[HN Gopher] Huawei releases an open weight model trained on Huaw...
___________________________________________________________________
Huawei releases an open weight model trained on Huawei Ascend GPUs
Author : buyucu
Score : 299 points
Date : 2025-07-02 07:36 UTC (15 hours ago)
(HTM) web link (arxiv.org)
(TXT) w3m dump (arxiv.org)
| gkbrk wrote:
| Weights are available on gitcode [1].
|
| [1]: https://gitcode.com/ascend-tribe/pangu-pro-moe-model
| ssddanbrown wrote:
| Just a warning, the license [1] specifically blocks EU use:
|
| > 3. Conditions for License Grant. You represent and warrant
| that You will not, access, download, install, run, deploy,
| integrate, modify, or otherwise use the Model, directly or
| indirectly, within the European Union.
|
| [1] https://gitcode.com/ascend-tribe/pangu-pro-moe-
| model/blob/ma...
| p2detar wrote:
| What's the reason behind this? What am I missing?
| worldsavior wrote:
| The company is Chinese, I presume that's why.
| Toritori12 wrote:
| Is Llama4 also chinese(0)?
|
| [0]: https://dionwiggins.substack.com/p/llama-4-is-
| banned-in-the-...
| axus wrote:
| Worse, it's American!
| Iolaum wrote:
| Most likely EU AI act regulations they don't see any value
| in bothering with.
| dr_dshiv wrote:
| Hence, the lack of European innovation in AI
| p2detar wrote:
| I don't agree. Tools like DeepL were and still are better
| than Google Translate long before chat bots became a
| thing. The French-made Mistral AI is pretty decent as
| well.
| seanw444 wrote:
| Saw some benchmarks recently that put Mistral well behind
| basically every other competitor. Don't have them on hand
| unfortunately.
| p2detar wrote:
| FWIW, I refactored 500+ Junit 4 to Junit 5 tests with
| locally running Mistral 8B on an M3 MBP. It worked
| flawlessly, but surely I cannot attest for other use
| cases.
|
| Edit: it was 8B, not 7B.
| lawlessone wrote:
| What specifically is it about the law slowing innovation?
|
| Is it something these companies do that they worry
| violates it?
| Keyframe wrote:
| Even so, why would the licensor put it in and force it
| through a license. It's on the licensee to check the laws
| and regulations they themselves operate in.
| pfortuny wrote:
| Wow this is a huge caveat: a guarantee that they are using
| data and not complying with GDPR.
| imiric wrote:
| Shocking. At least they acknowledge it.
| trhway wrote:
| It isn't acknowledging. It is just a legalese to wash
| their hands away from following whatever EU restrictions
| and requirements may be applicable here otherwise.
| 2Gkashmiri wrote:
| If you download to your PC and run locally, what will
| happen?
| rusk wrote:
| The same thing that usually happens when you violate TOC
| ...
|
| Probably safe enough on your own computer, but could have
| consequences if it's a work computer.
| 2Gkashmiri wrote:
| >could have consequences if it's a work computer.
|
| consequences for employer who "might" get a license audit
| done on their machines.
|
| does it really happen so often that a random employer in
| the eu would have to be concerned?
| rusk wrote:
| consequences for you when you unwittingly open up a back
| door or expose your organisation to a data breach
| viraptor wrote:
| You'll be both breaking their licence and potentially
| your local European data laws.
| lukan wrote:
| " potentially your local European data laws."
|
| If run locally, why?
| johnisgood wrote:
| I called him out in another thread. It makes absolutely
| no sense. He is talking against himself, judging by his
| comments.
|
| To answer your question, he modified my comment (see the
| parentheses):
|
| "> The point is, who gives a damn about (doing an illegal
| thing) in reality, on their (private property where
| nobody is likely to see that)?"
|
| So... at best what he said is purely theoretical. He
| admitted it himself: "nobody is likely to see that".
| Though I am not sure I agree with it, but then again, in
| reality, no one gives a fuck, at least not in Europe.
| viraptor wrote:
| There are likely multiple potential issues here, but one
| specific example: Processing and storage of PII without
| consent/authorisation is not allowed, regardless of
| whether you do it yourself or for others. And you can't
| guarantee that this model does not contain private
| information hoovered up by accident.
| 2Gkashmiri wrote:
| so what. will they send cops after you.
|
| breaking license will do what? whats up with licenses and
| violations? you and me are random people on internet
| Barrin92 wrote:
| >breaking license will do what?
|
| The same thing breaking any license does. If you do it in
| your basement, nothing by definition. If you incorporate
| it in a service or distribute it as part of a project,
| well then you're on the hook. (and that is what license
| holders tend to care about)
| seydor wrote:
| energy consumption
| nord73 wrote:
| Picture your PC as a cheery little planet in the EU's
| cosmic backwater, sipping a digital Pan-Galactic Gargle
| Blaster. You download Pangu Pro MoE, hit "run," and
| expect to chat with an AI wiser than Deep Thought.
| Instead, you've hailed a Vogon Demolition Fleet. Your
| machine starts moaning like Marvin with a hangover, your
| screen spews gibberish that could pass for Vogon poetry,
| and your poor rig might implode faster than Earth making
| way for a hyperspace bypass.
|
| The fallout? This AI's sneakier than a two-headed
| president--it could snitch to its creators quicker than
| you can say "Don't Panic." If they spot your EU
| coordinates, you're in for a galactic stink-eye, with
| your setup potentially bricked or your data hitchhiking
| to a dodgy server at the edge of the galaxy. Worse, if
| the code's got a nasty streak, your PC could end up a
| smoking crater, reciting bad poetry in binary.
| shit_game wrote:
| To translate for those not familiar with the writings of
| Douglas Adams:
|
| nord is suggesting it's possible that the physical
| computer running this model could be used as a "hub" for
| potential spyware, or be overloaded with workloads that
| are not related to the actual task of running the model
| (and instead may be some form of malware performing other
| computational tasks). It could potentially perform data
| exfiltration, or act discriminatorily based on your
| percieved location (such as if you're located within the
| EU). At worst, data loss or firmware corruption/infection
| may be of concern in case of license violation.
|
| I'm not sure I would _outright disagree_ that this as
| possible, but with some caveats. I would think the reason
| that the license stipulates that usage within the EU is
| forbidden due to the EU AI Act (here is a resource to
| read through it: https://artificialintelligenceact.eu/ai-
| act-explorer/).
| 2Gkashmiri wrote:
| how will the "open weights" know that the pc is running
| within EU? again, you are not talking about software that
| actually runs in your pc but the file that the software
| reads and loads into memory for its own use.
| tliltocatl wrote:
| Sorry, sounds like total bullshit. The weights aren't
| going to do anything. And if you are worried about the
| code, with current deployment practices of curl | sudo
| bash there are much more low-hanging fruits out there.
| That's not even mentioning the possibility of running the
| model on a PC without internet access (no matter how good
| the new Chinese AI is, it's still not good enough yet to
| convince you to let it out of the box).
| arccy wrote:
| with mcp, and the right tools, it's effectively already
| out of the box
| tliltocatl wrote:
| Don't give it mcp then (and I struggle to understand why
| would anyone give a stochastic model such access even if
| it is trained on very American NSA-certified hardware
| approved by Sam Altman himself).
| 2Gkashmiri wrote:
| you can use existing apps that take random huggingface
| files, do you expect weights to somehow coax the software
| to do exfiltration?
|
| same. i call bull on this.
|
| remember how they convinced huawei was public enemy
| without evidence because nokia and others were unable to
| compete with them?
| 2Gkashmiri wrote:
| >If they spot your EU coordinates how.
|
| can anyone give a technical answer how will weights get
| to know this fact?
| jaggs wrote:
| No it's actually worse. Approximately three seconds after
| you install the model in offline mode on your computer, a
| small detector van will come and park outside your door
| with an antenna on the roof, and relay your position to a
| Chinese ICBM for immediate targeting.
