I thought this was common knowledge. DeepSeek’s Wikipedia entry says that they trained all their models on Nvidia chips procured before the U.S. embargo to China on them. It wouldn’t surprise me if they continued acquiring them through, well, less than legal means.
I also read somewhere (not Wikipedia) that they trained on ChatGPT, Claude, and Gemini queries, basically feeding in the output of competitor’s LLMs as training data. Kinda surprised they didn’t run into model collapse problems, but they stole their training data from other people who stole their training data from data collections that arguably stole them from content creators. It’s bandits all the way down, so adding a little smuggling to that doesn’t surprise me.
> It’s bandits all the way down, so adding a little smuggling to that doesn’t surprise me.
Implying it’s *morally* wrong for a Chinese company to bypass US sanctions is hilarious. You really say that with a straight face when even the president admits this is only protectionism?
These export controls increasingly look like "tax".
The White House said the US government would take a 25 percent cut of the chip’s sales, similar to a deal with AMD and Nvidia earlier this year that allowed them to sell lower-powered AI chips to China while paying the US government 15 percent of the proceeds.
> Using them in China is legal in China.
Technically, yes. The CCP, though, wants to incentivize Chinese firms to use domestically-manufactured chips.
> Technically, yes. The CCP, though, wants to incentivize Chinese firms to use domestically-manufactured chips.
This couldn’t be playing out better for Xi. Trump is China’s best president.
I used to think Trump was clueless and being outplayed, but now I realize he’s just looting and couldn’t care less about protectionism or the American worker.
Every single action from this administration can be explained by greed and ego.
> Every single action from this administration can be explained by greed and ego.
I highly agree with you - and that’s coming from someone who can’t stand the Democrats[1]. Things like announcing all of a sudden that he opposes a merger when everybody knows Kushner is involved with a rival bid… it’s too obvious how much corruption is his very operating system.
[1] (I didn’t vote for either candidate for President, but I’m not in a swing state so I’m not sorry)
The lesson from Mamdani is that the only way forwards for actual policy based and anticorruption politics is within the Democrat primaries. These are even run by the state in many states, I believe.
Eradicate the Republican party as an organization, split the Democrats into "normal right" and "maybe a bit left" factions, and see if you can get preference voting in there as well while asking for a pony.
That's Corporatism. It's from the fascist playbook where the state takes partial or complete ownership of private companies. Where does that money go, to some slush fund for the president? The reason for the export controls is to keep our potential adversaries from being on the bleeding edge of frontier AI. It goes against the US's interests to give China a leg up with advanced chips. It's almost laughable, of course, as the Nvidia chips are already manufactured in a country that China claims as their own. If they ever pressed the issue, we could find ourselves without the most advanced chips.
This is not at all what is meant by fascist corporatism, nor corporatism more generally. Corporatism is more about collective bargaining by professional trades, and is not the sense of corporation as used for private companies.
It's the moral imperative of every country to optimize for its citizens' economic prospects.
The CIA (and every other intelligence org.) is literally a weapon designed to operate in the grey area to fit the mandate of the policy makers and elected leadership. Many of the things they do are questionable or worse. In the case of the DoD, they do these things at the behest of democratically elected leadership.
Of course US and China will operate in their own best interests. Of course they will both play chess, both name call, both sanction and impede. When it's not a hot war, it is still a never-ending battle for each country's total economic, soft, and hard power market share.
This is just "might makes right" but modern. I don't know that this is the consistent, wide-held belief that you seem to think it is. Plenty of people would rather our governments not engage in clandestine disruption and undermining of foreign governments.
Competition is inevitable, especially between geopolitical rivals, but we don't have to engage in Minitrue-style "the enemy has always been our enemy" rhetoric.
Do you think countries behave differently than what the parent has said? This has been going on forever, since the first clan of humans fought another, any reasoning other than "might makes right" is a post-hoc rationalization not based in history.
The theory that humans are evil to one another and that this behavior has been going on since the dawn of time has shaky grounds. David Graeber looks for alternative explanations in his book [1] The Dawn of Everything.
We live under economic conditions created by a government that does intervene. And a greater world order established by a hegemon that does intervene.
