I like a lot of what Dario writes, but in this case I just can't follow the reasoning. Everything I've picked up about how DeepSeek did what they did (including going a level lower than CUDA to better take advantage of the limited hardware[1], and the balance of techniques used[2]) points to some very smart Chinese engineers having out-smarted US ones (to put it in terms that matter to US folk, because I'm European and I ordinarily wouldn't care):
This post comes across as overly defensive of the US export controls and leaning on the authoritarian regime angle far too much to feel like it isn't just a way to shore up interest in US-based AI companies and widen the moat (or just make sure someone else shores it up politically while they catch up technically).
Anyway, this will always be a deep pocket race. But I wish it wasn't so much about brute-forcing GPUs and wasting power to (as yet) uncertain outcomes as far as model capabilities are concerned, and to me what DeepSeek achieved was to point out ingenuity and better techniques should be something that both OpenAI and Anthropic ought to be pursuing instead of burning cash.
To be sure, DeepSeek did great work, and this is a bit aside from TFA. But the PTX thing is a bit of meme? What do we think torch.compile and triton and llvm's nvptx backend are doing under the hood? The warp-specialization thing quoted in [1] cites to a _2014_ paper[2] out of Stanford ...
Yeah, well, it's not _just_ PTX. Think about what you would do if you had to work in a resource-constrained system (that's a mindset I closely relate to since I still do C++ for MCUs, and it makes you dig _under_ the libraries to save resources).
Totally, they did great work under their constraints. Training in FP8, the MLA thing they introduce in DeepSeek-V2, etc. I just take particular issue with the attention the PTX thing is getting because (a) it's not like other labs don't do stuff like that and (b) it doesn't contribute nearly as much to their outcome as the other algorithmic and operational improvements they've made.
He is basically saying, with his inside knowledge of Anthropic's current capabilities: "they did for $5m what we could probably do for $10m or $15m if we launched the training run today without any new optimizations." So on the one hand that's very impressive, both because the cost effectiveness is significantly higher and because even replicating SoTA outside of OpenAI/Anthropic is very difficult. On the other hand, it's not too surprising that a company that needs to economize on compute will find ways to do so; neither Anthropic nor OpenAI would consider it worthwhile to have their best researchers prioritize cutting down on training costs or compute requirements; they have near infinite capital and are focused on breakthroughs in making their best models as good as possible. I don't think it's accurate to say that Deepseek "outsmarted US engineers"; they had a very different objective function than Deepseek, so they pushed much harder on the engineering optimizations for better cost performance.
Everyone seems to rag on OpenAI/Anthropic for spending so much money and take it as a symptom of capitalist waste, but this reality seems great to me - massive amounts of money is basically being funneled from VCs toward progress in machine learning. Once expensive breakthroughs are made, it is only a matter of a few years until people make the engineering optimizations to make those breakthroughs cheap.
Just want to emphasize the progress in cost that Dario highlights: fixed AI capabilities are becoming 4x cheaper every year. That is absolutely insane. US GDP growth averaged 3-4% over the last 250 years and look how far that has taken us. Moore's Law averaged ~40% annual growth in transistor density and look how far that has taken us in just 65 years. 4x growth is AI capability/cost per year is absolutely insane.
yeh. I see Dario saying "let's protect the US more" for no reason other than bias and "of course they improved over time" which feels like a mighty strong strain of copium. Very disappointing for a leader of an organization i respected. Assuming he speaks for Anthropic, and it seems he does, Past tense.
The "good" news is that neither Anthropic nor OpenAI are based in Europe, so I don't have to feel like they're burning my taxpayer money.
The "bad" news is that so far I've yet to see a serious contender emerging from that part of the world (although the Middle East has invested heavily in LLMs due to obvious cultural differences).
You can tell as an academic he wants to give DeepSeek props, at least that is what I would like to believe. The first third has the presenter leaning into "cheap chinese" too many times.
He writes with the assumption that the reader understands the danger in China developing a lead in military technology. Hence his very explicit wording:
"To be clear, the goal here is not to deny China or any other authoritarian country the immense benefits in science, medicine, quality of life, etc. that come from very powerful AI systems. Everyone should be able to benefit from AI. The goal is to prevent them from gaining military dominance."
Without taking a position on unipolar vs. multi-polar:
Dario makes an astounding implicit assumption:
- China originating labs cannot acquire chips providing 80-90% similar utility without the US within the next 2-3 years.
I'll make an observation, re: DeepSeek's incentives that drove them to create the innovations from the V2 and V3 papers.
DeepSeek, compared to American AI labs, are much more compute constrained, but in a unique way. Their chips are more memory bandwidth constrained (depending on type anywhere from 50% to 80% less bandwidth).
Therefore, each dollar/hour of investment towards memory optimization is worth MORE to DeepSeek than to American labs.
In the V2/3 paper, they've demonstrated exactly that with these memory optimization techniques.
1. MLA -> reduces KV cache by nearly 80% compared to GQA. By the way, this was published in V2 in May 2024.
2. FP8 matmul (while still accumlating in FP32 gradients) without losing significant quality.
3. DualPipe scheduling and reworking of Hopper SM's allocation on communication vs. computation -> DeepSeek's V3 paper has 2 full pages of hardware suggestions for "hardware designers" (read NVIDIA)
Export controls in a global market create different incentives in parties. The resulting incentives will change, and agents (using it as an traditional economics term) will change their capital allocation strategy.
> All of this is to say that DeepSeek-V3 is not a unique breakthrough or something that fundamentally changes the economics of LLM’s; it’s an expected point on an ongoing cost reduction curve. What’s different this time is that the company that was first to demonstrate the expected cost reductions was Chinese.
Says the CEO whose product [1] costs 15-50x times more. (This is not just the DeepSeek's API, but also 3p providers hosting the same model)
> DeepSeek does not "do for $6M5 what cost US AI companies billions". I can only speak for Anthropic, but Claude 3.5 Sonnet is a mid-sized model that cost a few $10M's to train
Ok, that's still at least 3-10x cost reduction (assuming "a few $10M" lowerbound of $20M). And for a model that he later implies is 2x larger than Sonnet. So that's 6-10x efficiency improvement. Nice!
> Since DeepSeek-V3 is worse than those US frontier models — let’s say by ~2x on the scaling curve.
What curve? Does he mean the simplistic performance / model params curve? That does not take into account that DeepSeek v3 is a MoE (can't compare MoE and dense param # in a naive way), nor the other architecture changes (KV compression, etc.).
Also, if Sonnet 3.5 is 2x smaller, then why is inference 15-50x more expensive than DeepSeek v3's? Does Anthropic not have good GPU engineers? Are they just running at insanely high margins? As a consumer I don't care how big your model is behind the scenes. I care about API costs or inference efficiency when hosting the model myself.
[1] Product that is mostly comparable and in some ways quite ahead.
> DeepSeek does not "do for $6M what cost US AI companies billions". I can only speak for Anthropic, but Claude 3.5 Sonnet is a mid-sized model that cost a few $10M's to train (I won't give an exact number). Also, 3.5 Sonnet was not trained in any way that involved a larger or more expensive model (contrary to some rumors).
^ This is publicly new information, and the 2nd part especially contradicts consequential rumours that had been all-but-cemented in closely-following outsiders' understanding of Sonnet and Anthropic. Completely aside from anything else in this article.
Also, though it's not "new information": "Making AI that is smarter than almost all humans at almost all things [...] is most likely to happen in 2026-2027." continues to sail over everybody's head, not a single comment about it, even to shit on it. People will continue to act as though they are literally blind to this, as though they literally don't see it.
> People will continue to act as though they are literally blind to this, as though they literally don't see it.
Or like they see it and have learned the appropriate weight to give unsupported predictions of this type from people with a vested interest in them being perceived as true. It not only not new information, its not information at all.
Anthropic is, according to themselves, using RLAIF... which is basically using LLM as a judge / reward model. So maybe he means that the models they use for RLAIF are not (much?) more expensive than Sonnet 3.5 (e.g. previous Sonnet or Haiku 3 :)).
What allies? Is he aware that his current President is alienating and antagonising almost every single country out there that can historically be considered an ally?
Multipolar world is probably going to have multiple wars on the "edges of control", similar to the one now raging in Ukraine.
It may also have some positives, too, but situations with several similarly sized powers competing for primacy have historically been pretty bloody. Unlike Apple and Google, the US, China, Russia etc. have military forces at their disposal.
US unipolar hegemony has averted an all-out war, but it has also been very bloody, or at least immiserating, for people at its periphery.
Multipolarity doesn't imply we go back to the 1910s. The idea would be to strengthen multilateral institutions that put a check on things like the World Wars.
Humans are pretty much the same as they were in the 1910s, or the entire history before.
If there is anything that keeps us from tearing each other's throats out, it is a) democracy, which in most countries makes going into a war of aggression somewhat harder (I know, the US is a huge exception, but Trump I. was partially elected on the basis of Clinton being perceived as a war hawk), and b) the intuition that weapons are now so destructive that there is nothing to win, except scorched earth.
Still that didn't stop Putin from launching his war; miscalculation such as his ("the enemy is a paper tiger and will fold immediately") is very possible even today.
IIRC the same delusion led Saddam to invade Iran in 1980.
well, from the historical perspective, it's more easy(or equally hard) to achieve a peaceful multipolar world than a eternal unipolar world
i mean, all 'rule-based international order' died in history, the Holy Roman Empire, the First Empire of France, and the ancient Chinese empire died in every 300 years
even there is no China, even the empire doesn't implode, will africans and south americans be willing to mining for the west for all their lives? there will be Sankaras and Castros in every several decades, every hegemony has an end
Can you name a single period of peaceful multipolar world in situations where the Great Powers could actually reach one another militarily?