| HSO wrote:
| a siren will go off and in 10 secs your computer will
| explode
| sigmoid10 wrote:
| GDPR is not the issue here, the new AI act is. Since this
| is an open-weight release it is not bound by the training
| data disclosure rules, but it probably didn't go through
| the evaluation that is required above a certain number of
| FLOPs. That's why many recent big player model releases had
| a staggered release in the EU.
| seydor wrote:
| does anyone comply with gdpr & Ai act? Even for mistral i m
| not sure, the best we can say is "we don't know"
| waffleiron wrote:
| There is not a single AI model that fully complies with
| GDPR. How can you inform everyone, even those not named by
| actual name but otherwise identifiable, that their data is
| being processed and give them the ability to object when
| the data they train on isn't public.
|
| Literally the same for all other open weights, this is just
| legal ass covering where most others don't even do that.
| yard2010 wrote:
| There's something nefarious about this.
| johnisgood wrote:
| And who thinks that, for even a second, that an European (in
| this case) will not download, install, and try to run this
| just because the LICENSE says you can't?
|
| FYI, this is not intended to be offensive to Europeans, I am
| European myself. That is not the point. The point is, who
| gives a damn about the LICENSE in reality, on their PERSONAL
| computer? Serious question.
| _joel wrote:
| Quite a few, actually.
| viraptor wrote:
| The licence is not there for enforcement from their side.
| It's a legal protection for Huawei. Essentially "We told
| you it's not for the EU. If you get sued don't try to put
| it on us."
|
| Also any company of a serious size will have lawyers
| interested in licences of everything you're running.
| johnisgood wrote:
| I am not talking about companies. I edited my comment.
| Emphasis on "[their] PERSONAL" computer.
|
| I know that companies would probably not. But
| individuals?
| viraptor wrote:
| > The point is, who gives a damn about (doing an illegal
| thing) in reality, on their (private property where
| nobody is likely to see that)?
|
| I'm not sure which part of that you find confusing. Some
| people will estimate benefit>risk and won't care.
| johnisgood wrote:
| What? I do not find anything confusing. You live in a
| Marvel world if you think a LICENSE is going to stop
| people from using a product. But like you said, it is not
| intended to be for enforcement purposes, but Huawei is
| trying to save its own ass.
|
| So what is your answer? Mostly companies only? That is a
| fair answer, but you are the one who said this:
|
| > You'll be both breaking their licence and potentially
| your local European data laws.
|
| Again, who cares, dude? Companies might, but individuals
| probably give a rat's ass. So why leave that comment?
|
| And just for the record, if you quote someone, quote them
| verbatim, otherwise it is not a quote.
| linotype wrote:
| Breath.
| johnisgood wrote:
| Been there, done that.
|
| That said, I agree that it is my fault that self-
| contradicting virtue signaling hypocrites always find a
| way to irk me.
|
| And I think it is good for the world to know that the
| LICENSE often means jack shit, unless when companies of
| significant size are involved.
|
| Again, we all agree that they put it there to cover their
| own asses, not that Europeans cannot download, install,
| and run their product, right?
| linotype wrote:
| Yes we're all aware of the unlimited rights of Europeans,
| including subjecting the rest of the world to annoying
| cookie notifications.
| johnisgood wrote:
| Europeans definitely do not have unlimited rights, and I
| do not agree with the annoying cookie crap either.
| coldtea wrote:
| It's probably the inverse:
|
| they might license it to companies in the US, but don't
| want to have to deal with the changes and bureucracy
| needed to support individuals.
|
| The statement's purpose is to say the equivalent "if
| you're a European and do run it, it's on you, this is not
| a product we release or support for the European market,
| don't expert support, liability, etc".
| johnisgood wrote:
| I get that. What I do not get it some other commenters
| "scaring" Europeans attempting (or thinking) to run this
| product.
|
| I mean, this other commenter literally said:
|
| > You'll be both breaking their licence and potentially
| your local European data laws.
| benreesman wrote:
| I'm really torn on the whole thing. I consider myself a
| patriotic American and would never do anything to
| undermine the security of my country or its allies (using
| the same definition of national security that the serious
| sworn oaths use, "all enemies foreign and domestic",
| which makes NSA backdoors that compromise American
| devices squarely a "domestic enemy").
|
| But loyalties don't change facts and China is where
| serious hackers are rising on merit, doing a lot with
| limited resourves, giving zero fucks about empty slick
| talk.
|
| If we wanted to hobble the PRC's technical rise we should
| have subsidized wasteful NVIDIA use and had Altman/YC be
| in charge: they'd still be gladhanding about how to pump
| their portfolio companies sticker price and avoid
| "systemic shocks" to the stock market anchored on NVDA.
| johnisgood wrote:
| Just for the record, I would never run this product, but
| it has nothing to do with the LICENSE itself.
| coldtea wrote:
| Well, people say such things even for watching pirated
| shows, which to be truthful, almost everybody does...
|
| Some just are narc types.
| m3kw9 wrote:
| Why would they open themselves up to liability in the
| rest of the world where it is allowed?
| Arnt wrote:
| It's plausible deniability. Someone at Huawei presumably
| thinks there's a chance that exporting this to Europe might
| be a legal problem at some point in the future. So they
| added a restriction, enough for plausible deniability.
| coldtea wrote:
| It's not exactly "plausible deniability" in the common
| sense of the term.
|
| It's not supposed to make them appear as plausibly
| denying that some European can download and use this.
|
| It's role is to signal that if someone does, it's on him,
| not them, and he wont have any support, liability claims,
| etc as if they could if it was a product intended for
| their use.
| coldtea wrote:
| Legalese and licenses aren't to make sure no X will
| download/install/or run something.
|
| It's to make it a matter of legal record that you stated
| they should abstain.
|
| Copyright warnings on music and DVDs never stopped people
| pirating them either.
| johnisgood wrote:
| I know. That is why I do not get what is the big fuss
| about.
| randomNumber7 wrote:
| Most companies abide the law. So no self hosted LLM for
| europeans.
| johnisgood wrote:
| ... what? Self-hosted LLMs are precisely for individuals.
| randomNumber7 wrote:
| A lot of companies and research institutes in the EU
| would like to be able to use a locally hosted LLM for
| their employees so they don't have to worry what data
| they give away.
|
| Also it is not rational for any individual to buy the
| hardware for running a serious LLM and then let it idle
| 99.9% of the day.
| johnisgood wrote:
| Why would not these companies or research institutes in
| the EU not be able to run locally hosted LLMs for their
| employees though?
| bit1993 wrote:
| Try selling pirated copies and see what the warnings are
| really about.
| coldtea wrote:
| When CDs and DVDs were a thing people wanted, there were
| people selling pirated copies on every corner, so...
| Quarrel wrote:
| For those that would not remember, this was a real thing in
| the late 80s and 90s relating the cryptography.
|
| There were serious laws limiting the export a "modern"
| cryptography software from the USA.
|
| Some of us had to face up to the serious challenge of
| connecting to an FTP server and downloading PGP and risking
| violating US law to download a software package.
|
| A few years later we had to decide "Do you want the secure
| Netscape, or the insecure Netscape?".
|
| I'm sure we all chose the ethical choice.
| johnisgood wrote:
| You should elaborate on this for the unacquainted.
| randomNumber7 wrote:
| A lot of companies and research institutes in the EU would
| like to be able to use a locally hosted LLM for their
| employees so they don't have to worry what data they give
| away.
|
| They will certainly not violate EU laws and also probably
| not the licence.
| coliveira wrote:
| I wonder if you'd say the same if the license were coming
| from Microsoft of Apple...
| johnisgood wrote:
| Why would I not? Of course I would.
| knowitnone wrote:
| I doubt the Chinese ever care about licensing so I would not
| care about following their license
| JKCalhoun wrote:
| I know people get upset when _open source_ is used when _open
| weight_ is more correct (happily here _open weight_ is
| specifically being applied).
|
| My question: is _open weight_ even interesting? What does that
| really offer? Does it allow one to peer into the biases (or
| lack thereof) of a model? Does it allow one to train a
| competing model?
