187] Later that day, it became known that both Sweden[188] and Norway sent military forces to support the Danish Defence in protecting Greenland.[189] The Ministry of Defence on 14 January 2026 announced "an increased military presence in and around Greenland, comprising aircraft, vessels and soldiers, including from NATO allies".[190] After consultations between a group of European countries,[191] Germany announced it was sending a small contingent to Greenland on a reconnaissance mission with other European nations.[192] On 15 January 2026, a French military contingent arrived in Greenland.[26]
- ibid.
Several European leaders including France's Emmanuel Macron advocated using the EU's Anti-Coercion Instrument, a security and trade policy tool known as the "trade bazooka", to target the US in retaliation.[258][259][260][29][261] Under the not-previously-used instrument, the EU may adopt countermeasures, sometimes described as sanctions, including tariffs, restrictions on public procurement, and measures affecting trade and investment.[252][262] Bernd Lange, chair of the European Parliament Committee on International Trade, supported its activation.[263] By 21 January EU countries were described as being increasingly open to using the ACI against the US, with Germany stating it will ask the Commission to explore its use.[264]
French people should be occupied by Guyanans who have no historic connection to France. Apparently that makes sense to some people when it's the other way around.
If that's correct, then yeah, I struggle to find much to be upset about with this list. $106M federal loan against what they've built feels, I don't know, about correct for a telecom. Light even. The tax abatement stuff sounds gross but not surprising.
(There are other forms of subsidy than this list would show, though! So, yeah, always going to be an argument.)
You have to be absolutely blind if you can't even read history going back just a few decades, which makes it evidently clear that we have more motivation to work together than pretty much any continent out there. Doesn't help that we all know that the two world wars we've had so far, has come from our continent, so yes, we have a lot of motivation to work together.
Eh hard disagree there, the way I see it there is no more squabble between France and Poland than there is between Texas and California or between Spain and Romania compared to New york and Virginia. In fact, I would say we have LESS of it.
The only thing we need is to actually stop ignoring the forest for the tree and see that the world moved on beyond our petty local borders and we're lucky enough to have found a system where we're stronger together, to the point where we matter in the world and can't be ignored or pushed aside, while also protecting the core of each of our own cultural societies (langugage, history, ...), and only fools would let that go because they're too close minded to see it.
Must be incredibly humbling experience to not realize EU is a union of independent countries and the US is a federation of states until you work for one of the delegations, wasn't it? Most of us pick up on that even without professional experience about it.
> See the thing is when we studied law in Europe a long time ago we were sold on this dream that Europe would become like the United States
Eeh, what school was this? I think most public schools (at least those me and people near me) attended, it was taught that EU is nothing like the US, as we're all individual countries and everyone wants to remain as such, but we still want to collaborate although we're somewhat different from each other, and been killing each other since forever. I guess that's why second world war and similar things are such a big part of our history lessons, rather than diving deep into the history of US or Australia or whatever.
> It really takes the soul out of you. The way people work in the US and China is completely different. They DO things. Europeans just talk about doing things, and then talk behind each others backs.
For me the difference is between politicians and people actually doing the work. And my personally experience points towards this split existing equally everywhere, in China, in the US, in Europe and everywhere else I've been. The higher up the political ranks you get, the more people like to talk, the less they like doing.
I feel like this is very evident in the US too, looking at how productive and efficient companies are vs research labs living on grants vs federal government vs local governments, the scale is completely different, and same goes in Europe as well, at least from what I can tell.
Evidentially, in the high ranks of the European Parliament, things are exactly the same as higher ranking politicians everywhere, they just like to talk and grandstand and pretend to know what they're talking about. But I think this is true in the US too, or in China, these sort of people just like talking and discussing things forever, for better or worse.
> it was taught that EU is nothing like the US, as we're all individual countries and everyone wants to remain as such
It's written into the treaties that this will not be so. The goal of the EU is "ever closer union" and the people at the top of it routinely talk about a federal Europe in which the individual states disappear.
Individual states do not disappear in a more federal Europe. Otherwise that wouldn’t be a confederation… It’s about delegating some authority and decision making to a common entity, but each nation state stays sovereign. I don’t understand how people are confused about that concept, there are lots of federated entities in the world, with various level of delegation to the federal level. And that works well. The distinctions between states aren’t erased, the culture between states isn’t erased.
You can't be sovereign and also subservient to another power. That's a contradiction! Texas is not sovereign, for example.
The people who call themselves federalists don't actually want a federation. They just use the name because it sounds more acceptable. Read Spinelli's Draft Treaty from 1984 or his Ventotene Manifesto. These documents don't describe the creation of a federation but rather the creation of a unitary state.
> It's written into the treaties that this will not be so.
What exact treaty are you talking about, and what does it verbatim say? I've heard this so many times before, yet no one is able to link to concretely what they refer to. No texts I've seen has called for the dissolution of nation states in the EU or any of the other political structures.