I mean, the fact that the Romans and the Chinese never clashed over future Kazakhstan is not a proof of a peaceful multipolar world. It is simply a corollary of the fact that neither empire was able to expand into the steppe.
I just want to ask if when we talk about export controls on AI chips, is this going to create export controls on general consumer goods in the near future.
Given Moore’s law and the efficiency that will clearly come from optimizing chips for AI and competition increasing the amount of VRAM on these devices to run models locally.
Is creating an export regime today going to mean that in 3 or 4 years general smart phones and high end laptops are all going to be subject to export controls?
Keep in mind that at one time computers that consumed entire building are less powerful than my apple watch.
That would be a great way to give Non-American companies an advantage in the market. Imagine if Apple has to constantly worry about complying with US export controls while Samsung, a Korean company, can just ship whatever they want with zero additional paperwork as long as they avoid US suppliers. Multiply that over all consumer goods that might include AI.
Dario certainly had no qualms about working with authoritarian regimes when he worked for Baidu!
Also, if the US seeks a permanent advantage over China by 2027 with self-replicating AI, so much so that it can dictate whatever terms militarily. Wouldn't that just force China to send a few cruise missiles to TSMC before that happen?
Unless the author makes a compelling case about why AI breaks the MAD status-quo between nuclear powers, I will assume that their appeals to NATSEC are an attempt to artificially create moat for their company.
I also find it counterintuitively reassuring that we have already had the power to blow up the planet for decades. But at least in theory - GPT10 level AI could help develop missile defense systems that actually work at scale, which would eliminate the MAD status quo, no?
What about the missile guidance / launch systems it could develop? The question at the at the limit of development becomes is nuclear warfare slightly attacker biased or slightly defender biased?
At least they're all coming out of the woodwork and start telling more details about their own training runs, costs, efficiency rates and so on. Interesting to see how an open-weights model could force their hand like that.
The technical summary is excellent. But I worry that prominent US voices couching AI progress in Cold War style war rhetoric is likely to be... self-fulfilling. Surely some of the people paying closest attention to these articles are Chinese politicians and military leaders - it would definitely seem alarming to me if the Deepseek CEO kept writing things like: "whoever reaches this threshold of AI intelligence first will accelerate into world military domination - we must beat the US."
There's a reason why Oppenheimer & co. weren't non-stop publishing internationally accessible op-eds during WWII about how important it is that Germany not develop a scaled uranium fission bomb before the US.
If you read enough history books (and have a passable grasp of German, and tolerate the weird lead types used in the press for those days) you will be able to find _a lot_ of opinion pieces from our recent history that play along the same lines, but for earlier industries (like automotive).
From my perspective it’s very convenient that all the new information and competition supports his existing priors that:
1) we need to do various forms of regulation to entrench US closed source market leaders, which happens to increase his company’s value
2) the best way towards improvement in models is not efficiency but continuing to burn ever increasing piles of money, which happens to increase his company’s value
It's shortsighted to believe that export controls are relevant. China will be able to manufacture as good or better chips than the US in the not too distant future anyway.
This is a choice between the lesser of two evils, or what could be called "drinking poison to quench thirst." Export controls might lead China to develop its own advanced chips, but without them, China will certainly gain powerful AI capabilities
Point being, they will gain those capabilities either way. The only question is whether we in the US choose to hamstring ourselves with bureaucracy while we watch them do it.
The only way this would work is to create a kind of great firewall of America keeping open source Chinese models out while the remainder of the globe uses something like DeepSeek.
It would literally be the dumbest soft-power move in the history of the US.
While Russia has certainly found ways around the sanctions, look at their economy... Putin is spending his reserves out from under the next decade. They're rather fucked.
Clearly china could last longer under similar conditions, but they're also looking weaker than they did 5 years ago.
> And it doesn’t matter if Trump defends Taiwan and makes China back down - the fabs will have been bombed to bits.
What would prevent the US from directly attacking China's domestic fabs in this scenario? Defending Taiwan doesn't seem as proxy-ish as Ukraine considering China considers it their territory already.
That's a extremely dangerous, disproportionate and self destructive strategy. What would prevent China from bombing US targets as revenge?
Both mainland and Taiwan belong to China, it's written even in the Taiwanese constitution. The controversy is in the definition of China, basically they are supposed to unify mainland while the mainland wants to unify them. A balance will be reached at some point either peacefully or violently.
Peace is more likely if the US stops pouring oil on fire.
The US should stay out of this situation if its leadership has its own people's interest in mind. A hot war between US and China will cause many millions to die and possibly end of human civilization. Let that sink in.
It seems unlikely the US will take military action to protect an ally.
It seems even more unlikely that, should the US be prepared to bomb land targets in China, that they would do anything beyond overt military targets that are being used to attack Taiwan.
When trump has recently been asked what he’ll do when China attacks Taiwan he has just said sanctions and tariffs.
My going hypothesis is that the US cannot politically solve their Modern Monetary Theory national debt blackhole problem and is therefore itching for WW3 to be able to "reset" debts when the dust settles. Taiwan seems like a good opportunity for that, but maybe there is some sanity left in Congress.
> All of this is to say that DeepSeek-V3 is not a unique breakthrough or something that fundamentally changes the economics of LLM’s; it’s an expected point on an ongoing cost reduction curve. What’s different this time is that the company that was first to demonstrate the expected cost reductions was Chinese.
Excellent analysis and I largely agree. The part where he looses me here is the certainty in the claim around fundamental changes to the economics...
This is open weights with a (largely) open recipe. Arguably, Meta has already been at it but now there are two labs producing frontier models at this level. A wishful analysis could imagine that there _might_ be more and that _could_ result in a qualitative shift in how the frontier model research is being done and hence affect the economics but Dario seems to have largely ignored this aspect.
> Export controls serve a vital purpose: keeping democratic nations at the forefront of AI development.
As an outsider looking in, I’m not thinking of the USA as a particularly democratic country any more. And I didn’t think its export controls were about democracy, but rather a trade weapon.
"Making AI that is smarter than almost all humans at almost all things will require millions of chips, tens of billions of dollars (at least), and is most likely to happen in 2026-2027."
What an incredible claim to just slip in two-thirds of the way through! Where have I heard it before?
1965, H. A. Simon: "Machines will be capable, within twenty years, of doing any work a man can do."
Stepping back, it feels like the techies working on ai have been celebrating the expense of it all. That techies have been wanting bigger and bigger clusters as some kind of bragging rights and that suits the suits who see it as a most to protect their crazy stock validations. The money was being made not by the miners but by nvidia selling shovels they could name the price of.
Necessity is the mother of invention. Deepseek is demonstrating that if the incentives are aligned to want efficiency then good engineers - and there are plenty of them in every country - can make things fly.
The author seems to be of the opinion that the creators of DeepSeek will either be unable to, or will not see the value of optimizing the 'second stage' RL component of the 'new' (post pre-training RL) way of training frontier foundation models. Every competent programmer in China is now looking for low level ptx optimizations for EVERY SINGLE STAGE of the pipeline. They will now, likely not publish any of it.
>Even if the US and China were at parity in AI systems, it seems likely that China could direct more talent, capital, and focus to military applications of the technology. Combined with its large industrial base and military-strategic advantages, this could help China take a commanding lead on the global stage, not just for AI but for everything.
China spends 1.5% of its GDP on its military. The US spends 3.5%. I get that the two countries are engaged in competition for dominance, but why is China the bigger threat here?
Also, there are a great deal of groups and institutions out there that are pushing for more diplomacy, more cooperation, and a ratcheting down of tensions. If Dario is going to get political anyway, why not go that route?
It's also funny that there's a scary implication "China could do XYZ and focus on military applications" when Anthropic and Palantir teamed up last year to offer Claude models to US Defense orgs
That post is really dishonest, which is expected from Anthropic's CEO talking about the competitors. Dario conveniently omits that R1's approach was validated months before o1-preview came out, and DeepSeek releases the weights and architecture unlike Anthropic and OAI which produce black box models in the name of "safety". Then he tries to diminish their innovation, dismissing it as a normal step they're also capable of, saying v3 and R1 are competing with models that are 9-10 months old. Which is not really true.
And then there goes the usual techno-feudal interjection:
>If China can't get millions of chips, we'll (at least temporarily) live in a unipolar world, where only the US and its allies have these models. It's unclear whether the unipolar world will last, but there's at least the possibility that, but there's at least the possibility that, because AI systems can eventually help make even smarter AI systems, a temporary lead could be parlayed into a durable advantage
So Dario envisions a unipolar world where a black box corporation (his own one, of course) controls a perpetually self-improving machine god and distributes scraps to the masses that are too unwashed and dangerous to control it themselves, while at the same time preventing other countries from breaking out and making their own self-improving machine god because they're evil. A boot stomping on your face forever, regardless of who you are.
That's coming from an "ethical, helpful, and honest" guy.
>To be clear, the goal here is not to deny China or any other authoritarian country the immense benefits in science, medicine, quality of life, etc. that come from very powerful AI systems. Everyone should be able to benefit from AI. The goal is to prevent them from gaining military dominance.
And of course he immediately tries to deny that thought.
This piece ignores recent news that DeepSeek is doing inference on Huawei’s Ascend 910C GPUs.
The simplified export control analysis Amodei gives here gets a lot more complicated when increasing export controls potentially ends up spurring additional R&D in Nvidia competitors.