|
| Would _open source_ be something different and preferable -- or
| are "weights the new source" in this LLM world we are finding
| ourselves in?
|
| I'm trying to educate myself.
| HPsquared wrote:
| I really don't get why there's any confusion. These models
| are LITERALLY compiled binary data. Weights are definitely
| not source. Source is "the source from which the thing is
| generated" i.e. the training data (or a script to assemble
| it) and all scripts, procedures etc required to make the
| binary blob.
| FirmwareBurner wrote:
| Will the US also sanction countries buying Ascend GPUs*?
|
| *) They're NPUs not GPUs, since they can't render graphics
| chenzhekl wrote:
| US Warns That Using Huawei AI Chip 'Anywhere' Breaks Its Rules
|
| https://www.bloomberg.com/news/articles/2025-05-13/us-warns-...
| throawayonthe wrote:
| https://archive.is/GrvVo
| teloli wrote:
| Oh boy do they love the free market!
| roenxi wrote:
| They may have misunderstood it as free from Huawei.
| kesor wrote:
| > The agency's Bureau of Industry and Security said in a
| statement Tuesday that it's also planning to warn the public
| about "the potential consequences of allowing US AI chips to
| be used for training and inference of Chinese AI models."
|
| The A.I. war is intense, and you can clearly see who is s**
| their pants here.
| v5v3 wrote:
| I am presuming that only applies to US companies? To stop
| them using it at their sites outside of the USA.
| kesor wrote:
| No, it applies to anyone who wants to do business with the
| US. https://www.bis.gov/press-release/department-commerce-
| rescin...
| atlasduo wrote:
| So, it is "buy made in US products or face sanctions" now?
| This will end well, I bet.
| tliltocatl wrote:
| This is just beyond stupid. Sanctions as in "not letting them
| use our advantage" might make some sense, but "not letting us
| use their advantage" is just another level of retardity.
| quantum_state wrote:
| Very much indeed ...
| ronsor wrote:
| Yes.
|
| Huawei GPUs are illegal worldwide now, according to the US
| government.
| falcor84 wrote:
| Cue the Team America: World Police music
| firesteelrain wrote:
| Yes because Huawei uses US technology. Huawei can't be
| competitive without US and EU technology (ASML, Tokyo
| Electron, Applied Materials, Lam Research, KLA, Cadence,
| Synopsys, or Mentor, etc)
| FirmwareBurner wrote:
| Here's a thought exercise: If EU and US are the inventors
| of silicone tech at the basis of all products, why are they
| struggling to beat China in the free and open AI market and
| need to resort to kneecapping it?
|
| Like, if you're so good at something why are you scared of
| the unproven newcomer beating you?
| firesteelrain wrote:
| The issue is not fear but strategic control. The United
| States and European Union form the backbone of the global
| semiconductor ecosystem, from design software like
| Cadence and Synopsys to manufacturing tools from ASML and
| Applied Materials. Huawei's chips depend on this
| infrastructure. When entities use US-origin technology to
| build AI hardware, they fall under export control
| regardless of where the chips are used. This is not about
| competing in a fair market but about controlling the
| development and use of advanced computing in areas like
| defense, surveillance, and critical infrastructure.
| esafak wrote:
| You just explained _why_ they are scared.
| firesteelrain wrote:
| Not scared - aware. Strategic assets like compute and
| semiconductor capability shape global influence.
| Controlling those assets is about ensuring they are not
| used to undermine national security interests. That is
| not fear. It is policy rooted in leverage and long-term
| risk assessment.
| hgomersall wrote:
| Maybe we could write some code that we can use to use them as
| gp-npus (graphics-processing NPUs).
| yanhangyhy wrote:
| huawei's Ascend GPUs is the only choice for many chinese company
| for now. Huge win for huawei.
| xbmcuser wrote:
| The world needs Huawei and China to get competitive on its node
| size with TSMC and Nvidia.
| amelius wrote:
| That would be great, if you ignore geopolitical concerns. Alas,
| AI technology is a double-edged sword and any competition in
| consumer space will likely be mirrored in an arms race which
| (given their current manufacturing capabilities, cheap labor)
| China would win.
|
| Anyway, they would need to duplicate ASML first which will
| probably not happen in the foreseeable future.
| xbmcuser wrote:
| Don't count the Chinese out when ever they have been side
| lined they have copied then innovated and gotten ahead. Now
| their economy in real terms is actually bigger than the US
| economy and more well rounded so sanctions or any kind of
| restrictions boost their local companies by removing any
| competition.
|
| And looking at how the recent wars and skirmishes in Ukraine,
| Israel and Pakistan/India have gone I think western military
| superiority is no longer real. In a conventional war of today
| US will lose and most likely nuke the other country so I
| think its best for the world if China gets so powerful that
| USA accepts that it no longer is the sole super power and we
| can avoid a nuclear war. As that is where we are heading
| either a multi polar world or a nuclear holocaust.
| bilbo0s wrote:
| _which will probably not happen in the foreseeable future_
|
| Can you elaborate on why this will not happen in the
| foreseeable future?
|
| Because in my version of the foreseeable future I see it
| happening quite readily.
|
| EUV is not the magic that everyone believes it to be. It can
| be replicated by us, (the US), at our convenience. It can
| even be replicated by the Chinese and Japanese. (Personally,
| I'd throw the Koreans and the Russians into that pool as
| well.)
|
| But that ain't even the point. The point is that, in that
| particular game, there is always more than one way to skin a
| cat so to speak. It's not at all a foregone conclusion that
| EUV is the best way to skin a cat, and it'd certainly be a
| bad bet to assume it is the _only_ way to skin a cat. Those
| are the questions I would hope that we, (the US), are
| focusing research efforts on, and we should assume that China
| is also focusing research efforts in that direction.
|
| PS - Please no one bring up Trump gutting research. Here I'm
| only speaking of clear strategic research priorities in an
| ideal, (ie - collaborative), political environment.
| Obviously, politically erected structural concerns impact the
| viability of any research strategy we want to implement. I'm
| just talking about what _I_ think would be ideal.
| owebmaster wrote:
| I have the impression that if the US removed the chips exports
| control, the government of China would impose a imports
| control. They have so much more to gain from creating a real
| contender to Nvidia/TSMC/Apple/Google.
| seydor wrote:
| Sic transit gloria nvidii
| HPsquared wrote:
| Linguistic deep lore: "invidia" is Latin for "envy".
| madmask wrote:
| Also italian
| fakedang wrote:
| That was the reasoning behind the NVidia name choice btw -
| the cofounders wanted the competition to be envious of their
| company's capabilities.
|
| https://en.m.wikipedia.org/wiki/Jensen_Huang
| fennecbutt wrote:
| Explains invidious, I suppose.
| WithinReason wrote:
| So it's extra fitting!
| smitterle wrote:
| Sorry for the nitpick: I'd expect Gen Sg to be nvidiae - i is
| for o declension
| seydor wrote:
| and what would be the plural of nvidia?
| Intermernet wrote:
| "now don't do it again!"
| abdulhaq wrote:
| 5 years of Latin finally pays off
| Quarrel wrote:
| Best thing I've read today.
|
| Bravo.
| hearsathought wrote:
| Sic transit gloria nvidiae
| snickmy wrote:
| thank you for conjugating correctly the 1st declension to the
| genitive, we all, that wasted many years studying latin, te
| salutant!
| randomNumber7 wrote:
| So the Huawei Ascend 920 is produced by SMIC on a 6 nm process.
|
| I always thought sceptical of the US sanctions, but that they
| backfire so fast is insane.
|
| Out of China's perspective it might make sense to take out the
| wests AI capabilities soon.
| seydor wrote:
| even if their chips weren't as good, what stops a company from
| training a large model for a very long time in less capable
| hardware? Is there a way to overcome memory limitations
| somehow?
