Hacker News Reader

Comment by ericmay | Hacker News Reader

I remember listening to that and thinking, this sounds like a guy who is primarily focused on selling chips and is brushing away legitimate concerns in a way that, I just did not find very compelling. A little bit too much protest you might say. Not to say there isn't or wasn't validity to his arguments, but in light of the current circumstances it seems to me that those arguments are quite a bit less valid than they were, and I've certainly swung my personal opinion from being on the fence to a hard no to selling these chips to Chinese buyers.

If the Chinese Communist Party are going through such great lengths to obtain these advanced Nvidia chips, it further invalidates a lot of what I've been reading here and elsewhere about China "catching up" or that being the leading models isn't what matters (commoditization). You don't do stuff like this unless you really need something.

View on HN

Talking to friends who worked in diplomacy - honey traps are extremely common and often tried quite routinely. It makes sense if you think about it. The failure rate on these is going to be fairly high and even if it’s a small success (ie a one night stand sort of thing) you can get a little compromat. The cost is probably fairly negligible, a lot of this will be fairly low risk as far as spying goes.

also, men are redickulous. And they can go in any number of beneficial direction from turning informant, selling GPUs to the chinese, or just a pee tape, etc. Just like a lottery of fun ways to screw with an adversary.

Jensen was a tech geek? who had a midlife crisis…

> honey traps are extremely common and often tried quite routinely.

WSJ reports that U.S. intelligence says Chinese intelligence approached Alice to become an asset after she met Jensen in March 2024. So, she was an opportunistic info source, not targeted in advance.

My favourite honey trap story comes from the cold war. A Russian diplomat in Denmark was the target of Danish intelligence. They tried to honey trap him at a party, but he didn't respond to their agent's advances. So, the Danes supposed that he might be gay, and sent a male honey trap to the next party, but again got no response from the Russian. They shrugged their shoulders and gave it up.

Unbeknownst to them, the Russian was actually at that time trying to make contact with western intelligence, as he was keen to defect! He was just a bit too awkward to respond to flirting.

He did eventually manage to defect to Britain, and died in 2025. His name was Oleg Gordievsky.

I think it's worth remembering that China isn't a monolithic bloc, and CCP doesn't drive everything, that there are tons of private enterprises there that would benefit, even if the Chinese government strategically would prefer to push Huawei to grow the domestic capability. There being a ton of private demand internal to China doesn't invalidate Jensen's view that it's a strategic blunder to let Huawei get a foothold in the market. Within the CCP, there's also a lot of inter-regional competition to show growth in their own regions, because that's the primary qualification for advancement within the party.

That said, I don't have any special insight into the specifics of who was driving the demand backing Megaspeed.

I find this hard to believe. Covid made it clear the CCP is very much in charge and in control.

Which part? You find it hard to believe that there are private enterprises there that don't always do exactly what CCP wants? Or that there're conflicting incentives inside the CCP, so that "what CCP wants" isn't simple?

You can't just take away someone's boogeyman like that.

Political officers are assigned to most Chinese tech companies.

there's a good analysis that says Alice was a fixer and/or white-glove. those are common-in-China roles that have to do with obfuscating big financial operations, not necessarily with any political (CCP) involvement. obviously, CCP approves, if tacitly, everything that happens in China.

> invalidates . . . . China "catching up"

China catching up does not mean they are not still behind.

Also, China is a big country with many companies. While some may be trying to catch up via local development, others may still prefer getting western hardware right now e.g. in order to get a competitive advantage over those catching up.

Over the last 40-50 years they haven’t had the try that hard to get that technology, the overwhelming majority was served up on a plate.

I'm a bit out of the loop, what part of what he said is wrong? He said they have more power and the more capability to produce chips, so they can essentially just brute force things vs the US if they wanted to: is that not the case?

No it's not the case because if they could do it they would already have done it. Selling them Nvidia chips doesn't change that.

But here's a link to the podcast episode if you want to give it a listen. It will inform a lot to the discussion taking place here and help you understand a few of the things he said that are probably now wrong: https://www.dwarkesh.com/p/jensen-huang

If they don't need NVIDIA then why are they attempting so hard to bypass export controls.