There are lots of people working on ASICs for inference. There are already people using FPGAs for some vertically integrated AI stacks, so it's only a matter of time until we start discussing literally hard-coding models (or flashing them) to hardware to make them go faster.
AI is a race to the bottom, and a race to over-spend. Seeing this as competition will bankrupt the players, make enablers rich, and end up killing a lot of people when the need to bolster confidence in its capabilities overrules basic common sense.
Maybe if US wasn't led by a bully that care's not about democratic ideals, but only on the rule of power it'll be easier to support the notion of US the leader of the free world.
But just take a look at how he handled Denmark-Greenland issue, Pure bullying.
In the Trumpian world, US has no allies, only vassals.
I'm kind of tired of the entire narrative that "US good, everyone else bad", as if only US deserves to hold powerful technology because it will be used for "the greater good", rather than employ it in military applications, as if Anthropic didn't partner with Palantir.
And it amazes me how many people can't seem to see past the "US is good, everyone else are bad" smoke mirror.
I'm sure most of us agree with you, but independently of that, please don't take HN threads on generic tangents, and certainly not generic nationalistic flamewar tangents. The problem is that it's tediously repetitive and inevitably turns nasty.
First- the author of this blog post is the CEO of Anthropic, an American AI company. Of course they are going to argue for export controls if it can hurt their competition. They get to benefit from anticompetitive practices with none of the legal risk! So it's not even about keeping the "bad guys" from having things, it's really about making more profit. Also, you can typically assume that anything a CEO publishes is in a pursuit to raise stock prices, it has nothing to do with morality. If CEOs were moral people most of them wouldn't have become CEOs (there are probably a few CEOs that are truly good people, but I'll bet there are more "wolf in sheep's clothing" CEOs that use an image of morality to improve their company's reputation).
Second- aside from the specifics of this post, you don't have to believe that the US is good to see the logic of export controls. There are a number of countries that are openly hostile to the US (Iran and North Korea), and also some that are semi-privately working against the US (like China). These countries will take any advantage that they can- they will steal IP and use the US's development resources to catch up to the current state of the art at a reduced cost. So you don't have to put this in terms of good vs. bad, just think of it in terms of things that benefit the US and things that don't benefit the US. Whether you think people are evil is irrelevant, it is very logical for the US to put export controls on things that it doesn't want to give away for free. The US wants to preserve its advantage and make it as costly and difficult for these other countries to catch up.
Absolutely. Chinese companies shouldn't have chips because their government has "committed human rights violations, has behaved aggressively on the world stage". And the US government hasn't?
To say the two are remotely comparable is either insincere or naive. Like I'm not saying the US always acts utterly virtuously, but they aren't running concentration camps and chemically castrating entire ethic groups.
They also aren't imprisoning people for speech, which is important when we're discussing who we want controlling AI in the future.
Don't forget Central America and Iran; supporting military dictators in Southeast Asia (Marcos in the Philippines and Suharto in Indonesia); invading Haiti and raiding it of the entirety of its gold reserves, condemning its people to poverty -- one could go on and on.
DeepSeek model not providing answer on the Tiananmen Square and ChatGPT providing answer on ethnic cleansing of Palestine are two sides of the same coin.
I made no such comparison, but if you really want to start comparing government crimes and human rights abuses, I think you'll find we can make a stronger moral case arguing for chip bans for US companies than Chinese companies. https://www.youtube.com/watch?v=8BOBKTmaQ9M
if you're chinese you'll know some chinese were sterilized because they _didn't_ believe in a religion
muslims were pardoned of child policy because of their ethnics, it's a privilege of them
people can not imagine something they never saw, try dig your own history, you'll find why it's always about religonal genocide, religonal sterilization, and religonal concentration camps
I find this line of reasoning a whole lot less convincing now than I would have 2 weeks ago, and it wasn't all that convincing two weeks ago.
When the controlling party of the US is discussing concentration camps for immigrants while also happily calling to revoke birthright citizenship and deport citizens of the US that criticize Trump or the Republican party (eh - who am I kidding - it's the maga/heritage party now)...
It feels like we have no legs left to stand on here, and the support was DAMN shaky to begin with, seeing as we were routinely knocking over governments that we see as inconvenient in our geopolitical sphere of control for the last 100+ years.
Essentially, I think your argument is about 3 elections stale.
> To say the two are remotely comparable is either insincere or naive. Like I'm not saying the US always acts utterly virtuously, but they aren't running concentration camps and chemically castrating entire ethic groups.
They absolutely did directly and indirectly several times in their history. A few of those times were recently:
> In September 2020, it was revealed that ICE had performed mass hysterectomies on immigrant women in several detention centers, reminiscent of the long-standing US policy of sterilization of black and brown women. 2
> The US currently operates a system of slave labor camps, including at least 54 prison farms involved in agricultural slave labor. Outside of agricultural slavery, Federal Prison Industries operates a multi-billion dollar industry with ~ 52 prison factories , where prisoners produce furniture, clothing, circuit boards, products for the military, computer aided design services, call center support for private companies. 1, 2, 3
> In the present day, ICE (U.S._Immigration_and_Customs_Enforcement), the police tasked with immigration enforcement, operates over 200 prison camps, housing over 31,000 undocumented people deemed "aliens", 20,000 of which have no criminal convictions, in the US system of immigration detention. The camps include forced labor (often with contracts from private companies), poor conditions, lack of rights (since the undocumented aren't considered citizens), and forced deportations, often splitting up families. Detainees are often held for a year without trial, with antiquated court procedures pushing back court dates for months, encouraging many to accept immediate deportation in the hopes of being able to return faster than the court can reach a decision, but forfeiting legal status, in a cruel system of coercion. 1, 2
> During the 2020 coravirus pandemic, it was found that a law that empowered police to arrest those for not social distancing, lead to 80% of those arrested being black and latino.
True, providing complete military support, with special deliveries of bombs and ordinance to Israel, knowing full well that it will go towards obliterating a city and killing tens of thousands of civilians is more humane. It's not genocide if an ally is doing it!
Oh, people are thinking it now for sure, but it takes a while to retool your relationships and build out a new trade network so you can safely tell a major trading partner to fuck themselves to threats of 100% tariffs.
If Europe and Latin America want to buddy up with the concentration camp gang, I really don't care. We do not need fascists and communists as our partners.
Funny you mention fascists. As for China being a concentration camp gang, that's not going to matter when their former big brother/ally is just relentlessly bullying them and trying to extract all sorts of concessions under duress, and the alternative has demonstrated to be a reliable ally and partner.
Once again, I don't really care if Europe decides that is who they want to be friends with. If your country thinks that trade is more important than the country currently directly committing a modern holocaust, I don't want us to be allied with you at all.
Yeah. I personally think we should just sink the middle east into the ocean because not a single problem there will ever be resolved in a meaningful way.
Ironically, if Trump keeps running this playbook, it's going to end with a USA/Russia/Israel bloc and a Europe/LatAm/Africa/China bloc. Which one of those sounds like the evil empire to you?
Note how interesting it is that you can read about these on the internet in the US on US hosted servers by US companies. At least we own up to it and don't just throw you in a mental asylum like they do to anyone who dares to stand up to the CCP.
Article 1 Section 8 does not enumerate Congress with the power to provide healthcare, so Medicaid should onoy be a state level program.
I know removing it harms those at the low end of the income spectrum, and that is a bummer, but I really would prefer Congress ammend the Constitution instead of just passing laws of which they have no authority to pass.
That is... bonkers. At the risk of feeding a troll, I find it difficult to take the argument that voluntary contributions to public health through charity is more "moral", which in the case of health care I would argue is equivalent to "effective", than a system implemented by an organization with enormous power (the government) which at least theoretically has a direct duty to its stake holders (voters), who in turn have the power to enact change in that system if it's not serving them? (by voting)
They just kind of have to give up their power and hope?
Buddy, one cursory glance at history will show that hoping gets you nowhere.
I assure you I'm no troll. I think difference in our perspectives is that I form the basis of my "moral" at the individual vs. the collective.
"Moral" for me means that individuals are empowered to own their private property and should only need to sacrifice it to society for public goods, where I take the economic definition of a private good: nonrivalous and nonexcludable.
Forcing all persons to pay taxes to cover healthcare for only a subset of the population is, to me, akin to forcing all your friends to give to the charity you like because you like that charity and want it to be able to do more, where that more is a level of spending above what you can or are willing to provide.
Economically, this creates deadweight loss: people's individual preferences are violated because they are forced to spend money for something of which they receive no benefit, or at least the direct/indirect utility occupies a lower utility than the opportunity cost of those specific taxes.
I'm not saying that such a policy won't result in undue death. But since I use the individual as the basis for morality, I consider it more moral to have some death than it is to steal from others to prevent it.
Well according to the unitary executive theory the US president has complete authority over all government actions and congress is only able to give legitimacy to the president's actions. Under this regime the president would be allowed to stop the execution of any law passed by congress.
I think this is insane and a complete destruction of the balance of power written in the constitution that congress can't enact enforceable laws but only "suggestions" for the president
> UET is a constitutional law theory that gives the President sole authority over the executive branch.
I think UET applies only to the Executive branch, which to me makes sense as he is the head of the Executive branch.
It would seem a violation of checks and balances for Congress to be able to install unfirable persons in the Executive. The checks and balances come from Congress's subpoena and investigatory powers, which can ultimately result in impeachment and removal of office of the President if he is derelict in his duties of executing the laws set forth by Congress.
Though I would agree that UET would violate the balance of the Constitution if it applied outside the Executive branch.