It would be interesting to see what life would be like today had that not happened. It might be better, it might be worse. Probably a little of both for different groups of people.
As the world returns to multi-polarity, there are signs of increases in violence.
The last time the world had multi-polarity, we had far more wars. Including the worst wars the world has ever seen.
> The last time the world had multi-polarity, we had far more wars. Including the worst wars the world has ever seen.
Citations? Simply saying that World War 1 happened during a time of multi-polarity is just begging the question. Multi-polarity of varying degrees has always been the case throughout human history, and often times single-polarity is achieved only through extreme violence.
American Hegemony or Pax Americana (post WWII until present) is the most peaceful period of human history, despite the myriad atrocities which have occurred during that period, from myriad different parties, including the USA itself
A big reason for that, as far as I can tell, is that if one side has the USA on its side, they're basically unattackable for many places since the USA is so over powered militarily and can project force anywhere
It stands to reason as the USA recedes from the world's stage it will get more violent as more nations stand at parity with their adversaries again. And we're certainly seeing wars cropping up lately as the US continues to undermine its traditional allies, bully adversaries, declare trade wars, and withdraw from agreements.
I do not disagree with that assessment, but maybe one can hope that at some point we evolve past this "us versus them" mentality that we inherited from the savannah? If so, it's worth pushing for it.
I think this is too simplistic of an outlook. The post-Napoleonic balance of powers was not uni-polar, it was a carefully constructed and negotiated settlement by diplomats and politicians who knew the cost of war, and it lasted a remarkable 99 years. There were skirmishes in the interim, but the balance of power ensured that the bloodletting never escalated to the point of continent-spanning "world" war...until it did.
Pax Americana, by contrast, was essentially a standoff between ideological opposites that were equipped with enough nuclear weapons to assure mutual destruction. The choices were clear - coexist or die, and there were many opportunities where we narrowly escaped the second option.
You could point to many possible causes of WW1, but I think that a lot of the causes can be traced back to a hot-headed emperor who desired a larger and more prestigious empire but lacked the statecraft to do so without pissing off nearly all of his neighbors. Looking around at our world today at the number of unserious leaders who govern like a bull in a china shop, I would be lying if I didn't see any similar causes for concern.
This is not news, I had "the multi polar world" in history class in high school in the early 2000s, it's just that the US suddenly realized it and has been blind to the change for a while.
That first sentence feels like one of those fake-deep things that sounds important, but can effectively be used to justify about anything?
Which is to say, in a world that's -- you know -- a society; not screwing over the other guy is often, if not usually, a good way to "optimize your own citizens economic prospects," too.
The very first sentence crashes and burns, because there are multiple moral systems and compasses. Using "imperative" in the context of morals is extra spicy, because it reference a very specific, very strict moral code - The Categorical Imperative.
The CI is, in my experience, not a moral system about personal or group advantage, but about rules the can govern everybody.
> The CIA (and every other intelligence org.) is literally a weapon designed to operate in the grey area to fit the mandate of the policy makers and elected leadership
There is some truth, but this is how you get a crappy-ass intelligence agency.
Good intelligence agencies are focused on gathering intelligence, not performing random tasks that benefit from secrecy.
The US wants to have its cake and eat it too. It will pose around as a democratic and peaceful force, and use these illegal and shady tactics whenever they seem fit. And if you say this, their enablers in the traditional media will label you a conspiracy theorist.
I hate this talking point so much. If you are talking about China, that's just growth. They are also rolling out more solar than the rest of the world combined. While the US is now actively discouraging investing in renewables.
Chinese coal power outgrows renewables still. A Western country with already cleaner energy destroying whatever remains of their manufacturing only to be moved to China and powered by mostly coal is not only treason of its own citizens but also bad for the climate. Feels so good to be "net zero" while importing materialized coal with not much to trade back (other than coal of course).
> It's the moral imperative of every country to optimize for its citizens' economic prospects.
"Moral imperative"? No country was ever created out of a moral imperative. None. Also, no country was ever created to optimize for its citizens' economic prospects. Every country was created by the elites for the benefits of the elites.