>RESOLVED to continue the process of creating an ever closer union among the peoples of Europe, in which decisions are taken as closely as possible to the citizen in accordance with the principle of subsidiarity,
Now you might say, ah! This text also says that decisions must be taken as closely as possible to the citizens. Yes, the treaties do say that, the EU just doesn't do it. Of course it makes no sense, the claim about subsidiarity is in direct contradiction to the first part of the sentence, you can't be both in an ever-tightening political union and also have decisions made by local governments. In practice the first part is chased and the second part is ignored. EU decisions are taken in Brussels - far away from the average citizen - even about many issues that don't need to be decided there.
If you want to understand the real origins of the EU and where it's going, you need to read the Ventotene Manifesto. Written during WW2 by a group of imprisoned Italian communists led by Altiero Spinelli, it calls explicitly for a communist revolution that would unite all of Europe. Spinelli's document can be found here:
It says things like, "In order to respond to our needs, the European revolution must be socialist" and "By this dictatorship of the revolutionary party a new State will be formed, and around this State new, genuine democracy will grow". It goes on like that in quite some detail.
Although the document is insane Spinelli went on to be one of the founding architects of the EU. He is officially credited as a "pioneer of the EU":
He became one of the first members of the European Commission and was a member of the European Parliament. He was also one of the primary authors of the "Draft Treaty for establishment of the European Union" in 1984 that directly led to the creation of the modern EU itself via the Maastricht Treaty, which imported a lot of its ideas. The Draft Treaty can be found here:
It not only mentions the goal of ever closer union (with no mention of subsidiarity), but mandates it legally in Article 11, which defines "cooperation" to mean control by member states and "common action" to mean things directly controlled by the EU institutions. It then states in plain English that a topic may be moved from cooperation to common action, but never in reverse. It also states that powers go through this one way transfer to the EU based only on a vote of the EU Parliament, with member state parliaments themselves getting no say in it.
So the idea of abolishing national states has always been very visible among the EU's "founding fathers".
> we were sold on this dream that Europe would become like the United States
Sorry but that's just made up bullshit. The EU is not a "United States of Europe", and thinking along that line will ineviatably lead to tons of misconceptions.
> Then you get there all bright eyed and it's just a bunch of boomers squabbling to fuck each other over to get 0.03% tax advantages on peach liquor exports for their region.
Congratulations, you summed up how politics work in all regions of the world.
The European parliament and US congress are both just as united and fragmented.
We disagree internally until we find a common posture that we adopt.
I'm not sure why you talk California public sector instead of something involving multiple states that have differing views, that makes your comparison useless, that's why I say US congress.
Note that your misunderstanding of the situation and lack of knowledge about it is not rare per se, to give an exemple that's why a fair amount of brexiter politician were surprised the EU didn't break itself in internal fight over the UK deal (while they were used to the inside squabble when they were a member, and they failed to notice the similarity with their own parliament compared to their country).
The EU-related structures in Brussels are about as united/fragmented as the US Congress, indeed.
But the national governments are nothing like US state governments. A lot more political fragmentation, plus they still consider themselves executives of sovereign nations with various vital interests and also historic resentments. Few Californian politicians will base their programs on suspicions towards New Mexico. In Poland, "don't kowtow to the Germans" is still a living idea and current PM Tusk is regularly accused of doing precisely that.
The European nations are not unified similarly to the USA because we do not have the common history of being a bunch of immigrant colonies glued together.
What is even your point in all those messages ? If we were to take all of your comments in this thread, and we all agree with no exception to them, it doesn't change anything to the conversation, it doesn't make any point, it does't provide anything interesting, you're just dead set on commenting in a loop that the sky is green
Think about it, what are you saying that adds to a German company successfully launching into space from Norway ?
It's an opinion just like all the others here... every morning before US wakes up there are a bunch of articles by EU readers for EU readers patting themselves on the back for getting something built without US influence.
Inevitably it turns into the same tired discussions that either:
A) Actually, it's not completely EU because it has x% of American parts
B) America Bad
Then, the one person who comes in to give a different opinion is met with your comment. I see it constantly.
That was not a useful metaphor, but by all means try again (if you want) and teach us something interesting (and concrete) based on your unique experience.
> I understand what is meant here, and of course said company need to push that forward, but French Guyana is full french territory, it's not even a foreign territory.
Does French Guyana have a usable space rocket? Cause if not then you are completely missing the point of what "space access" means here. It's not about a spot from where to start the rockets. It's about being able to build them. No matter French Guyana or Norway or central Berlin.
And that's obvious to everybody including yourself, so why you opened the French-nationalist angle here with lots of angry messages with all sorts of politics is unclear to everybody watching.