Do you really think the US is going to come out on top as the "lesser evil" if we start listing all the "bad things" each side has done related to human rights abuses and behaving aggressively on the world stage? https://www.youtube.com/watch?v=8BOBKTmaQ9M
I'm not sure how that relates to original comment. Do you mean you want everything that is or could be better than American technology banned/destroyed so we stay the best...?
Like, any global hegemony will be increasingly corrupt given the power that gives, IMO.
He's unlikely to be writing from a genuinely ideological point of view. He's playing a zero-sum game, in that every chip that goes to China is one that Anthropic and/or its cloud provider doesn't get its hands on.
So it makes sense to him to argue for export controls using whatever rhetorical flexes and flourishes he can come up with.
As long as the semiconductor fabs are running at capacity, yes, it's one big zero sum game. If you win one chip, I lose one chip, and vice versa.
This situation is temporary, of course. China previously had a large incentive to get their own leading-edge nodes into production, and now they have a Manhattan Project-size incentive.
They aren't running at full capacity on AI chips though - TSMC's main customer is iPhones as far as I know. So you could take away production time for them, though it's still zero-summy.
But TSMC is building new leading edge fabs right now.
It might already be more dangerous in the sense that it is unpredictable. China is autocratic and powerful but their goals are known and do not change at a whim.
You can't build diplomatic and trade ties in quicksand.
Economically, perhaps. Brexit certainly didn't help Europe as a whole, and German, French and Italian politics have been off-kilter for years, but it's still, by and large, an ethical and civilized place to live in.
I will freely grant that Euro money goes to completely obsolete heavy industries and luxury items rather than technology innovation (and blame all politicians for missing out on technology), but there are outliers like ASML that almost make up for the whole thing...
U.S. and Chinese hegemony are just different kinds of bad. We could put the pros and cons up and debate it endlessly, but in the end ranking them is pointless, humanity should just be trying to make a better alternative
I don't know. Living in a country where being able to shoot people is an absolute, immutable, inalienable right granted by God himself (more so than anything else in the Bill of Rights) but where access to food, shelter, healthcare and education are all expensive privileges subject to the whims of the free market, and where half of the population doesn't even believe viruses are real, and where you don't even need the popular vote to elect a President (it's just nice to have) because the electoral system is a series of 18th century compromises to keep slavers happy... doesn't seem like the best of all possible worlds to me.
> Living in a country where being able to shoot people is an absolute, immutable, inalienable right granted by God himself (more so than anything else in the Bill of Rights)...
2A provides access to weapons. It doesn't allow people to murder one another.
> but where access to food, shelter, healthcare and education are all expensive privileges subject to the whims of the free market,...
Article 1 Section 8 doesn't enumerate Congress the power to provide any of those things.
What's more, I beieve the Federal government shouldn't. Rights should only ever be negative unless being a counterparty to positive action taken by the government (e.g., it makes sense to provide public defenders since it is reactionary to the government trying to take away your rights).
Healthcare, education, etc. are better served by the private sector with charity as the supplement for the less fortunate.
> and where you don't even need the popular vote to elect a President (it's just nice to have) because the electoral system is a series of 18th century compromises to keep slavers happy...
The electorial system is designed to prevent tyranny of the majority and is biased towards inaction. This seems better than a pure democracy to me because most government action is bad: inefficient, introduces moral hazard, etc.
In a system of pure democracy, the bottom 51% can vote to tax the top 49%'s income away. Then the next iteration the new bottom 51% can vote the same. Eventually this could lead to purely even distributions. That seems wildly worse for society: all incentive to produce would be void.
Most people - even most Americans - aren't libertarian minarchists who believe everything the government does is harmful by default and that free market capitalism is an unalloyed moral good. History has shown that the kind of governance you want rarely works out well for anyone but oligarchs.
I don't believe everything the government does is harmful. Personally I'm quite fond of the FAA despite the many that complain about TSA.
But most of what the government does introduces dead weight to economic transactions or moral hazard. This is especially true when the government tries to provide social benefits beyond taking over a product or service that could perfectly well be provided by the private sector.
I would have little argument against the government regulating that private healthcare providers cannot discriminate on the basis of preexisting conditions. It would increase costs uniformly across the industry. I do have arguments for providing healthcare on the basis of income, as I believe the program would be better served by private charity. Those who argue the contrary do so only because they want to force all citizens to contribute to the private charity (run publically) of healthcare-for-the-poor, because they want to solve someone else's problem but don't want to have to bear the burden themselves.
It does if you're quickly converging into the Handmaid's Tale universe. May the odds be ever in their favor - I'm quite happy we don't carry guns to make it through the day...
You do not have a right to other people's work. I should not be forced to give you food because you are too lazy to do it yourself. If you do not work, you do not eat, just like Jamestown.
You do not deserve to eat just by existing (unless you are disabled or medically cannot work). Saying this is true implies that you have a right to other people's labor, which you do not.
Land, water, air and light used to produce food are not owned by any particular human either. The monopolization of that land (at the point of a gun most often) is the beginning of the issue. The overuse of that land, often subsidized heavily by taxpayers, in order to generate exorbitant profit for the few is a rot in human society. Those owners are the actual ones who don't work and expect a reward for it
I like a lot of what Dario writes, but in this case I just can't follow the reasoning. Everything I've picked up about how DeepSeek did what they did (including going a level lower than CUDA to better take advantage of the limited hardware[1], and the balance of techniques used[2]) points to some very smart Chinese engineers having out-smarted US ones (to put it in terms that matter to US folk, because I'm European and I ordinarily wouldn't care):
This post comes across as overly defensive of the US export controls and leaning on the authoritarian regime angle far too much to feel like it isn't just a way to shore up interest in US-based AI companies and widen the moat (or just make sure someone else shores it up politically while they catch up technically).Anyway, this will always be a deep pocket race. But I wish it wasn't so much about brute-forcing GPUs and wasting power to (as yet) uncertain outcomes as far as model capabilities are concerned, and to me what DeepSeek achieved was to point out ingenuity and better techniques should be something that both OpenAI and Anthropic ought to be pursuing instead of burning cash.
To be sure, DeepSeek did great work, and this is a bit aside from TFA. But the PTX thing is a bit of meme? What do we think torch.compile and triton and llvm's nvptx backend are doing under the hood? The warp-specialization thing quoted in [1] cites to a _2014_ paper[2] out of Stanford ...
[2]: https://dl.acm.org/doi/10.1145/2555243.2555258
Yeah, well, it's not _just_ PTX. Think about what you would do if you had to work in a resource-constrained system (that's a mindset I closely relate to since I still do C++ for MCUs, and it makes you dig _under_ the libraries to save resources).
Totally, they did great work under their constraints. Training in FP8, the MLA thing they introduce in DeepSeek-V2, etc. I just take particular issue with the attention the PTX thing is getting because (a) it's not like other labs don't do stuff like that and (b) it doesn't contribute nearly as much to their outcome as the other algorithmic and operational improvements they've made.
He is basically saying, with his inside knowledge of Anthropic's current capabilities: "they did for $5m what we could probably do for $10m or $15m if we launched the training run today without any new optimizations." So on the one hand that's very impressive, both because the cost effectiveness is significantly higher and because even replicating SoTA outside of OpenAI/Anthropic is very difficult. On the other hand, it's not too surprising that a company that needs to economize on compute will find ways to do so; neither Anthropic nor OpenAI would consider it worthwhile to have their best researchers prioritize cutting down on training costs or compute requirements; they have near infinite capital and are focused on breakthroughs in making their best models as good as possible. I don't think it's accurate to say that Deepseek "outsmarted US engineers"; they had a very different objective function than Deepseek, so they pushed much harder on the engineering optimizations for better cost performance.
Everyone seems to rag on OpenAI/Anthropic for spending so much money and take it as a symptom of capitalist waste, but this reality seems great to me - massive amounts of money is basically being funneled from VCs toward progress in machine learning. Once expensive breakthroughs are made, it is only a matter of a few years until people make the engineering optimizations to make those breakthroughs cheap.
Just want to emphasize the progress in cost that Dario highlights: fixed AI capabilities are becoming 4x cheaper every year. That is absolutely insane. US GDP growth averaged 3-4% over the last 250 years and look how far that has taken us. Moore's Law averaged ~40% annual growth in transistor density and look how far that has taken us in just 65 years. 4x growth is AI capability/cost per year is absolutely insane.
yeh. I see Dario saying "let's protect the US more" for no reason other than bias and "of course they improved over time" which feels like a mighty strong strain of copium. Very disappointing for a leader of an organization i respected. Assuming he speaks for Anthropic, and it seems he does, Past tense.
The "good" news is that neither Anthropic nor OpenAI are based in Europe, so I don't have to feel like they're burning my taxpayer money.
The "bad" news is that so far I've yet to see a serious contender emerging from that part of the world (although the Middle East has invested heavily in LLMs due to obvious cultural differences).
You can hear it directly from Dario
Navigating a world in transition: Dario Amodei in conversation with Zanny Minton Beddoes https://www.youtube.com/watch?v=uvMolVW_2v0
You can tell as an academic he wants to give DeepSeek props, at least that is what I would like to believe. The first third has the presenter leaning into "cheap chinese" too many times.
He writes with the assumption that the reader understands the danger in China developing a lead in military technology. Hence his very explicit wording:
"To be clear, the goal here is not to deny China or any other authoritarian country the immense benefits in science, medicine, quality of life, etc. that come from very powerful AI systems. Everyone should be able to benefit from AI. The goal is to prevent them from gaining military dominance."