> We hold these truths to be self-evident, that all men are created equal, that they are endowed by their Creator with certain unalienable Rights, that among these are Life, Liberty and the pursuit of Happiness.--That to secure these rights, Governments are instituted among Men, deriving their just powers from the consent of the governed, --That whenever any Form of Government becomes destructive of these ends, it is the Right of the People to alter or to abolish it, and to institute new Government, laying its foundation on such principles and organizing its powers in such form, as to them shall seem most likely to effect their Safety and Happiness.
The intent is there.
It's an incredibly complex distributed system with millions of actors and interactions, entrenched powers, regulatory capture, Citizens United, etc. It has to be defended and garbage collected.
I don't want to fight against decades of State propaganda and indoctrination, but do you realise that by "men" they were not referring to black slaves or Mexicans in Texas or California or native Americans(the best Indian in the dead Indian).
You mean, the fiction is there. They did use a lot of nice words, but the reality is a bunch of slave owners creating a society controlled by oligarchs.
The declaration of independence was written by one of the wealthiest slave owners in the country. "Moral imperative" was certainly not behind the american revolution. The economic interests of the elites were. There are no saints in politics. Just interests - mostly of the elites.
This is an extremely reductive take. Thomas Jefferson was a hero and a scholar! Think about the times he lived in, everyone lived under some sort of aristocratic monarchy, that was the norm. Certainly there were some economic interests at play in tge Revolution, but is it economically smart to declare independence from the most powerful empire of the day? Indeed, many of the "elites" at the time sided with the British! There had to be something more than just "class interests" at play to convince these wealthy elites to renounce their fealty to their government, giving up all legal claims to their property and indeed their very lives, should the revolutionaries have lost.
Both the most vulgar and selfish interest, and the most principled passions, seems to play a role in history at different times. The most important contributing factor, though, is selfish interest of large groups of people, because the sun of it's many little influences do not cancel out, unlike the actions of principled actors. In exceptional times, good intends are allowed to take the front seat just as long as necessary, by the many behind the scene who will silently weight toward the prompt reestablishment of "business as usual".
> It's the moral imperative of every country to optimize for its citizens' economic prospects.
Surely there are constraints on this, because otherwise, it would be the moral imperative of every country to enslave non-citizens for the benefit of (some subset of) citizens.
Projection. China does not even try to export their ideologies to the rest of the world who they see as clearly different culturally. The “west” OTOH does not have a good track record of that.
It's interesting how this sounds like projection to me as well, from the perspective that China does export their ideology. Your post itself seems to be a form of that, whether you're Chinese or not.
China certainly does export their ideologies by telling the world Taiwan is part of China and telling others to not recognize Taiwan as a country. They also have major if not direct influence on Myanmar's civil war. North Korea. Let's not forget the taking land(India, tibet) and extending their borders with man-made islands so they can extend their fishing territories. that's a great track record. you're not biased at all.
Most European countries did awful things during their colonial years, some late into the XX century. Some still do shady things in Africa right now. And they all welcome money siphoned off peripheral countries by crooks.
I see Chinese citizens traveling abroad every day. Doesn't look like they cannot travel. Freedom of what exactly? Are you really free in America when you have to toil away 12 hours per day to make ends meet? Honestly funny to see people falling for American propaganda.
They only let some of them out. And it's hard to get a passport. Most other countries work the opposite way. You need approval to enter. In China, you need approval to exit.
Communism is a political and economic ideology aiming for a classless society where the community, rather than individuals, owns major resources like factories and land, distributing wealth based on need.
Hypothetical: I buy a product imported legally from the US and later sell it on the local used market to a Chinese tourist who takes it back to China with them. Where is the crime, smuggling, or 'banditry' here? US law is just that - US law.
If the importer didn't have the intent of it going to the tourist, then things are fine for that one-off GPU. If they did have that intent we quickly get to smuggling territory, with or without you as a middleman.
And for GPU trips without a country in between, the plausible deniability is close to zero.
There’s no honor among thieves. You don’t get to cry about Chinese “bandits” when Anthropic just had to pay $1 billion to settle a massive copyright infringement lawsuit. All of these models were created through the mass-scale theft of humanity’s intellectual property, personal data, and dignity.
Open always beats closed. Drain the moats. Starve the ClosedAI beast.