| umgefahren wrote:
| In general yes, you can (and do) shard the model over
| multiple GPUs. If you want to do that yourself look at
| DeepSpeed or FSDP . There is a communication overhead though
| and the speed at which the GPUs can communicate is key. Thats
| where NVLink comes in btw. So yes, it's actually what you can
| and do do. However this limits your ability to iterate on the
| models quickly and from what I've read a lot of times the
| foundational labs throw out their models because by the time
| they are done training they are already outdated.
| checker659 wrote:
| NVLink isn't magic
| rfoo wrote:
| > the speed at which the GPUs can communicate is key
|
| Guess what a telco equipment company is good at :p
| corimaith wrote:
| Did they do it with EUV or just more multipatterning DUV? If
| it's the latter then it's not anything particular to worry
| about or unexpected.
| bayindirh wrote:
| If the chip is working, has a comparable performance and
| slows the gap or reverses the trend, does it really matter?
| imtringued wrote:
| Intel struggled with 10nm and you're shrugging off 6nm as if
| it was nothing?
| nickysielicki wrote:
| Process node naming is completely made up. It's been this
| way for a very long time.
|
| https://community.cadence.com/cadence_blogs_8/b/breakfast-
| by...
|
| It has nothing to do with the size of anything.
| linotype wrote:
| In fact, process node naming has always been made up.
| nickysielicki wrote:
| Not really? The physical gate length matched the node
| name all throughout the 2000s.
| linotype wrote:
| even that concept was in fact made up.
| trq01758 wrote:
| "China has EUV" would be news of the century. No doubt they
| are trying.
| saubeidl wrote:
| The only ones that have EUV are the Europeans.
|
| Despite all the anti-EU propaganda in these parts, the entire
| AI stack is powered by us.
| roenxi wrote:
| If you want the anti-EU points on that subject; the EU once
| (the 90s) represented around 40% of the world's
| semiconductor fabrication. Then the Asians decided that was
| a bit silly and the situation should be rationalised.
|
| Things like ASML are the tattered remnants of the EU being
| a powerful force in this market. China might not be able to
| replicate ASML, we'll see. But if the EU can compete
| meaningfully with the Asians at semiconductor manufacturing
| that'd be a shocking development.
| saubeidl wrote:
| I have a different interpretation of this. As our economy
| and expertise grows, we move further and further into
| foundational parts of the stack while leaving the
| implementation details to less developed countries.
|
| All modern semiconductors need ASML machines.
|
| All modern AI is based on research by Sepp Hochreitner.
|
| The www is based on work done by CERN.
|
| Most factories are run on Siemens automation systems.
|
| We've moved on from lower-level concerns like building
| chips to a higher level, directionally controlling the
| way economies and societies develop.
| ReptileMan wrote:
| And it fails spectacularly when you see EU economic
| growth.
| saubeidl wrote:
| It only fails if you see economic growth as the metric to
| optimize for, which I would argue is another symptom of
| societies still focused on those lower levels of the
| stack.
|
| Economic growth is an enabling factor for a higher level
| metric, quality of live. It succeeds spectacularly there.
| ReptileMan wrote:
| EU countries have enough money to either pay for overly
| generous social services or defense, not for both. A
| stronger economy would have been enough to pay for both.
| saubeidl wrote:
| Defense will pay for itself through economic stimuli.
| ReptileMan wrote:
| ... and other fantastic stories
| saubeidl wrote:
| It's basic keynesianism.
| patapong wrote:
| Unfortunately, apparently these lower-level concerns also
| seem to include building an industry around and
| benefitting from these innovations...
| veber-alex wrote:
| Calling TSMC "implementation details" is delusional.
|
| If the only thing you needed to make cutting edge chips
| were some machines from ASML every country in the world
| would be making their own chips right now.
| delfinom wrote:
| And the Europeans bend the knee to the US to follow
| sanctions on technology that the US doesn't even produce.
| randomNumber7 wrote:
| Why is EUV so hard to build? If the europeans deveoloped it,
| why cant the chinese (with or without reverse engineering)?
| corimaith wrote:
| EUV was an international effort by multiple (of the
| richest) countries specializing in different areas. There
| is no one country that does it all. When someone says China
| can do it, it's tantamount to saying China can outcompete
| the world. Maybe they can, but its not going to be an easy
| effort.
| randomNumber7 wrote:
| Why is the technology then owned mostly by a single dutch
| company (if I understand it correctly)?
| HPsquared wrote:
| It's not, at all. The EUV light source itself comes from
| an American company, for instance. The mirrors come from
| Zeiss. ASML integrate the parts into a system.
| botro wrote:
| The International Space Station was an international
| effort by multiple (of the richest) countries
| specializing in different areas. There is no one country
| that does it all.
|
| China built their own.
| bilbo0s wrote:
| You would think people like HN User corimaith would have
| learned from the ISS fiasco.
|
| I've gotten to the point where I no longer even listen to
| people who underestimate the Chinese. Whatever points
| they're making can be safely ignored. Indeed, strategic
| sense _demands_ we ignore them. The time for
| underestimating the Chinese is long past. That 's not the
| reality we live in any longer.
| corimaith wrote:
| On the contrary, the use of sanctions precisely
| recognizes the Chinese ability to innovate and the need
| to not give them our technology on a silver plate while
| they are it. Tens, if not hundreds of billions are
| already being funded into the chip industry's r&d and
| bringing manufacturing back home.
| randomNumber7 wrote:
| While I agree somehow I think the time to do this
| (sanctions) was the time where we actually delivered them
| our technology.
|
| Now it is more like throwing pebbles at someone who is
| stronger while hoping he doesn't come over to punch back.
| ReptileMan wrote:
| Do you remember the speed with which Chinese phones went from
| absolute crap to legitimate flagships? Their automobiles?
| From having no AI as to speak of to Deepseek? China advancing
| at breakneck speed when they really want is something to be
| worried about if you want to be technologically ahead of
| them.
| randomNumber7 wrote:
| Their government reasons with cause and effect while the
| western world argues about ideologies of different colours.
|
| If I would have to bet my money I would bet it on China.
| bayindirh wrote:
| Sanctions are generally a stopgap measure. They can't create
| any meaningful gap for a very long time.
|
| I live in a country which has experienced some hard and soft
| embargoes over the years, and let's look what has it done.
|
| - We wanted to buy drones, and denied. Now we are one of the
| biggest drone manufacturers in the world.
|
| - We denied air defense systems. We are developing a whole
| arsenal of missiles and rockets now, incl. standoff/cruise
| missiles.
|
| - We denied planes. Our 4.5th generation fighter program got a
| great speed boost.
|
| - We denied advanced naval technology. We built stealth ships,
| fast coastguard boats and all navigational systems which goes
| inside them.
|
| - We denied optical pods for drones and aerial vehicles. We
| built our own in 6 months.
|
| etc. etc...
|
| Sanctions and embargoes are the biggest catalyst for a country
| to advance their tech at tremendous velocity.
| gabrielgio wrote:
| > They can't create any meaningful gap for a very long time.
|
| That highly depends on the size and natural resource
| available to the country.
| sillystu04 wrote:
| To play devils advocate, a lack of sanctions also allows
| countries to advance technology at a tremendous velocity.
|
| The Ford Motor Company's dealings in Nazi Germany and the
| Soviet Union significantly helped the adoption of mass
| production in those countries.
| heavyset_go wrote:
| They work for countries that kneecap themselves by being at
| the whims of capital and the market. Countries that don't
| constrain themselves like that can laugh at sanctions because
| they don't matter in that context.
| sigmoid10 wrote:
| Who cares about 35% inflation when you got your own drones
| to bomb Kurds, am I right? Which btw. was the original
| reason why Obama didn't want to give Turkey combat capable
| drones. He wasn't sassy or tried to curb competition. The
| US presumed (correctly with the benefit of hindsight) that
| they would be used to go after supposed PKK positions
| inside and outside Turkey. Sanctions do work, but when an
| opponent is hell-bent on their objective, they will always
| be willing to sacrifice in order to attain their goals
| nonetheless. Even at the cost of their own population. You
| can see the same thing in Russia: They are literally
| bleeding their middle class into poverty under all those
| sanctions just to keep going in Ukraine.