Without taking a position on unipolar vs. multi-polar:
Dario makes an astounding implicit assumption:
- China originating labs cannot acquire chips providing 80-90% similar utility without the US within the next 2-3 years.
I'll make an observation, re: DeepSeek's incentives that drove them to create the innovations from the V2 and V3 papers.
DeepSeek, compared to American AI labs, are much more compute constrained, but in a unique way. Their chips are more memory bandwidth constrained (depending on type anywhere from 50% to 80% less bandwidth).
Therefore, each dollar/hour of investment towards memory optimization is worth MORE to DeepSeek than to American labs.
In the V2/3 paper, they've demonstrated exactly that with these memory optimization techniques.
1. MLA -> reduces KV cache by nearly 80% compared to GQA. By the way, this was published in V2 in May 2024.
2. FP8 matmul (while still accumlating in FP32 gradients) without losing significant quality.
3. DualPipe scheduling and reworking of Hopper SM's allocation on communication vs. computation -> DeepSeek's V3 paper has 2 full pages of hardware suggestions for "hardware designers" (read NVIDIA)
Export controls in a global market create different incentives in parties. The resulting incentives will change, and agents (using it as an traditional economics term) will change their capital allocation strategy.
> All of this is to say that DeepSeek-V3 is not a unique breakthrough or something that fundamentally changes the economics of LLM’s; it’s an expected point on an ongoing cost reduction curve. What’s different this time is that the company that was first to demonstrate the expected cost reductions was Chinese.
Says the CEO whose product [1] costs 15-50x times more. (This is not just the DeepSeek's API, but also 3p providers hosting the same model)
> DeepSeek does not "do for $6M5 what cost US AI companies billions". I can only speak for Anthropic, but Claude 3.5 Sonnet is a mid-sized model that cost a few $10M's to train
Ok, that's still at least 3-10x cost reduction (assuming "a few $10M" lowerbound of $20M). And for a model that he later implies is 2x larger than Sonnet. So that's 6-10x efficiency improvement. Nice!
> Since DeepSeek-V3 is worse than those US frontier models — let’s say by ~2x on the scaling curve.
What curve? Does he mean the simplistic performance / model params curve? That does not take into account that DeepSeek v3 is a MoE (can't compare MoE and dense param # in a naive way), nor the other architecture changes (KV compression, etc.).
Also, if Sonnet 3.5 is 2x smaller, then why is inference 15-50x more expensive than DeepSeek v3's? Does Anthropic not have good GPU engineers? Are they just running at insanely high margins? As a consumer I don't care how big your model is behind the scenes. I care about API costs or inference efficiency when hosting the model myself.
[1] Product that is mostly comparable and in some ways quite ahead.
Where does he imply that it's 2x larger than Sonnet?
I inferred that from the other statement:
> Since DeepSeek-V3 is worse than those US frontier models — let’s say by ~2x on the scaling curve, which I think is quite generous to DeepSeek-V3
He says that it's 2x worse. So if it has the ~same quality [1], it would imply it's 2x larger. Unless I misunderstood what he meant there, of course.
[1]: "DeepSeek produced a model close to the performance of US models" -- in his own words.
Ah, I read that as him saying the quality isn't as good-equivalent to a model with 50% less compute. But you might be right.
> DeepSeek does not "do for $6M what cost US AI companies billions". I can only speak for Anthropic, but Claude 3.5 Sonnet is a mid-sized model that cost a few $10M's to train (I won't give an exact number). Also, 3.5 Sonnet was not trained in any way that involved a larger or more expensive model (contrary to some rumors).
Wow!
The narrative is running away in every direction.
There are popular infographics floating around today which explain to the layman that deepseek invented MoE.
Shit... I was just looking at the DeepSeekMoE paper, almost exactly from 1 year ago:
https://arxiv.org/html/2401.06066v1
The red star in the first chart is so funny.
^ This is publicly new information, and the 2nd part especially contradicts consequential rumours that had been all-but-cemented in closely-following outsiders' understanding of Sonnet and Anthropic. Completely aside from anything else in this article.
Also, though it's not "new information": "Making AI that is smarter than almost all humans at almost all things [...] is most likely to happen in 2026-2027." continues to sail over everybody's head, not a single comment about it, even to shit on it. People will continue to act as though they are literally blind to this, as though they literally don't see it.
> People will continue to act as though they are literally blind to this, as though they literally don't see it.
Or like they see it and have learned the appropriate weight to give unsupported predictions of this type from people with a vested interest in them being perceived as true. It not only not new information, its not information at all.
I find that really, really, really hard to believe, given the current approaches.
we're getting used to it
and, personally, i think, if any CEO in this industry dare to say "we won't get a super AI in 2028"
many people will be disappointed, some people will be scared, and one prisident be pissed off
Anthropic is, according to themselves, using RLAIF... which is basically using LLM as a judge / reward model. So maybe he means that the models they use for RLAIF are not (much?) more expensive than Sonnet 3.5 (e.g. previous Sonnet or Haiku 3 :)).
Do you have a link to Anthropic saying they use RLAIF?
https://www.anthropic.com/research/constitutional-ai-harmles...
> Thus, in this world, the US and its allies might take a commanding and long-lasting lead on the global stage.
This is just like Captain America giving a pre-battle speech to the Avengers—it's so inspiring! Hail Hydra!
> the US and its allies
What allies? Is he aware that his current President is alienating and antagonising almost every single country out there that can historically be considered an ally?
I just had to upvote this for the tongue in cheek call to arms, even though I am really sad that the US is turning out the way it is today...
Democratic USA must prevail against Communist China by having better models.
edit: (Please read this in a sarcastic voice, I think it's a crazy idea!)
OpenAI already announced they will be copying DeepSeek implementation, and that they will make sure not to share it with anyone else than themselves
otherwise it will become to a multipolar world(or 'bipolar', said the article)
that would be horrible, a multipolar world? what a nightmare
As users, we would have more choices, more competition, and access to cheaper AI, more freedom (if DeepSeek stays open), higher quality models.
The same with chips, if no sanctions on exports to NVIDIA, we would have access to Huawei Ascend chips (used by DeepSeek to have lower runtime costs)
Yes, last time we had a "multipolar world" it was absolutely terrible.
Powerful people are playing chess and the rest of us are pawns - whatever could go wrong?
Depending on your perspective, pawns that get promoted?
Worry not, we have lots of drugs we can prescribe for this in the US
Multipolar world is probably going to have multiple wars on the "edges of control", similar to the one now raging in Ukraine.
It may also have some positives, too, but situations with several similarly sized powers competing for primacy have historically been pretty bloody. Unlike Apple and Google, the US, China, Russia etc. have military forces at their disposal.
US unipolar hegemony has averted an all-out war, but it has also been very bloody, or at least immiserating, for people at its periphery.
Multipolarity doesn't imply we go back to the 1910s. The idea would be to strengthen multilateral institutions that put a check on things like the World Wars.
Humans are pretty much the same as they were in the 1910s, or the entire history before.
If there is anything that keeps us from tearing each other's throats out, it is a) democracy, which in most countries makes going into a war of aggression somewhat harder (I know, the US is a huge exception, but Trump I. was partially elected on the basis of Clinton being perceived as a war hawk), and b) the intuition that weapons are now so destructive that there is nothing to win, except scorched earth.
Still that didn't stop Putin from launching his war; miscalculation such as his ("the enemy is a paper tiger and will fold immediately") is very possible even today.
IIRC the same delusion led Saddam to invade Iran in 1980.
> democracy, which in most countries makes going into a war of aggression somewhat harder
no, it's just because some countrys lost their power of aggression, nothing related to democracy
the US is not the only one with dirty hands, every NATO member is part of it
just check France, the only reason it doesn't colonize african anymore is it's too weak
not a single french voter gives a fk about what france is doing in africa
well, from the historical perspective, it's more easy(or equally hard) to achieve a peaceful multipolar world than a eternal unipolar world
i mean, all 'rule-based international order' died in history, the Holy Roman Empire, the First Empire of France, and the ancient Chinese empire died in every 300 years
even there is no China, even the empire doesn't implode, will africans and south americans be willing to mining for the west for all their lives? there will be Sankaras and Castros in every several decades, every hegemony has an end
or, american exceptionalism?
Can you name a single period of peaceful multipolar world in situations where the Great Powers could actually reach one another militarily?
I mean, the fact that the Romans and the Chinese never clashed over future Kazakhstan is not a proof of a peaceful multipolar world. It is simply a corollary of the fact that neither empire was able to expand into the steppe.
so it's equally hard
and, i know you guys don't like it, but that's what communism aims for
it's an extremely glorious and great undertaking that has never been done by our predecessors
I just want to ask if when we talk about export controls on AI chips, is this going to create export controls on general consumer goods in the near future.
Given Moore’s law and the efficiency that will clearly come from optimizing chips for AI and competition increasing the amount of VRAM on these devices to run models locally.
Is creating an export regime today going to mean that in 3 or 4 years general smart phones and high end laptops are all going to be subject to export controls?
Keep in mind that at one time computers that consumed entire building are less powerful than my apple watch.
I don't expect so. Export controls are aimed at chips needed for training, not inference.
The H20 with 96GB of memory is currently available in China. We're a ways off from restricting consumer devices.
That would be a great way to give Non-American companies an advantage in the market. Imagine if Apple has to constantly worry about complying with US export controls while Samsung, a Korean company, can just ship whatever they want with zero additional paperwork as long as they avoid US suppliers. Multiply that over all consumer goods that might include AI.