If they don't need NVIDIA then why is the now former CEO of NVIDIA trying so hard to allow exprt controls to be relaxed.

And it's because brute forcing only allows you to not fall behind, but you can't catch up to frontier models or surpass.

Could it not be true that they could brute force it if need be, but it would be much easier to use the better chips if they can get them?

That's like asking why China is building nuclear and coal power plants when solar is so much cheaper. They are doing all of the above.

Let's assume for a second that the compromat story on Huang is true, how much would that have cost? It could be everywhere from 10k to 10m $, depending on how difficult it was to get access, but it would still be a rounding error for AI invest.

Chips aren't what maketh SI anyway. The US is ahead because of a trifecta of having the best (most) hardware, best data, and best talent. You might also throw in business culture. But the same way Google is leading from behind in the US AI eco system, China is behind the frontier on LLMs, but laying foundational work in robotics, multi-modal models and integrated factories that might prove a big advantage down the line.

Best talent, doubtful keep holding that belief. Right up until 2050…

I am trying to separate the current situation from his comments. Perhaps that is a mistake. He's trying to sell chips but that doesn't make him wrong.

They are catching up. Don’t delude yourself. This minor innocent little faux pas is the start of the AI bubble, it is in some ways poetic…

One other little crack in the dike. Was Sam Altman saying regular computer users shouldn’t have memory all memory should be allocated to data centers.

What will be the third crack in the dike?

But what’s the point ultimately? Delaying their progress by a couple years?

Not selling them chips virtually guarantees that they will learn how to make their own, very quickly. The end result of that is much, much worse for the USA.

(and could leave Taiwan in a much weaker position in the near future)

This. And this happened before, see the China Sovjet split

Who says we won't just progress as well and they stay behind a couple of years?

Where is the evidence that China is catching up considered a fiat d accompli?

Do you presume that China is destined to always at least catch up in every sector or field ever? That stinks of Chinese Exceptionalism to me.

If they could just "catch up" they would, and whether we sell them chips wouldn't matter because they would want to create their own Nvidia and export chips and make money too.

> Where is the evidence that China is catching up considered a fiat d accompli?

The ever closing gap between the capabilities of frontier models vs chinese open weight models.

Models aren’t everything. You need physical infrastructure at scale for inference, which is where the economic power comes from. Chinese hardware makes double the heat and needs at least double the energy input.

even if true, luckily for them, double the heat and double the energy input aren't an issue, and they have more physical infrastructure at scale too

And 3 years ago they were producing no heat and required no energy because they did not exist. Clearly they are catching up

It's like the ever closing gap between the smartest kid in class and the kid who sits behind him copying his work.

That's not a strategy that leads to actually being the best.

Exactly. And we better keep it this way. The fact that we can block them from copying us today means that they will try that much harder to come up with their own answers. Remember, they are miles ahead in manufacturing hardware and the software is not a problem anymore. Whatever moat cuda represented is gone now.

> fiat d accompli

The phrase is “fait accompli”, without a D, and is French for “accomplished feat”.

> Do you presume that China is destined to always at least catch > up in every sector or field ever? That stinks of Chinese > Exceptionalism to me.

Suppose you want to build something that requires technical knowledge (a GPU, a fighter jet, ...). You have engineers, but they don't yet know how to build it.

If no one on earth has ever built this thing, you need to consider the possibility that it can't be done at a cost you can pay. Perhaps it violates some law of physics, or it can't be manufactured without knowing some chemistry/physics that we are 100+ years from discovering.

If a company in some other country is already building this thing, you can be sure it is possible to build it. That is a critical piece of information! It means you don't need to discover (much) new physics. It means you can buy/steal a working copy and look at it to learn about it. It means there are engineers you can hire/spy on to understand their design process. It means you can more easily convince a bank/government to fund development, as they have proof your goals are achievable.

Inventing a thing is much harder than inventing a clone of a thing. The idea that China doesn't have the people and organization to copy things other countries make is the opposite of Chinese Exceptionalism. It is the belief that the Chinese are uniquely incapable of engineering work. This is obviously not true.

> The idea that China doesn't have the people and organization to copy things other countries make is the opposite of Chinese Exceptionalism. It is the belief that the Chinese are uniquely incapable of engineering work. This is obviously not true.