| coldtea wrote:
| > _Who cares about 35% inflation when you got your own
| drones to bomb Kurds, am I right?_
|
| Orthogonal. You can have 35% inflation AND no drones or
| good arms in general.
|
| Which is worse, especially if this lack makes it easier
| to get invaded or "regime changed" into a failed style.
|
| > _Which btw. was the original reason why Obama didn 't
| want to give Turkey combat capable drones._
|
| The welfare of the Kurds, or using them as a proxy force
| against Syria, Iran, and abandoning them whenever
| convenient?
| sigmoid10 wrote:
| >You can have 35% inflation AND no drones or good arms in
| general.
|
| I guess you can always be worse off as a country, no
| argument there.
|
| >if this lack makes it easier to get invaded or "regime
| changed" into a failed style.
|
| I don't understand what you mean. Are you saying it would
| be better if a country gets US weapons while being under
| threat of being "regime changed" into an enemy? Because
| that somehow seems even worse. And that's literally what
| the US was afraid of in Ukraine.
|
| >The welfare of the Kurds
|
| I suppose they don't care if the bombs are dropped by
| Turkish or US drones. But at least the sanctions delayed
| it by ~half a decade, so there's that.
| coldtea wrote:
| You said "who cares about 35% inflation when you got your
| own drones"
|
| Based on the parent you were replying to, I take it to
| imply: "don't make drones to avoid sanctions (even if it
| means not having arms), lest you get inflation", the
| latter of which you paint as the worse outcome.
|
| So, the point I made above is that if not having
| drones/arms etc makes it easier to be regime
| changed/turned into failed state, then, yes, 35%
| inflation would still be a small cost to pay to avoid it.
| roenxi wrote:
| > They work for countries that kneecap themselves by being
| at the whims of capital and the market.
|
| Typically it actually looks like the opposite. When I look
| at, eg, North Korea or Iran - if I were going to try and
| make them wealthy it is mostly internal policies that are
| the problem and not external ones.
|
| If North Korea set itself up with single digit % company
| and income tax combined with a strong rule of law, local
| education programs and a liberal economy it would barely
| matter what sanctions were imposed on them. A tide of money
| would flow in and they'd eventually be wealthy under their
| own power anyway if not. Although it isn't obvious why
| anyone would sanction a small well run country; there is a
| correlation between sanctions and incompetent governance.
| kingkawn wrote:
| This is the most profoundly arrogant thing I've read in a
| while, kudos! It's hard to stand out on the internet
| these days
| anticodon wrote:
| Most of the sanctions are whac-a-mole game played by the
| West (mostly by USA): as soon as one of the peripheral
| countries starts developing its economy, gains more
| wealth, starts to produce something the competes with
| products of Western companies, or attempts to avoid use
| of USD (because it allows immense enrichment of USA
| simply by printing it in whatever quantities), soon there
| starts:
|
| - firstly, media attack ("they're dictatorship!",
| "they're genociding someone", etc), preparing population
| for stricter measures
|
| - secondly, color revolution, which, if successful, puts
| puppet malleable government in power, and makes the
| country ultra-poor (most of the countries, where color
| revolutions staged by US/West succeeded, became
| significantly poorer)
|
| - if color revolution didn't work, there's always an
| option to just bomb the country, because it's always
| ignored if all the international treaties and laws are
| ignored if USA or Israel bomb any other nation.
| littlestymaar wrote:
| The US has a strong record of interfering with foreign
| countries in way that advantage their interests, but at
| the same time the fact that you bring the "color
| revolution" phrase is a strong marker that you are
| reading way too much Russian (imperialist) propaganda
| disguised as anti-imperialism, which really is no better
| than feeding on the neoconservative narrative.
| Y_Y wrote:
| I presume you're talking about Turkey/Turkiye.
|
| If so, do you think it makes a difference that Turkey is a
| NATO member, and on (relatively) good terms with the Western
| powers?
|
| For all the ideological differences and geopolitcal
| nervousness I don't think the US or EU see themselves as
| potentially fighting against Turkey, and so they don't feel
| the need to go to the trouble of strict sanctions or
| sabotaging local tech.
| littlestymaar wrote:
| > and on (relatively) good terms with the Western powers?
|
| In recent years the relationship between Turkey and Western
| countries has been OK-ish (though far from stellar, see the
| S-400-related tensions or the French-Trukish tension in the
| Mediterranean).
|
| But if you look at it on a longer perspective, the
| relationship used to be very tense, first there was the
| Cyprus crisis leading to pretty harsh western sanctions on
| military equipment, and then the cold war between Turkey
| and Greece in the Aegean see, with occasional real fire air
| combat and casualties.
| firesteelrain wrote:
| What country? Your website is out of Germany. Germany is in
| NATO.
| akarlsten wrote:
| Judging by his name, Turkey
| firesteelrain wrote:
| Okay, in that case it's not a good example because Turkey
| is a NATO country.
| littlestymaar wrote:
| Which didn't stop the US and other western countries to
| embargo them after their invasion of northern Cyprus (yes
| it was in 1974, but it's when the Turkish domestic
| defense sector really started so it's not irrelevant even
| if it's 50 years old).
| firesteelrain wrote:
| Survivor bias
|
| Turkey's defense advances took decades and came with
| major setbacks
|
| Not to mention many NATO incompatibilities
|
| Just look at all the other sanctioned countries
| postexitus wrote:
| inferred from username and stated facts, Turkey.
| firesteelrain wrote:
| Sorry I didn't pick up on it. Not familiar with Turkish
| names.
| perks_12 wrote:
| Hetzner is still a thing.
| seydor wrote:
| Turkey chose to develop a large defense industry. There were
| restrictions on them acquiring some US weapons (even though
| they are NATO member) but that wasn't the motivating factor,
| it was a political choice to develop large defense sector.
|
| Other countries like Iran however, do develop their own
| drones because of sanctions
| littlestymaar wrote:
| The strict tensions imposed after Turkey's intervention
| in/invasion of northern Cyprus in 1974 played a key role in
| the development of the Turkish defense industry.
| corimaith wrote:
| If this was true then countries could just impose import
| restrictions if they wanted to "advance their tech at
| tremendous velocity".
| zorked wrote:
| This is called "import substitution" and was incredibly
| popular in the age before neoliberalism. If your country is
| rich today you probably did a lot of that in the past.
| randomNumber7 wrote:
| It also depends if the country has a realistic capability
| to catch up with their own tech. So for Congo it would
| probably be a bad move. China is a different story imho.
| polytely wrote:
| Adding more sanctions seems to be the default US move it seems,
| to the point that there are so many sanctioned entities that
| they've formed a paralel world economy.
|
| https://www.theatlantic.com/international/archive/2018/08/wh...
| blackoil wrote:
| 10s (maybe 100s) of billions that could have gone to Nvidia are
| going to Huawei, so not surprising they are able to make the
| progress and pulling along SMIC with them. Most of the
| sanctions against Huawei were because they were a credible
| threat to US companies, so again not surprising.
| seydor wrote:
| What do we know about the production capacity of those NPUs ?
| asking for a friend
| wordofx wrote:
| And the spying they have been doing.
| mtkd wrote:
| In 2023 Huawei surprised with the Kirin 9000S in the Mate 60,
| this seems to get forgotten when talking about GPU moats and
| sanction effectiveness
| lostmsu wrote:
| Same as Pixel 8 performance according to
| https://nanoreview.net/en/soc/hisilicon-kirin-9000s and
| https://browser.geekbench.com/android-benchmarks
| yorwba wrote:
| You cannot just point to some amount of a sanctioned good being
| manufactured to claim a sanctions _backfire_ , you need to look
| at the difference between with and without sanctions, and also
| consider the cost.
|
| Huawei was going to work on GPUs anyway, SMIC was going to fab
| chips anyway. How much of the total GPU compute is the result
| of increased investment after sanctions, and how much was
| already planned? And how does it compare to the alternative of
| importing Nvidia GPUs for the same amount of money?