Dario certainly had no qualms about working with authoritarian regimes when he worked for Baidu!
Also, if the US seeks a permanent advantage over China by 2027 with self-replicating AI, so much so that it can dictate whatever terms militarily. Wouldn't that just force China to send a few cruise missiles to TSMC before that happen?
>Dario certainly had no qualms about working with authoritarian regimes when he worked for Baidu
Maybe it's worth listening to a Western AI researcher who worked for a Chinese company and came to the conclusion "China must not beat us to AGI/ASI".
> force China to send a few cruise missiles to TSMC before that happen?
Why on earth would they do that? China considers Taiwan to be rightfully theirs and would consider TMSC to be a valuable asset.
Unless the author makes a compelling case about why AI breaks the MAD status-quo between nuclear powers, I will assume that their appeals to NATSEC are an attempt to artificially create moat for their company.
I also find it counterintuitively reassuring that we have already had the power to blow up the planet for decades. But at least in theory - GPT10 level AI could help develop missile defense systems that actually work at scale, which would eliminate the MAD status quo, no?
What about the missile guidance / launch systems it could develop? The question at the at the limit of development becomes is nuclear warfare slightly attacker biased or slightly defender biased?
I imagine that argument is a pretty easy sell. Of course governments will want to create moats that protect their incumbents.
AI is potentially so much deeper than MAD status quo.
At least they're all coming out of the woodwork and start telling more details about their own training runs, costs, efficiency rates and so on. Interesting to see how an open-weights model could force their hand like that.
The technical summary is excellent. But I worry that prominent US voices couching AI progress in Cold War style war rhetoric is likely to be... self-fulfilling. Surely some of the people paying closest attention to these articles are Chinese politicians and military leaders - it would definitely seem alarming to me if the Deepseek CEO kept writing things like: "whoever reaches this threshold of AI intelligence first will accelerate into world military domination - we must beat the US."
There's a reason why Oppenheimer & co. weren't non-stop publishing internationally accessible op-eds during WWII about how important it is that Germany not develop a scaled uranium fission bomb before the US.
If you read enough history books (and have a passable grasp of German, and tolerate the weird lead types used in the press for those days) you will be able to find _a lot_ of opinion pieces from our recent history that play along the same lines, but for earlier industries (like automotive).
People _really_ ought to study more of our past.
Links and book recs are great - sounds like a helpful perspective
Not all of it is digitized. I happened to spend a few weeks in Germany once and had access to a library in Dusseldorf.
From my perspective it’s very convenient that all the new information and competition supports his existing priors that:
1) we need to do various forms of regulation to entrench US closed source market leaders, which happens to increase his company’s value
2) the best way towards improvement in models is not efficiency but continuing to burn ever increasing piles of money, which happens to increase his company’s value
It's shortsighted to believe that export controls are relevant. China will be able to manufacture as good or better chips than the US in the not too distant future anyway.
This is a choice between the lesser of two evils, or what could be called "drinking poison to quench thirst." Export controls might lead China to develop its own advanced chips, but without them, China will certainly gain powerful AI capabilities
Point being, they will gain those capabilities either way. The only question is whether we in the US choose to hamstring ourselves with bureaucracy while we watch them do it.
I think it is even worse than this.
The only way this would work is to create a kind of great firewall of America keeping open source Chinese models out while the remainder of the globe uses something like DeepSeek.
It would literally be the dumbest soft-power move in the history of the US.
reminds me of every other tariff based failure in the past...
That is impossible. There's several layers deep of sole suppliers for them.
China only has to attack Taiwan to deny the US of TSMC’s production capacity.
Prediction: they will do this as soon as they see sufficient domestic chip production.
The attempts to inshore fabs to the USA is too little too late.
And it doesn’t matter if Trump defends Taiwan and makes China back down - the fabs will have been bombed to bits.
As it is, trump would probably just do sanctions and tariffs, which - like Russia - China will expect to weather.
While Russia has certainly found ways around the sanctions, look at their economy... Putin is spending his reserves out from under the next decade. They're rather fucked.
Clearly china could last longer under similar conditions, but they're also looking weaker than they did 5 years ago.
> And it doesn’t matter if Trump defends Taiwan and makes China back down - the fabs will have been bombed to bits.
What would prevent the US from directly attacking China's domestic fabs in this scenario? Defending Taiwan doesn't seem as proxy-ish as Ukraine considering China considers it their territory already.
That's a extremely dangerous, disproportionate and self destructive strategy. What would prevent China from bombing US targets as revenge?
Both mainland and Taiwan belong to China, it's written even in the Taiwanese constitution. The controversy is in the definition of China, basically they are supposed to unify mainland while the mainland wants to unify them. A balance will be reached at some point either peacefully or violently.
Peace is more likely if the US stops pouring oil on fire.
The US should stay out of this situation if its leadership has its own people's interest in mind. A hot war between US and China will cause many millions to die and possibly end of human civilization. Let that sink in.
It seems unlikely the US will take military action to protect an ally.
It seems even more unlikely that, should the US be prepared to bomb land targets in China, that they would do anything beyond overt military targets that are being used to attack Taiwan.
When trump has recently been asked what he’ll do when China attacks Taiwan he has just said sanctions and tariffs.
Maybe so. Thank you for the perspective.
My going hypothesis is that the US cannot politically solve their Modern Monetary Theory national debt blackhole problem and is therefore itching for WW3 to be able to "reset" debts when the dust settles. Taiwan seems like a good opportunity for that, but maybe there is some sanity left in Congress.
> All of this is to say that DeepSeek-V3 is not a unique breakthrough or something that fundamentally changes the economics of LLM’s; it’s an expected point on an ongoing cost reduction curve. What’s different this time is that the company that was first to demonstrate the expected cost reductions was Chinese.
Excellent analysis and I largely agree. The part where he looses me here is the certainty in the claim around fundamental changes to the economics...
This is open weights with a (largely) open recipe. Arguably, Meta has already been at it but now there are two labs producing frontier models at this level. A wishful analysis could imagine that there _might_ be more and that _could_ result in a qualitative shift in how the frontier model research is being done and hence affect the economics but Dario seems to have largely ignored this aspect.
Not sure why?
> Export controls serve a vital purpose: keeping democratic nations at the forefront of AI development.
As an outsider looking in, I’m not thinking of the USA as a particularly democratic country any more. And I didn’t think its export controls were about democracy, but rather a trade weapon.
"Making AI that is smarter than almost all humans at almost all things will require millions of chips, tens of billions of dollars (at least), and is most likely to happen in 2026-2027."
What an incredible claim to just slip in two-thirds of the way through! Where have I heard it before?
1965, H. A. Simon: "Machines will be capable, within twenty years, of doing any work a man can do."
Stepping back, it feels like the techies working on ai have been celebrating the expense of it all. That techies have been wanting bigger and bigger clusters as some kind of bragging rights and that suits the suits who see it as a most to protect their crazy stock validations. The money was being made not by the miners but by nvidia selling shovels they could name the price of.
Necessity is the mother of invention. Deepseek is demonstrating that if the incentives are aligned to want efficiency then good engineers - and there are plenty of them in every country - can make things fly.
The author seems to be of the opinion that the creators of DeepSeek will either be unable to, or will not see the value of optimizing the 'second stage' RL component of the 'new' (post pre-training RL) way of training frontier foundation models. Every competent programmer in China is now looking for low level ptx optimizations for EVERY SINGLE STAGE of the pipeline. They will now, likely not publish any of it.
There isn't a "the pipeline". You'd have to work at DeepSeek for your low-level work to affect it.
>Even if the US and China were at parity in AI systems, it seems likely that China could direct more talent, capital, and focus to military applications of the technology. Combined with its large industrial base and military-strategic advantages, this could help China take a commanding lead on the global stage, not just for AI but for everything.
China spends 1.5% of its GDP on its military. The US spends 3.5%. I get that the two countries are engaged in competition for dominance, but why is China the bigger threat here?
Also, there are a great deal of groups and institutions out there that are pushing for more diplomacy, more cooperation, and a ratcheting down of tensions. If Dario is going to get political anyway, why not go that route?
It's also funny that there's a scary implication "China could do XYZ and focus on military applications" when Anthropic and Palantir teamed up last year to offer Claude models to US Defense orgs
https://investors.palantir.com/news-details/2024/Anthropic-a...
That post is really dishonest, which is expected from Anthropic's CEO talking about the competitors. Dario conveniently omits that R1's approach was validated months before o1-preview came out, and DeepSeek releases the weights and architecture unlike Anthropic and OAI which produce black box models in the name of "safety". Then he tries to diminish their innovation, dismissing it as a normal step they're also capable of, saying v3 and R1 are competing with models that are 9-10 months old. Which is not really true.
And then there goes the usual techno-feudal interjection:
>If China can't get millions of chips, we'll (at least temporarily) live in a unipolar world, where only the US and its allies have these models. It's unclear whether the unipolar world will last, but there's at least the possibility that, but there's at least the possibility that, because AI systems can eventually help make even smarter AI systems, a temporary lead could be parlayed into a durable advantage
So Dario envisions a unipolar world where a black box corporation (his own one, of course) controls a perpetually self-improving machine god and distributes scraps to the masses that are too unwashed and dangerous to control it themselves, while at the same time preventing other countries from breaking out and making their own self-improving machine god because they're evil. A boot stomping on your face forever, regardless of who you are.
That's coming from an "ethical, helpful, and honest" guy.
>To be clear, the goal here is not to deny China or any other authoritarian country the immense benefits in science, medicine, quality of life, etc. that come from very powerful AI systems. Everyone should be able to benefit from AI. The goal is to prevent them from gaining military dominance.