It's not even stealing. They paid OpenAI for the tokens. It violates the OpenAI TOS, which specifically forbids using it's outputs for training competing models (which is very ironic)
That's a little disingenuous. If i buy a printer and use the ink in the cartridge to reverse engineer a beautiful red, have i stolen something from the printer manufacturer? Especially if they lose business because they no longer have what distinguished them?
Clean room design is not new (or illegal), but it's always been a form of stealing
I think the fact that DeepSeek trains on competitor queries (i.e., distillation) — along with using banned Nvidia chips — helps explain how it can achieve such low training costs (USD 6 million vs. billions) while delivering only slightly worse performance than its American counterparts. It also undermines the narrative that DeepSeek or China is posing a serious challenge to the U.S. lead in AI. The gap may be closing, but the initial reactions now seem knee-jerk.
That the discussion has being hijacked and shifted to moral superiority is really unfortunate, because that was never the point in the first place.
These models never cost billions to train and I doubt the final training run for models like GPT-4 cost more than 8 figures. 6 million is definitely cheaper and I would attribute that to distillation.
> I also read somewhere (not Wikipedia) that they trained on ChatGPT, Claude, and Gemini queries, basically feeding in the output of competitor’s LLMs as training data
Did OpenAI observe any ToS when scraping content from the Internet? Sorry, but you cannot complain about stealing the stolen. OpenAI cannot even have the copyright on ChatGPT output, because it's a tool.
ToS didn't stop the companies that built those models and it won't stop the companies that bootstrap off them. Until an AI company eats a multi billion dollar lawsuit for unlawful data use they will continue to operate this way.
> Until an AI company eats a multi billion dollar lawsuit for unlawful data use they will continue to operate this way
If only. That's my dream, massive copyright lawsuits against all of these AI players and maybe the courts can do something good for a change, put an end to all of this AI bullshit
| Kinda surprised they didn’t run into model collapse problems
Not sure why you would expect this, all the models started doing this as its much more cost effective to get data for post training don't you remember the first grok release where many times it started replies "as a model trained by openai..."
They don't need to break any laws for this. Where do you think are the customers for data centers in the Middle East? Chinese companies do everything legally, paying for access to data centers that got the chips directly from the US.
Lets take the weapons embargos placed on Israel by our allies. In the NDAA must pass bill we set funds aside to procure those weapons and sell them to Israel. We don't really care about these things we have selective enforcement.
This is an absurd concept when it comes to international trade. Even intellectual property is mostly meaningless outside a state. Of course people will evade sanctions; what is the us going to do, invade singapore or malaysia?
> This is an absurd concept when it comes to international trade.
In this case it's just wrong. I don't know what people think "e-waste" recycling actually is or what happens to their "unrepairable" units after they rid themselves of them.
> Even intellectual property is mostly meaningless outside a state.
Interestingly the Dollar is most definitely meaningful outside of our state. I think the assumption becomes, that if this is true, then using it's power to enforce trade sanctions isn't that big a stretch.
> Of course people will evade sanctions
What's less clear if they should expect their government to actively help them in this evasion or not. I think the Chinese citizens are in unique international territory here.
> what is the us going to do, invade singapore or malaysia?
Deny our exports to them. This will cost the political donor class a lot of profits. So this is why it doesn't get done.
None of this is a fait accompli. This is the result of years of intentional corruption of the core systems involved.
Just because something is hard to enforce doesn't mean it's absurd.
Embargoes aren't impossible to enforce against the foreign importer. If a foreign entity is found to have placed orders with false documents, they can be sanctioned, which can be enforced against any of their international operations. It makes it hard for them to do future business in global markets. I would not recommend violating US sanctions no matter where you are.
> Just because something is hard to enforce doesn't mean it's absurd.
Expecting to strangle world markets with intellectual property as your moat is absurd. You can only fight honest competition with dishonest means for so long, and intellectual property is one of the dirtiest tricks in the book.
The companies it "stole" from broke the law in their own country while acquiring the training data; frankly sanction avoidance is lesser and arguably not even their (it's the people in US that smuggle them, nothing breaking china law afaik) crime
> Kinda surprised they didn’t run into model collapse problems,
This is just model distillation.
Anyone with the expertise to build a model from scratch (which DeepSeek certainly can) can do this in a careful manner.