|
| Unless Huawei is getting better performance per dollar than
| Nvidia, this is them implementing a costly workaround, which is
| the point of sanctions: increasing cost.
| like_any_other wrote:
| Increasing short-term cost, at the price of losing leverage
| in the medium/long-term. Or even ending up on the wrong side
| of that leverage, if SMIC (who was just handed control of
| China's domestic market) ends up outcompeting
| TSMC/Intel/Samsung.
| randomNumber7 wrote:
| Without sanctions noone would buy the Huawei chips.
|
| With sanctions they get free money to develop better chips.
|
| == sanctions backfire
| corimaith wrote:
| Why not just put in import restrictions then?
| quantum_state wrote:
| Sanction motivates ... import restriction discourages and
| may not work in reality ...
| corimaith wrote:
| Sounds like a weak argument to me, businesses are profit
| making entities, they certainly would be motivated to
| keep their profits irrespective of the cause.
| yorwba wrote:
| I see now, the problem is that you're confusing expenses
| and profit. If Huawei had spent a quadrillion dollars to
| make the same number of GPUs, that wouldn't be an
| incredible sanctions backfire (look at how much money they
| got!) it would be an incredible sanctions success (look at
| how much money they spent to get what would've been much
| cheaper otherwise!)
| randomNumber7 wrote:
| I was looking at it under the assumption that the goal of
| the sanctions is to keep China away from beeing able to
| compete in AI not how much they have to pay for it.
|
| Also the money the spend on top is mostly domestic. They
| can build some more powerplants to run less efficient
| chips. They can pay their workers and their developers to
| build it.
| zmgsabst wrote:
| "Compete" is an economics term.
|
| Making it cost more is keeping them from being as
| competitive -- which is the point of sanctions.
| joshuaissac wrote:
| Their customers who would have bought Nvidia chips are
| instead giving money to Huawei even though they prefer
| Nvidia. With more money, they can deliver better tech
| faster. It is costing these customers more, but in
| return, they get domestic capability with better supply
| chain security, and a possibility that it can eventually
| become cheaper than imported chips. Meanwhile, Nvidia
| gets less money from those lost customers.
| MangoCoffee wrote:
| China's Made in China 2025 already lays out which key
| industries they will go after. Semis are always a key
| industry for China.
|
| Huawei already designed and developed its own chips before
| sanctions, and SMIC is developing their next node before
| sanctions.
|
| Whether sanctions work or not is another matter
| phkahler wrote:
| >> Huawei was going to work on GPUs anyway, SMIC was going to
| fab chips anyway.
|
| Sure, but China has started dumping a _lot_ of money into
| competing with ASML as a result of the sanctions. And no,
| they are not going the super complicated route of firing
| lasers at drops of tin to get EUV. They are trying to
| sidestep that costly complexity, and if they are successful
| they 'll be providing equipment to the top chip producers
| (around the world or just their own).
| starfallg wrote:
| SMIC is running into real problems without EUV. Just because
| they are able to produce something at 6/7nm, it doesn't mean
| that it is efficient or competitive. Right now, they do it
| because of strategic considerations.
| randomNumber7 wrote:
| What is so hard about EUV that they cant develop/reverse
| engineer themself?
| WhereIsTheTruth wrote:
| https://fr.wikipedia.org/wiki/Quand_la_Chine_s%27%C3%A9veill...
|
| It's funny how it's impossible to find English coverage of this
| book :')
| dist-epoch wrote:
| China would have invested in SMIC anyway, with or without
| sanctions. They consider semiconductors critical technology.
| tonyhart7 wrote:
| china literally "steal" talent from taiwan TSMC, of course they
| "progress" fast, it literally the same people
| codedokode wrote:
| You mean people are ready to exchange a democracy and freedom
| for a tiny increase in income?
| em500 wrote:
| They're not just doing this for money... They're doing it
| for a shitload of money!
| roenxi wrote:
| This is really exciting! They're laying out an architecture that
| may mean even small players with cheap GPUs can compete with the
| majors. The idea implies that eventually crowd-sourcing an open
| AI is probably technically feasible and we've got the Chinese
| actively researching how to do it to a high standard that
| competes with the monolithic models.
|
| I was sceptical of the US sanctions but this seems like a real
| win if this can be taken all the way to its logical conclusions.
| logicchains wrote:
| >The idea implies that eventually crowd-sourcing an open AI is
| probably technically feasible
|
| It's already technically feasible:
| https://www.primeintellect.ai/blog/intellect-2
| ryoshu wrote:
| People are doing it: https://nousresearch.com/nous-psyche/
| am17an wrote:
| Deepseek-R1 is at the level of GPT 4.1 already, it's open-
| weight, open-source and they even open-sourced their inference
| code.
| jjordan wrote:
| I don't know why everyone keeps echoing this, my experience
| with Deepseek-R1, from a coding perspective at least, has
| been underwhelming at best. Much better experience with GPT
| 4.1 (and even better with Claude, but that's a different
| price category).
| am17an wrote:
| I'm not arguing which model is better for your use-case.
| I'm saying in general as it's "powerful" as GPT 4.1 in a
| lot of benchmarks, and you can peak underneath the hood,
| even make it better for your said use-case
| seunosewa wrote:
| Do you mean V3? V3 is 4.1 level or above.
| iJohnDoe wrote:
| I got the impression that 03-mini or 03-mini-high were
| meant for coding? GPT 4.1 was meant for creative writing,
| not coding?
| conradev wrote:
| It's good at a lot of things: GPT-4.1
| scores 54.6% on SWE-bench Verified, improving by 21.4%abs
| over GPT-4o and 26.6%abs over GPT-4.5--making it a
| leading model for coding.
|
| https://openai.com/index/gpt-4-1/
| coliveira wrote:
| They are trained on these "benchmarks", that's why they
| score better.
| jorvi wrote:
| This is everyone with every model.
|
| People sang praise from the roof for Google's Gemini 2.5
| models, but in many things for me they can't even beat
| Deepseek V3.
| CamperBob2 wrote:
| What would be an example of 2.5 Pro failing against R1
| (which is what you'd actually want to compare it to)?
| jorvi wrote:
| R1 sometimes fails against V3 for me too, so its not a
| specific dig against Gemini.
|
| In terms of code and science, Gemini is way, way too
| verbose in its output, and because of that it ends up
| getting confused by itself and hurting the quality of
| longer windows.
|
| R1 does this too, but it poisons itself in the reasoning
| loop. You can see it during the streaming, literally
| criss-crossing its thoughts and thinking itself into
| loops before it finally arrives at an answer.
|
| On top of that, both R1 and Gemini Pro / Flash are
| mediocre at anything creative. I can accept that from R1,
| since it's mainly meant as more of a "hard sciences"
| model, but Gemini is meant to be an all-purpose model.
|
| If you pit Gemini, Deepseek R1 and Deepseek V3 against
| each other in a writing contest, V3 will blow both of
| them out of the water.
| CamperBob2 wrote:
| Agreed on the last point, V3 is terrifyingly good at
| narrative writing. And yes, R1 talks itself out of
| correct answers almost as often as it talks itself into
| them.
|
| But in general 2.5 Pro is an extremely strong model. It
| may lose out in some respects to o3-pro, but o3-pro is so
| much slower that its utility tends to be limited by my
| own attention span. I don't think either would have much
| to fear from V3, though, except possibly in the area of
| short fiction composition.
| Zambyte wrote:
| In my experience, all reasoning models feel (vibely) worse
| at structured output like code versus comparable non-
| reasoning models, but far better at knowledge-based
| answering.
| benreesman wrote:
| https://i.imgur.com/we5ggtS.jpeg
| mensetmanusman wrote:
| Yeah the sanctions will (not sarcastically) actually improve
| the world on a number of fronts. Increasing diversity of
| compute, forcing decentralization of manufacturing, etc. etc.
| Der_Einzige wrote:
| The sanctions will (not sarcastically) massively harm the
| world because Nvidia may no longer be a free money cheat
| code. I like having an easy economic strategy for
| investing...