Which broadly leads to two conclusions:

They can't do it. If they could they would. If you could invest and create your own Nvidia, ASML, TSMC, &c. you would do that especially if you are China due to its stated goals and actions it takes that we can all see toward those goals.

They can do it and so selling them the chips doesn't matter other than helping to speed up their ability to do it, so we don't want to sell them the chips and thus speed up their capabilities.

But your viewpoint also holds true for the United States and Europe. Anything China can do we can also do, otherwise you're just supposing Chinese Exceptionalism.

It feels like Space Race thinking, honestly. If the US delays China and ends up with an AGI god a few years in advance, that can still be a significant advantage even if China can catch up. This especially would hold true if a theoretical AGI could maximally leverage RSI so the improvement curve becomes exponential.

It's much closer to nuclear arms-race thinking.

If the US or China developes an AGI that can leverage RSI, the game is over. It is over in the same way that WW3 is over as soon as the missiles are in the air: it doesn't matter who has the fastest or highest yield missiles, because there is a finite amount of time before there is nobody left to care about the advantages of one side. Except at least WW3 is over in about an hour, whereas an AGI could take years or decades to complete whatever thing it has been misaligned to desire.

AI alignment research is still in its alchemy era, and global cyber security is underfunded and bad. As a species, we simply do not have the tools to deal with AGI that can self-improve.

What stops another power from developing its own AGI a year or two later? Or from catching up once there are two competing RSI AGIs? The Soviets were a few years behind on getting nukes but they still got them.

The first AGI protecting its turf and sabotaging or corrupting the second AI to be its agent.

I think the verdict is still out on whether having an AGI God fly out of your butt would serve as a strategic advantage in the great power struggle.

AI has been identified as strategic national priority for both the Chinese and United States.

NVIDIA’s commercial interest is pretty clear… it’s beneficial to them. to sell everyone to maintain the dominance of CUDA. That is not necessarily the national interest. And remember that espionage and infiltration is not a one-sided affair, the US may well have the ability to sabotage Chinese development efforts. You just don’t know.

End of the day, corporate CEOs are very powerful, but they’re not as powerful as the government. My guess is that Jenson truly didn’t think that this individual was a spy, and figured that the questionable legality of chip exports to Chinese proxies could be addressed by ingratiating the administration.

Why wouldn’t they have done that by now then? Where is their ASML or TSMC? They’ve had decades to catch up already.

Their ASML and TSMC are...ASML and TSMC.

> They’ve had decades to catch up already

No, they haven't. They spent the first decade after WW2 repairing the damage left behind by the Japanese and their Civil War. Subsequently they managed to starve tens of millions afterwards in several missteps, and then righted themselves.

They then spent three to four decades lifting about thrice the US population out of extreme abject poverty, and then in the last 20 years catching up. And catch up they have; they are already the leaders in green technology and EVs, already manufacture some of their own chips, already have indigenous military aerospace capability, which they are now expanding to civilian use. They make almost everything in the average household.

> Not selling them chips virtually guarantees that they will learn how to make their own, very quickly. The end result of that is much, much worse for the USA.

Are you proposing we guarantee access to avoid potential access?

Keeping them dependent upon you guarantees that they never surpass you and that you can milk the relationship for all its worth economically and geopolitically. When they have their own fabs spun up it's possible that you are the one eventually stuck with inferior tech, and you lose all sort of revenue + leverage you'd otherwise have from the relationship.

Except it's not a guarantee. They can still surpass you. It might be a little harder, a little less likely - but it's still completely possible.

It make sense. China had X amount of demand. If the demand is met they are satisfied, if the demand remains unmet they seek alternatives. That seeking of alternatives primarily means huge amounts of money, that would go to Nvidia, going to direct competitors of Nvidia to "chip" away at their competitive advantage.

The long term result is the same regardless of who makes the chips but the real strategy is delaying China's ability to surpass the US. It takes a lot of time and money to essentially copy the entire infrastructure required to build an Nvidia GPU. Nvidia has enough demand as is.