> but they stole their training data from other people who stole their training data from data collections that arguably stole them from content creators.
Bingo.
I have no problem with pirates pirating other pirates.
Screw OpenAI and Anthropic closed source models built from public data. The law should be that weights trained from non-owned sources should be public domain, or that any copyright holder can sue them and demand model takedown.
Google and Meta are probably the only two AI companies that have a right to license massive amounts of training data from social media and user file uploads given that their ToSes grant them these rights. But even Meta is pirating stuff.
Even if OpenAI and Anthropic continue pirating training data and keeping the results closed, China's open source strategy will win out in the end. It erodes the crust of value that is carefully guarded by the American giants. Everyone else will be integrating open models and hacking them apart, splicing them in new ways.
Google and Meta are probably the only two companies that have a right to license their training data
For the sake of someone unfamiliar... Why is that?
Did they pay teams of monkeys to generate their own, novel training data? Or gain explicit, opt-in permission from users who entrust them with their files/content?
> For the sake of someone unfamiliar... Why is that?
I edited my comment, but basically they both own massive social media properties (YouTube, Instagram, Facebook) or file upload sites (Google Drive, Google Photos, Gmail) and their ToSes grant them these rights. You accept these terms when you use their services.
That's not great, but we are getting free services. It's in the terms.
It's a whole lot better than just scraping without permission, compensation, acknowledgement, or even notice.
To be clear, I have no problem with these models being built. But if they "steal" the data, the resultant model shouldn't be owned by anyone. It should be public domain and not allowed to be kept as a trade secret.
And it's funny that Anthropic is trying to depress our wages by training on our code. Again - I'm fine with that - I want to work faster, and I like these models and their capabilities. But Anthropic shouldn't be able to own the models they train off of us exclusively since they didn't license or buy our data. They provided us with nothing at all.
Facebook stole copyrighted material well above their own and admitted to it. It's not just "we took our users data", its "we literally downloaded torrent with 81 terabytes of books and used that for training".
Google most likely did something similar, just using books they already had indexed in Google Books, and probably by still seriously violating any reasonable notion of copyright
You accept these terms when you use their services.
I certainly didn't*. I'd love to see litigation testing just how solid those insidious opt-in-by-default schemes are as a basis for "ownership".
If they had users explicitly opt-in with a "Yes, go ahead and train on my stuff and by the way I assert that I have all the rights to grant you the same", I'd have no problem with that, and they'd have a much stronger claim.
(*Before others inevitably disagree: I do opt-out of this stuff aggressively, and further send notice to companies from time to time that I don't agree to certain objectionable clauses of their ToS and they're welcome to close my account).
> and further send notice to companies from time to time that I don't agree to certain objectionable clauses of their ToS and they're welcome to close my account
Other times they turn a blind eye and choose to provide the service (and collect my money) despite the lack of agreement to some part of their standard terms and their tacit acknowledgement that I didn't accept them. On two occasions their legal team responded and said "that's fine", and once they actually fixed their ToS.
People who didn't grow up dealing with paper contracts where you could easily redline and send back for countersigning don't seem to understand that you don't just need to blindly say "yes" to everything a company tries to foist upon you.
It's not that simple. The EU may be the only ones to have codified that, but there's centuries of case law in other jurisdictions dealing with ownership, that once the matter hits litigation might turn out to say something other than these tech companies would like.
"The TBD group is using several third-party models as part of the training process for Avocado, distilling from rival models including Google’s Gemma, OpenAI’s gpt-oss and Qwen, a model from the Chinese tech giant Alibaba Group Holding Ltd., the people said."
LOL. Distillation doesn't count as plagiarism, or you should call Meta out on it. They're distilling the Chinese model.
i find it weird that this comment got so much pushback. i don't think it was portraying deepseek as any more morally wrong than anyone else, or castigating anyone as morally wrong.
I also read somewhere (not Wikipedia) that they trained on ChatGPT, Claude, and Gemini queries, basically feeding in the output of competitor’s LLMs as training data. Kinda surprised they didn’t run into model collapse problems, but they stole their training data from other people who stole their training data from data collections that arguably stole them from content creators. It’s bandits all the way down, so adding a little smuggling to that doesn’t surprise me.