| mensetmanusman wrote:
| The world doesn't have to optimize policy to increase the
| profits of a single American company.
| seydor wrote:
| also increase smuggling, theft, espionage, crime, sabotage.
|
| There are much better ways to increase diversity
| mensetmanusman wrote:
| China spends the normalized equivalent of America's defense
| spending on suppressing their own citizens.
|
| From a western standpoint, this is amazingly helpful
| because it's a form of Chinese self destruction and waste.
| brabel wrote:
| Why would the Chinese self destructing be "amazingly
| helpful " to the West? This sounds like spiteful vitriol.
| alexnewman wrote:
| I think vs spending it on defense spending which would
| further make us mil parity difficult.
| SR2Z wrote:
| The current government of China becoming more friendly to
| the rest of the world would help the West (and the rest
| of the world).
|
| This is not jingoism. The PRC has not been around very
| long and it's already conquered a neighbor, fought
| multiple wars for land, and continues to build up its
| military to subjugate Taiwan.
|
| Being replaced with a liberal democracy would DEFINITELY
| be an improvement, even if these tariffs won't do it.
| zeld4 wrote:
| I'm struggling to see which is worse, using AI to police
| their own people, or using AI to genocide in Middle East.
| avn2109 wrote:
| >> ... spends the normalized equivalent of America's
| defense spending...
|
| I'd be interested in seeing the numbers for that claim
| broken down if you can cite them. From napkin math it
| seems hard to make the budgets line up, unless we're
| doing a very large purchasing power parity adjustment?
| jjcc wrote:
| There exists such numbers/information circulated mainly
| inside Chinese (language) media/social media in form of
| "screenshot" but no links. Screenshot as a way of hiding
| source is a common format for this type of information
| because the links will disclose the media that spread the
| information. Then normal (Chinese) audience will know the
| credibility of the information. Give you an example,
| "epoch times" is a common source of such type of
| information. The nature of the media is well-known to
| Chinese audience.
|
| The real equivalence to US defense budget in term of size
| is actually the infrastructure construction budget. While
| both budgets boost the economy , infrastructure budget
| improves the life of local people. Now as the most cities
| in coast areas run out of project to build, the over
| capacity cultivated in early years is poured to other
| directions: rural areas, undeveloped provinces, and even
| overseas especially Africa and Latin America. It's
| amazing that China changes very fast year by year as I
| visited some rural areas.
|
| Ionically this behavior of infrastructure building sounds
| like Chinese MAGA to me: mind our own business, focus on
| improve ourselves instead of spread values to other
| countries.
| FilosofumRex wrote:
| PRESIDENT TRUMP: "You don't think we can. You don't think
| we do that to them? We do. So we do a lot of things."
| https://singjupost.com/transcript-maria-bartiromo-
| interviews...
| reactordev wrote:
| SETI@Home style peer2peer open GPU training network is
| something I'm looking into as well.
| coolspot wrote:
| Possible and has been done, but super-slow and inefficient
| resulting in long training times for small models. To keep
| compute occupied you need to pass gradients very fast.
| pk-protect-ai wrote:
| Do you mean this one?
|
| https://blog.lambdaclass.com/introducing-demo-decoupled-
| mome...
| reactordev wrote:
| This is what piqued my interest in the first place
| reactordev wrote:
| Yes but could you break it up into chunks of sets of
| gradients to compute? I know that compute needs the full
| chunk to compute a set. Again, things I'm exploring but
| ultimately no different than just having the full dataset
| on disk and just scaling out compute nodes in ro mode.
| SubiculumCode wrote:
| I suppose its exciting, but whether that is a good thing
| depends entirely on how much you think AI technologies pose
| existential threats to human survival. This may sound
| hyperbolic, but serious people are seriously thinking about
| this and are seriously afraid.
| foursilly wrote:
| Since the license ban the use and installation in EU, I would
| ask: It is possible to formulate a license that claims: "The
| restriction A is motivated to protect our ass but we will not
| directly or indirectly enforce it against you"?, Such kind of
| phrasing in the license could be categorized or called "isolating
| clause" but I don't know if judges could consider it a
| circumvention of the law.
|
| Edited several times, I should add: IANAL, but this sounds
| similar to meta releasing llama weights. I think that the spirit
| of the European law is to control concrete uses of AI and not a
| broad distribution of weights and architecture. So my question
| is: Does the EU AI act ban this distribution?, I think it
| provides more competition and options for Europeans.
|
| Edited: Thinking a little more, installing open weights could
| allow backdoors (in the form of a way to manipulate intelligent
| agents via specials prompts designated to control the system), so
| perhaps from a national security point of view some care should
| be taken (but I personally hate that). So another question: Is
| there a way to control if open weights can create back doors (via
| prompt injection)?, I recall a paper in which prompt by symbols
| like 0?,#2! could put the system in a state in which one can read
| information that the LLM is asked to hide (that is a well known
| attack available to those that know the weights).
|
| Another question: Is fine tuning or Lora a way to eliminate o
| amilliorate such prompt attacks?, is there any python library to
| defend against such attacks. Download - install - modify by fine
| tune or lora - now you are protected.
| seydor wrote:
| It's not up to Huawei to tell EU citizens what to do. In fact
| they did not need to add this restriction to their license at
| all. As EU citizens we shoud know the laws of the land and
| protect ourselves by avoiding using these models like the
| plague.
| Fluorescence wrote:
| IANAL but the EU legislation is very broad about what it
| covers e.g.
|
| "AI systems should fall within the scope of this Regulation
| even when they are neither placed on the market, nor put into
| service, nor used in the Union."
|
| I don't really understand the limits of it's scope e.g. the
| difference between making a system available vs. controlling
| how it's used is not clear to me. I don't think you can
| escape the regulation of high-risk uses by offering a
| "general purpose" AI with no controls on how it's used.
|
| In terms of the open-source nature - I can see it being
| treated like giving away any other regulated product e.g.
| medication, cars, safety equipment etc. The lack of cost
| won't transfer the liability from the supplier to the
| consumer.
|
| https://eur-lex.europa.eu/legal-
| content/EN/TXT/?uri=CELEX:52...
| foursilly wrote:
| Thanks for all that information, I agree with you that the
| EU legislation is very broad. In my opinion, this justifies
| or motivates the inclusion of the ban in the EU.
| hammock wrote:
| What happens if you ignore overly broad EU regulations?
| Does your home country observe the EU's violation of its
| sovereignty and throw you in a home country jail? Does
| Brussels throw you in an EU jail? What country hosts the
| EU jail?
|
| Not to sound too snarky (just a little snarky), I'm just
| curious how it all works.
| coliveira wrote:
| As always, EU and USA courts will act as if they had
| jurisdiction over the rest of the world and motivate bans
| and similar measures against other countries.
| bee_rider wrote:
| Realistically lots of multi-national companies have an EU
| presence, so concerns about "violating sovereignty" are
| sorta moot. Huawei probably wants to do business in the
| EU which requires following EU law.
|
| As an aside: in general a sovereign country can do
| whatever they want in their own territory, this includes
| the right of the country to bind itself to treaties. So
| in your hypothetical,
|
| > Does your home country observe the EU's violation of
| its sovereignty and throw you in a home country jail?
|
| This doesn't look like a violation of sovereignty to me;
| the non-EU has decided to enforce an EU law. Why? I don't
| know, maybe it makes business easier for the
| multinational companies of the non-EU country.
|
| Countries can also do things like apply secondary
| sanctions to an entity. So, again hypothetically, the EU
| doesn't need to be able to enforce a ruling against you.
| They can make you toxic to anyone who wants to do
| business in the EU.
| lawlessone wrote:
| >What happens if you ignore overly broad EU regulations?
|
| Not much
|
| It's usually up to the the member states to implement the
| EU regulation.
|
| As you can see from Hungary with Orban in recent years
| though the EU's response is slow and lacking.
|
| When there is consensus things can move a lot faster
| though.
| baobun wrote:
| You could lose government contracts, get sanctioned, get
| banned.