China is a trading partner but there is still a serious catalyst for war in the form of the South China Sea. If we went to war with China in 5 years time, why would we sell them the tech they need to match our ability now? Let them figure it out on their own.

If the US wanted to delay "China's ability," wouldn't the strategy be to give them just enough GPUs to keep them reliant on Nvidia (and make developing alternative tech locally seem less critical), but not enough to develop superior models?

Friends close and enemies closer.

An economy kept dependent on American tech is unlikely to drop bombs on America and vice versa. Co dependency prevents war.

This isn’t what China does though, the moment new tech crosses the border they reverse engineer and start producing it locally. IP protections are meaningless there. How many times does this have to happen to sink in?

What needs to happen is that the people who are getting wealthy from the offshoring of manufacturing and labor to China need to stop getting wealthy. Until then any and all excuses to pretend that China will operate in good faith will be levied to try and make people seem like it's in their best interest to hollow out the USA to hand over to China.

The problem is, the man in the mirror and it ain’t just the Chinese that have taken advantage, the Japanese and the Koreans are pretty good too, and so are the Taiwanese and it’s only a matter of time for one or two countries in Southeast Asia to rise up too, and if it wasn’t for Russia’s stupidity, both Russia and the Ukraine would have certainly produced what three or four very, very good AI models to the mix…

Soon, China, if they aren’t already, will be the premier developer of Thorium reactors courtesy of Oak Ridge, Tennessee and they also are now number one or at the least 1A when it comes to high-speed rail, and if the Russians and the Ukrainians weren’t fighting each other, that high-speed rail probably would be already built out all the way to eastern Europe.

Huawei has become a super company basically because of all the sanctions, Wxmt will become the dominant memory maker worldwide in the next five years due to shortsighted thinking in the West.

People are bending over backwards to come to try to find some deep reason, but I think the reason is simple - the people in charge are not particularly farsighted nor insightful. It's simply short-term thinking without consideration of long-term consequences - it's not limited to just businesses.

For another example, the reason Chinese are not allowed on the 'international' ISS is because of the Wolf Amendment. [1] The idea was that if we just allowed them onto the ISS they'd steal all of our elite tech or something. Within a decade they had launched their own space station and carried out various revolutionary space activities.

And of course all of that happened while the ISS itself continues to deteriorate with no replacement in sight. And when China did an unprecedented sample return mission from the far side of the Moon and offered the whole world access to samples, US researchers had to jump through a bunch of hoops to receive them, again thanks to the same law.

It's just myopia.

[1] - https://en.wikipedia.org/wiki/Wolf_Amendment

Davidson Window...

https://en.wikipedia.org/wiki/Davidson_window

This is good evidence sanctions don't really work. Critical sectors will find ways around them and it only punishes the rank and file working class. If China or any sanctioned nation wanted those nvidia chips, it has ways to get them. So the military and intelligence get them, but the people don't. Same with sanctions on Russia now, which has not significantly slowed its military machine, but in fact its military machine is coming out with constant innovations (new Gerans, better EW, more spy sats, etc). I remember seeing a post mortem of some Russian drone. It had a Jetson chipset Russia was not supposed to be able to get. I imagine the smuggling market is far bigger than we might think and sanctions far less effective than we think.

Same when you sanction food imports. The elites who make decisions will always be fed. Sanctions are almost always an attack on the working class who do not make the high-level political decisions the sanctions are trying to address. Not to mention the idea of "making life hard on them so they revolt" almost never works.

Sanctions should be a bigger conversation. They are almost never needed and especially trade tariffs and such are always used for the most corrupt cronyism.

> Same when you sanction food imports. The elites who make decisions will always be fed. Sanctions are almost always an attack on the working class who do not make the high-level political decisions the sanctions are trying to address. Not to mention the idea of "making life hard on them so they revolt" almost never works.

There’s also the minor quibble that restricting food exports to starve another country are illegal under international law.

Nobody is starving in Russia (at least unavoidably).

That sanctions only harm the civilian economy is not an argument against them. Kinetic sanctions agains oil refining are also primarily impacting the general population and the military have enough oil. But they still impact the ability of the Russians to fund the war. Civilian infrastructure and economic assets are used by militaries too.

Reply on news.ycombinator.com