And of course he immediately tries to deny that thought.
Note: Author is the CEO of Anthropic.
This piece ignores recent news that DeepSeek is doing inference on Huawei’s Ascend 910C GPUs.
The simplified export control analysis Amodei gives here gets a lot more complicated when increasing export controls potentially ends up spurring additional R&D in Nvidia competitors.
There are lots of people working on ASICs for inference. There are already people using FPGAs for some vertically integrated AI stacks, so it's only a matter of time until we start discussing literally hard-coding models (or flashing them) to hardware to make them go faster.
I find this article to be exceptionally self-serving.
AI is a race to the bottom, and a race to over-spend. Seeing this as competition will bankrupt the players, make enablers rich, and end up killing a lot of people when the need to bolster confidence in its capabilities overrules basic common sense.
He’s lost my Claude pro subscription and my respect with this post
Someone might just be the Trojan horse … encouraging the current path from within …
Maybe if US wasn't led by a bully that care's not about democratic ideals, but only on the rule of power it'll be easier to support the notion of US the leader of the free world. But just take a look at how he handled Denmark-Greenland issue, Pure bullying. In the Trumpian world, US has no allies, only vassals.
I'm kind of tired of the entire narrative that "US good, everyone else bad", as if only US deserves to hold powerful technology because it will be used for "the greater good", rather than employ it in military applications, as if Anthropic didn't partner with Palantir.
And it amazes me how many people can't seem to see past the "US is good, everyone else are bad" smoke mirror.
I'm sure most of us agree with you, but independently of that, please don't take HN threads on generic tangents, and certainly not generic nationalistic flamewar tangents. The problem is that it's tediously repetitive and inevitably turns nasty.
"Eschew flamebait. Avoid generic tangents."
First- the author of this blog post is the CEO of Anthropic, an American AI company. Of course they are going to argue for export controls if it can hurt their competition. They get to benefit from anticompetitive practices with none of the legal risk! So it's not even about keeping the "bad guys" from having things, it's really about making more profit. Also, you can typically assume that anything a CEO publishes is in a pursuit to raise stock prices, it has nothing to do with morality. If CEOs were moral people most of them wouldn't have become CEOs (there are probably a few CEOs that are truly good people, but I'll bet there are more "wolf in sheep's clothing" CEOs that use an image of morality to improve their company's reputation).
Second- aside from the specifics of this post, you don't have to believe that the US is good to see the logic of export controls. There are a number of countries that are openly hostile to the US (Iran and North Korea), and also some that are semi-privately working against the US (like China). These countries will take any advantage that they can- they will steal IP and use the US's development resources to catch up to the current state of the art at a reduced cost. So you don't have to put this in terms of good vs. bad, just think of it in terms of things that benefit the US and things that don't benefit the US. Whether you think people are evil is irrelevant, it is very logical for the US to put export controls on things that it doesn't want to give away for free. The US wants to preserve its advantage and make it as costly and difficult for these other countries to catch up.
[flagged]
Please don't take HN threads further into flamewar hell, and certainly not nationalistic flamewar hell. It's repetitive and evokes worse from others.
https://news.ycombinator.com/newsguidelines.html
Absolutely. Chinese companies shouldn't have chips because their government has "committed human rights violations, has behaved aggressively on the world stage". And the US government hasn't?
To say the two are remotely comparable is either insincere or naive. Like I'm not saying the US always acts utterly virtuously, but they aren't running concentration camps and chemically castrating entire ethic groups.
They also aren't imprisoning people for speech, which is important when we're discussing who we want controlling AI in the future.
Do you remember when the US lied to everyone about WMDs and then caused the deaths of at least 100k Iraqi civilians?
Or overthrew democracies in South America, leading to the refugee crisis it now refuses to take responsibility for.
The entity that is least problematic (of the available options) and should be entrusted is the European Union.
> overthrew democracies in South America
Don't forget Central America and Iran; supporting military dictators in Southeast Asia (Marcos in the Philippines and Suharto in Indonesia); invading Haiti and raiding it of the entirety of its gold reserves, condemning its people to poverty -- one could go on and on.
I mean hell, one of the states (Hawaii) didn't even join willingly, it was annexed by President McKinley by force.
DeepSeek model not providing answer on the Tiananmen Square and ChatGPT providing answer on ethnic cleansing of Palestine are two sides of the same coin.
ChatGPT can't provide answers on any current events because it has to be updated first.
I made no such comparison, but if you really want to start comparing government crimes and human rights abuses, I think you'll find we can make a stronger moral case arguing for chip bans for US companies than Chinese companies. https://www.youtube.com/watch?v=8BOBKTmaQ9M
[flagged]
if you're chinese you'll know some chinese were sterilized because they _didn't_ believe in a religion
muslims were pardoned of child policy because of their ethnics, it's a privilege of them
people can not imagine something they never saw, try dig your own history, you'll find why it's always about religonal genocide, religonal sterilization, and religonal concentration camps
I am not talking about the one child policy here. I am talking about the ongoing sterilizations of Uyghurs.
then it will be more ridiculous
there're 55 minority ethics in china, and china sterilization one single ethic and spares all others, for what???
double check your source next time, adrian zenz? wuc? or the us embassy?
[flagged]
do _we_ look suspicious to you?
how many of us in this room with you right now?
I find this line of reasoning a whole lot less convincing now than I would have 2 weeks ago, and it wasn't all that convincing two weeks ago.
When the controlling party of the US is discussing concentration camps for immigrants while also happily calling to revoke birthright citizenship and deport citizens of the US that criticize Trump or the Republican party (eh - who am I kidding - it's the maga/heritage party now)...
It feels like we have no legs left to stand on here, and the support was DAMN shaky to begin with, seeing as we were routinely knocking over governments that we see as inconvenient in our geopolitical sphere of control for the last 100+ years.
Essentially, I think your argument is about 3 elections stale.
> To say the two are remotely comparable is either insincere or naive. Like I'm not saying the US always acts utterly virtuously, but they aren't running concentration camps and chemically castrating entire ethic groups.
They absolutely did directly and indirectly several times in their history. A few of those times were recently:
> In September 2020, it was revealed that ICE had performed mass hysterectomies on immigrant women in several detention centers, reminiscent of the long-standing US policy of sterilization of black and brown women. 2
> The US currently operates a system of slave labor camps, including at least 54 prison farms involved in agricultural slave labor. Outside of agricultural slavery, Federal Prison Industries operates a multi-billion dollar industry with ~ 52 prison factories , where prisoners produce furniture, clothing, circuit boards, products for the military, computer aided design services, call center support for private companies. 1, 2, 3
> In the present day, ICE (U.S._Immigration_and_Customs_Enforcement), the police tasked with immigration enforcement, operates over 200 prison camps, housing over 31,000 undocumented people deemed "aliens", 20,000 of which have no criminal convictions, in the US system of immigration detention. The camps include forced labor (often with contracts from private companies), poor conditions, lack of rights (since the undocumented aren't considered citizens), and forced deportations, often splitting up families. Detainees are often held for a year without trial, with antiquated court procedures pushing back court dates for months, encouraging many to accept immediate deportation in the hopes of being able to return faster than the court can reach a decision, but forfeiting legal status, in a cruel system of coercion. 1, 2
> During the 2020 coravirus pandemic, it was found that a law that empowered police to arrest those for not social distancing, lead to 80% of those arrested being black and latino.
etc., etc., etc.
https://github.com/dessalines/essays/blob/main/us_atrocities...
True, providing complete military support, with special deliveries of bombs and ordinance to Israel, knowing full well that it will go towards obliterating a city and killing tens of thousands of civilians is more humane. It's not genocide if an ally is doing it!
Just wait. By year 2 of Trump 2.0, China is going to be the global good guy that Europe and Latin America make their new best friend.
Year 2? We're almost there at _month_ 2.
Oh, people are thinking it now for sure, but it takes a while to retool your relationships and build out a new trade network so you can safely tell a major trading partner to fuck themselves to threats of 100% tariffs.
Well, the British did it to Europe, so I'm pretty sure Europe can emulate their winning strategy :)
Isn’t it like… week 2?
You're right. It's been that intense - or I've actually been feeling the shift since the election ended, not the inauguration.
If Europe and Latin America want to buddy up with the concentration camp gang, I really don't care. We do not need fascists and communists as our partners.
Funny you mention fascists. As for China being a concentration camp gang, that's not going to matter when their former big brother/ally is just relentlessly bullying them and trying to extract all sorts of concessions under duress, and the alternative has demonstrated to be a reliable ally and partner.
Once again, I don't really care if Europe decides that is who they want to be friends with. If your country thinks that trade is more important than the country currently directly committing a modern holocaust, I don't want us to be allied with you at all.
As opposed to the country openly supporting the ongoing genocide in Gaza, which would not be possible without its help?
Yeah. I personally think we should just sink the middle east into the ocean because not a single problem there will ever be resolved in a meaningful way.
Ironically, if Trump keeps running this playbook, it's going to end with a USA/Russia/Israel bloc and a Europe/LatAm/Africa/China bloc. Which one of those sounds like the evil empire to you?
Indeed, the US has committed human rights violations pretty consistently.
https://github.com/dessalines/essays/blob/main/us_atrocities...
Note how interesting it is that you can read about these on the internet in the US on US hosted servers by US companies. At least we own up to it and don't just throw you in a mental asylum like they do to anyone who dares to stand up to the CCP.