1. Who is "they"? Isar Aerospace and Arianespace are distinct entities.
2. Isar Aerospace is the first fully private European company (a startup) to do this. Arianespace and the Ariane rockets have heavy state/ESA involvement.
3. It is the first launch to orbit from geographically European soil. French Guiana is politically part of France, but geographically obviously part of South America.
4. Why would everyone have to launch from the same place, in your esteemed opinion? Why are some US payloads launched from California, some from Florida, and soon some from Texas and Louisiana?
5. Are all launch locations equally suitable for all orbits? Are all orbits equally suitable for all missions?
6. Norway is a lot closer to the rest of Europe than French Guiana is.
7. Why was SpaceX newsworthy, particularly on a startup-centric site, when NASA already exists?
> 2. Isar Aerospace is the first fully private European company (a startup) to do this. Arianespace and the Ariane rockets have heavy state/ESA involvement.
To go even further: ~60% of Ariane Group revenue is from the french state, for their nuclear launch system. The whole ariane space rocket is their side show, a way to reduce cost for France by getting the market to pay some of the R&D.
> 2. Isar Aerospace is the first fully private European company (a startup) to do this.
Sure. But that's different from "Europe now has sovereign access to space." don't you think? If anything, a state-owned manufacturer is more "sovereign access" than a private company who can decide who they do business with. (Even though state regulates exports but they could decide where to manufacture things.)
> 3. It is the first launch to orbit from geographically European soil. French Guiana is politically part of France, but geographically obviously part of South America.
Sure. Though Norway has been part of Europe for a while and their space port isn't from yesterday either. A milestone? Sure. Never seen before sovereignity? Um...
I don't have answers to your other questions as I'd ask them too.
To me the "Europe now has sovereign access" line by the CEO is a bit of hype, though technically it is true if you view Europe geographically, as he seems to do.
Andøya Spaceport has been a place for suborbital/sounding rockets, this is the first time a rocket has been launched into orbit from there.
We would be doing it now if it wasn't cutting into frontier companies' revenue.
Eg if you use Claude, you probably want fable architecting, a couple of opus under it managing sub project and sonnet doing the actual function code, because fable coding a "run a query and filter the result" is a massive waste of abilities. But their own sub agent downgrade is limited to one level so if you use fable it will never direct sonnet coders.
Qwen 3.8 27B being so good, so easy to run on premise on cheap hardware and having so little limitations compared to what the US decides can or cannot be done is really eye opening.
The main tradeoff I’ve found with running models like Qwen 3.8 locally is speed. On my hardware, they’re significantly slower than models served through OpenRouter, ChatGPT, or Claude.
That’s been fine for smaller tasks like analyzing a document, but for more involved work like refactoring code the latency makes it harder to iterate.
Yeah for live it's a bit on the slower side though still usable on the 4070 ti and 5080 our usual office desk has here, but I found it awesome running on background tasks rather than on demand, things like analyze this and that with this and that ruleset and applicable law warn me with source and justification for any error or missing part etc, qualify each returns with a ranking (certain with source, probable with reason, suggestion),... With its thinking mode enabled and it's great context size abilities it's really powerful, and each employee basically has their own assistant working in the background checking their files on its own as the day progresses
(the main reason is not just cost, it's data not going out and even mess leaving the EU, this essentially frees us of a lot of hurdles)
Yes, the generic system I built has unsloth/qwen3.8-27b with a 48000 context size running on each employee system with lmstudio, with a simple runner to watch folders and pass it to the llm alongside all context (our company design rules, our legal documents, our contractual rules, our applicable local law and reglementation, ...), return is then mailed to them or added to their personal dashboard event list, depending on their settings.
This is clearly pushing it memory wise and the gpu offloading is only partial but LM studio deal with it automatically and it's being fine even on the 4071 Ti desks (Ryzen 7500F and 32 GB of ram), employees get any feedback in ~20 minutes after they dropped a file (it's much smoother on the 5080 desks obivously), for live it's useless but as background helper it's great and the very large context allows us to fit all the rules we want in there.
One caveat has been to not ask it if everything is ok, but to find what's wrong - but always source and explain it and justify itself, never drown the user in warning in suggestions; goal is to help and provide a second pair of eyes not make them feel annoyed or unsecure.
And I found people to genuinely enjoy something that works for them on their own machine and is not tracked "by the boss", thus the assistant reference, than than a centralized mothership like we also have and they have access to. I also allow them to disable it if they want, I trust them with their work, but a second pair of eyes is always great.
(my previous workhorse for this was Qwen3-14b but it's missing a lot more edge cases)
Accordit to reddit talk, Fable 5.1 is worse than Opus 4.6 and 8B models are smarter than Qwen 3.8 Max, I wouldn't take anything said there with any more reliability than an instagram short.
Using kids data to train? No no. Talking to kids about certain matters (it's still fairly easy to make those model go into drugs or sex or suicide or whatever). No no.
reply