There's a big difference between allowing the information to exist and "owning up to it." The US just has a different strategy for dealing with it.
And that strategy, notably, does not include camps or asylums for people like you and me.
[flagged]
Article 1 Section 8 does not enumerate Congress with the power to provide healthcare, so Medicaid should onoy be a state level program.
I know removing it harms those at the low end of the income spectrum, and that is a bummer, but I really would prefer Congress ammend the Constitution instead of just passing laws of which they have no authority to pass.
And thus is shown that legal reasoning is not equivalent to moral reasoning
In my opinion, it would be more moral to allow charity to cover healthcare for low wealth individuals over a government provided solution.
But even supposing a government funded solution is best, why push to the feds what could be solved by the state?
That is... bonkers. At the risk of feeding a troll, I find it difficult to take the argument that voluntary contributions to public health through charity is more "moral", which in the case of health care I would argue is equivalent to "effective", than a system implemented by an organization with enormous power (the government) which at least theoretically has a direct duty to its stake holders (voters), who in turn have the power to enact change in that system if it's not serving them? (by voting)
They just kind of have to give up their power and hope?
Buddy, one cursory glance at history will show that hoping gets you nowhere.
I assure you I'm no troll. I think difference in our perspectives is that I form the basis of my "moral" at the individual vs. the collective.
"Moral" for me means that individuals are empowered to own their private property and should only need to sacrifice it to society for public goods, where I take the economic definition of a private good: nonrivalous and nonexcludable.
Forcing all persons to pay taxes to cover healthcare for only a subset of the population is, to me, akin to forcing all your friends to give to the charity you like because you like that charity and want it to be able to do more, where that more is a level of spending above what you can or are willing to provide.
Economically, this creates deadweight loss: people's individual preferences are violated because they are forced to spend money for something of which they receive no benefit, or at least the direct/indirect utility occupies a lower utility than the opportunity cost of those specific taxes.
I'm not saying that such a policy won't result in undue death. But since I use the individual as the basis for morality, I consider it more moral to have some death than it is to steal from others to prevent it.
Well according to the unitary executive theory the US president has complete authority over all government actions and congress is only able to give legitimacy to the president's actions. Under this regime the president would be allowed to stop the execution of any law passed by congress.
I think this is insane and a complete destruction of the balance of power written in the constitution that congress can't enact enforceable laws but only "suggestions" for the president
> UET is a constitutional law theory that gives the President sole authority over the executive branch.
I think UET applies only to the Executive branch, which to me makes sense as he is the head of the Executive branch.
It would seem a violation of checks and balances for Congress to be able to install unfirable persons in the Executive. The checks and balances come from Congress's subpoena and investigatory powers, which can ultimately result in impeachment and removal of office of the President if he is derelict in his duties of executing the laws set forth by Congress.
Though I would agree that UET would violate the balance of the Constitution if it applied outside the Executive branch.
To be fair you should have socialised healthcare 70 years ago like the rest of the civilised world.
[flagged]
Do you really think the US is going to come out on top as the "lesser evil" if we start listing all the "bad things" each side has done related to human rights abuses and behaving aggressively on the world stage? https://www.youtube.com/watch?v=8BOBKTmaQ9M
[flagged]
I live here, so I want it to be the best.
I'm not sure how that relates to original comment. Do you mean you want everything that is or could be better than American technology banned/destroyed so we stay the best...?
Like, any global hegemony will be increasingly corrupt given the power that gives, IMO.
Same. Be nice if we were the best by our own merits though, not by holding back our peers.
In a zero sum game kinda way, the game that America understands.
He's unlikely to be writing from a genuinely ideological point of view. He's playing a zero-sum game, in that every chip that goes to China is one that Anthropic and/or its cloud provider doesn't get its hands on.
So it makes sense to him to argue for export controls using whatever rhetorical flexes and flourishes he can come up with.
There's no zero sum games in a growing industry like this.
As long as the semiconductor fabs are running at capacity, yes, it's one big zero sum game. If you win one chip, I lose one chip, and vice versa.
This situation is temporary, of course. China previously had a large incentive to get their own leading-edge nodes into production, and now they have a Manhattan Project-size incentive.
They aren't running at full capacity on AI chips though - TSMC's main customer is iPhones as far as I know. So you could take away production time for them, though it's still zero-summy.
But TSMC is building new leading edge fabs right now.
[flagged]
It might already be more dangerous in the sense that it is unpredictable. China is autocratic and powerful but their goals are known and do not change at a whim.
You can't build diplomatic and trade ties in quicksand.
I mean, in the end, USA is still better than the alternative.
The civilizational alternative is Europe. The economic alternative, that we can debate...
We are not alternative to anything. We are so hopelessly behind USA, and it doesn't seem it can be changed.
Choice is - USA or China. And I certainly know in which world I'd rather live.
Europe, especially UK, France and Germany degraded so much it is no longer any alternative.
Economically, perhaps. Brexit certainly didn't help Europe as a whole, and German, French and Italian politics have been off-kilter for years, but it's still, by and large, an ethical and civilized place to live in.
I will freely grant that Euro money goes to completely obsolete heavy industries and luxury items rather than technology innovation (and blame all politicians for missing out on technology), but there are outliers like ASML that almost make up for the whole thing...
UK, German and French cities are no longer civilized places to live in.
With respect this is a fairly daft comment
How so?
U.S. and Chinese hegemony are just different kinds of bad. We could put the pros and cons up and debate it endlessly, but in the end ranking them is pointless, humanity should just be trying to make a better alternative
USA is still better for whom?
Can't remember which countries where bombed by China in last 50 years, but I can give you plenty of examples for the USA
I don't know. Living in a country where being able to shoot people is an absolute, immutable, inalienable right granted by God himself (more so than anything else in the Bill of Rights) but where access to food, shelter, healthcare and education are all expensive privileges subject to the whims of the free market, and where half of the population doesn't even believe viruses are real, and where you don't even need the popular vote to elect a President (it's just nice to have) because the electoral system is a series of 18th century compromises to keep slavers happy... doesn't seem like the best of all possible worlds to me.
> Living in a country where being able to shoot people is an absolute, immutable, inalienable right granted by God himself (more so than anything else in the Bill of Rights)...
2A provides access to weapons. It doesn't allow people to murder one another.
> but where access to food, shelter, healthcare and education are all expensive privileges subject to the whims of the free market,...
Article 1 Section 8 doesn't enumerate Congress the power to provide any of those things.
What's more, I beieve the Federal government shouldn't. Rights should only ever be negative unless being a counterparty to positive action taken by the government (e.g., it makes sense to provide public defenders since it is reactionary to the government trying to take away your rights).
Healthcare, education, etc. are better served by the private sector with charity as the supplement for the less fortunate.
> and where you don't even need the popular vote to elect a President (it's just nice to have) because the electoral system is a series of 18th century compromises to keep slavers happy...
The electorial system is designed to prevent tyranny of the majority and is biased towards inaction. This seems better than a pure democracy to me because most government action is bad: inefficient, introduces moral hazard, etc.
In a system of pure democracy, the bottom 51% can vote to tax the top 49%'s income away. Then the next iteration the new bottom 51% can vote the same. Eventually this could lead to purely even distributions. That seems wildly worse for society: all incentive to produce would be void.
Most people - even most Americans - aren't libertarian minarchists who believe everything the government does is harmful by default and that free market capitalism is an unalloyed moral good. History has shown that the kind of governance you want rarely works out well for anyone but oligarchs.
I don't believe everything the government does is harmful. Personally I'm quite fond of the FAA despite the many that complain about TSA.
But most of what the government does introduces dead weight to economic transactions or moral hazard. This is especially true when the government tries to provide social benefits beyond taking over a product or service that could perfectly well be provided by the private sector.
I would have little argument against the government regulating that private healthcare providers cannot discriminate on the basis of preexisting conditions. It would increase costs uniformly across the industry. I do have arguments for providing healthcare on the basis of income, as I believe the program would be better served by private charity. Those who argue the contrary do so only because they want to force all citizens to contribute to the private charity (run publically) of healthcare-for-the-poor, because they want to solve someone else's problem but don't want to have to bear the burden themselves.
It does if you're quickly converging into the Handmaid's Tale universe. May the odds be ever in their favor - I'm quite happy we don't carry guns to make it through the day...
You do not have a right to other people's work. I should not be forced to give you food because you are too lazy to do it yourself. If you do not work, you do not eat, just like Jamestown.
Yes, all the poor are simply lazy, and you yourself have never ever benefited from the collective labor of society.
You do not deserve to eat just by existing (unless you are disabled or medically cannot work). Saying this is true implies that you have a right to other people's labor, which you do not.
You forgot "are a child".
Land, water, air and light used to produce food are not owned by any particular human either. The monopolization of that land (at the point of a gun most often) is the beginning of the issue. The overuse of that land, often subsidized heavily by taxpayers, in order to generate exorbitant profit for the few is a rot in human society. Those owners are the actual ones who don't work and expect a reward for it
> In the end, AI companies in the US and other democracies must have better models than those in China if we want to prevail.
How about everyone has access to the best model because it's Open Source and openly developed?
he's coping so hard.
CEO of an AI company asks for regulation against a country that is out-competing him.
Are we still capitalists?
I think the answer is - no.
another American Nationalist that see competition as a "threat" to their monopoly, how Great!!
I call this capitalism when its convenient for us, socialism when we are struggling.
I hope one day, China starts blocking advanced tech from US, for the good of the whole world
I hope they learn from the terrible job of stewardship of tech the west has done and don't repeat our mistakes.
[flagged]
[flagged]