|
| Already happening to Huawei and presumably the EU market
| is significant to them.
| Mars008 wrote:
| Arrest of Telegram's cofounder in France is an example of
| EU laws and enforcement.
| ricardobeat wrote:
| Continuing that quote:
|
| > for example of an operator established in the Union that
| contracts certain services to an operator established
| outside the Union in relation to an activity to be
| performed by an AI system that would qualify as high-risk
| and whose effects impact natural persons located in the
| Union.
|
| > this Regulation should also apply to providers and users
| of AI systems that are established in a third country, to
| the extent the output produced by those systems is used in
| the Union
|
| Otherwise it seems to reach way beyond what it actually is.
|
| Explicitly prohibiting EU usage in the license is probably
| a move to reduce liability under the eyes of those "used in
| the Union" clauses.
| Fluorescence wrote:
| It would be a misreading to think that example implies a
| more limited scope. The passage as a whole is pretty
| clear why they are they so broad: in order to avoid
| circumvention. I can understand why - it seems to be both
| a necessary yet unacceptable way to write laws!
|
| The passage continues:
|
| "To prevent the circumvention of this Regulation and to
| ensure an effective protection of natural persons located
| in the Union, this Regulation should also apply to
| providers and users of AI systems that are established in
| a third country, to the extent the output produced by
| those systems is used in the Union."
|
| An AI would come under this regulation even it's just the
| outputs that are used in the EU. Interesting to think
| about what that could lead to.
| paganel wrote:
| > protect ourselves
|
| "Protect" ourselves against whom? I'm a EU citizen
| (unfortunately), and I'm fully on board with China against
| Brussels. Which is to say, don't try to speak for everyone in
| this God-forsaken so-called union.
| HPsquared wrote:
| For security, I'd always treat ANY LLM generated code as
| untrusted until reviewed.
| foursilly wrote:
| The problem is with AI agents when you give them some
| control.
| HPsquared wrote:
| Those should also be treated as untrusted.
| jandrese wrote:
| 100%, which makes me nervous when I hear stories about AIs
| that write and execute their own code. It's just asking for
| trouble.
| tartoran wrote:
| They'll blame it on AI from now on. This could have serious
| implications and further erosion of any responsibility tech
| companies have will probably accelerate.
| coliveira wrote:
| That's correct, independent of the source of the LLM.
| chaosharmonic wrote:
| > Since the license ban the use and installation in EU, I would
| ask: It is possible to formulate a license that claims: "The
| restriction A is motivated to protect our ass but we will not
| directly or indirectly enforce it against you"?, Such kind of
| phrasing in the license could be categorized or called
| "isolating clause" but I don't know if judges could consider it
| a circumvention of the law.
|
| Maybe not the exact thing you're talking about, but that
| description reminds me of the Alliance for Open Media -- their
| codec licenses are royalty-free, but the same terms revoke your
| usage rights if you sue anyone for the use of these formats.
| Havoc wrote:
| Strange that they'd ban EU via license but not US
| seydor wrote:
| They wanted to exclude EU reviewers for when publishing their
| paper to a journal. EU reviewers are notoriously obnoxious and
| often chosen to be reviewer #2.
| mzl wrote:
| Yes, as a reviewer of a paper I am assigned a number, and
| must by nominative determinism fulfill the expectations...
| 123yawaworht456 wrote:
| EU has a nebulous AI act that prohibits a myriad things.
| JSR_FDED wrote:
| Sanctions are at best a stopgap measure. Ideally they would buy
| enough time to shore up domestic capabilities.
|
| Instead, cutting research funding and discouraging foreign
| students/researchers from coming to the US means that there will
| be depleted US capability just when China finds its groove.
| Havoc wrote:
| Time to sell nvidia shares?
| nsoonhui wrote:
| I hope someone can enlighten me, as it's not immediately clear
| the significance of it.
|
| Does this mean that Huawei phone which has been hurt badly by
| sanction will now stand a fighting chance because of homegrown
| GPU?
|
| How good or bad these GPU compares to the SOTA GPU in the west?
|
| And does this mean that Huawei has the ability to crank out the
| GPU commercially?
| cenamus wrote:
| I think this video gives a decent overview on Huawei in general
|
| https://www.youtube.com/watch?v=NMFdYFmeiBA
| randomNumber7 wrote:
| From the hardware huawei can build competitve phones. It's just
| hard to justify buying a phone without the google appstore.
| tazjin wrote:
| No, not at all. Huawei is targeting markets like China
| (obviously) and Russia, where the Google App Store is
| irrelevant.
|
| I've been using phones without Google Play for years.
|
| My point is that YMMV based on where you are.
| qkhhly wrote:
| > Does this mean that Huawei phone which has been hurt badly by
| sanction will now stand a fighting chance because of homegrown
| GPU?
|
| oh, man, "stand a fighting chance"? huawei phone sales has
| already been back and surpassed apple in china.
|
| https://www.counterpointresearch.com/insight/china-smartphon...
| crowcroft wrote:
| If current LLMs hit a scaling wall and the game becomes about
| efficiency, I wonder if there's going to be space in the market
| for small models focussed on specific use cases.
|
| I use Gemini to extract structured data from images and the flash
| model is great at this. I wonder how much effort it would be to
| create a smaller model that would run on something like a NUC
| with an AMD APU that is good enough for that one use case.
|
| Or perhaps you end up with mini external GPU sticks that run use
| case specific models on them. Might not be much of a market for
| that, but could be pretty cool.
| hmottestad wrote:
| I was looking for one to use for named entity extraction and
| found this fine tune here: https://huggingface.co/dslim/bert-
| base-NER?utm_source=chatgp...
|
| Its only 108 million params.
| bjord wrote:
| throwback to that brief period where people would mine bitcoin
| (ineffectively) using ASICs in their USB ports
| crowcroft wrote:
| Yes, and people buying random GPUs for ether etc. I'm not a
| huge fan of what crypto has become but there was something
| exciting about hacking stuff together at home for it which is
| currently missing in AI IMO.
|
| Maybe it's not really missing and the APIs for LLMs are just
| too good and cheap to make homebrew stuff exciting.
| bjord wrote:
| no, I think you're right--there's definitely something
| missing right now
|
| but more likely it's going on and we're just not seeing it
|
| in general, though, I think once a certain amount of money
| is involved, people just start to get rabid and everything
| becomes a lot less fun
| lawlessone wrote:
| Maybe more accessible tools i think?
|
| It's possible to run models locally, fidget with temp etc
|
| Being able to change other things on the fly like
| identify weights most used for a prompt and just changing
| those to see what happens is much harder.
|
| I've tried both LLMS and image generators on my machine
| locally and while it's gotten in easier it's a long task
| just setting up. Especially if you run into driver
| issues.
| snickmy wrote:
| that's already the case, and it's called model distillation.
| You use LLMs to generate labels but then you use a dedicated
| smaller model (usually NN) to run at 1000x cheaper cost of
| inference.
| crowcroft wrote:
| I think beyond the technical aspect it's a product and
| packaging problem.
|
| All the effort is in productizing foundational models and
| apps built on top of them, but as that plateaus distilled
| models and new approaches will probably get more time in the
| sun. I'm hopeful that if this is the case we will see more
| weird stuff come available.
| Mars008 wrote:
| > I wonder if there's going to be space in the market for small
| models focused on specific use cases.
|
| just recent discussion on HN: "Small language models are the
| future of agentic AI"
|
| https://news.ycombinator.com/item?id=44430311
| MangoCoffee wrote:
| People on HN are so delusional it's funny.
|
| The semiconductor industry is always a key industry for China.
| They laid it out in Made in China 2025.
|
| Huawei designed and developed its own chip before any sanctions,
| and the Chinese government never stopped throwing money into the
| semiconductor industry.
|
| In the short term, money can go to Nvidia, but it won't be long
| before China creates its own "Nvidia" like BYD.
|
| The sad part is America elected Trump, and he's gutting American
| research and cutting out the CHIPS Act.
|
| One country is going for the long term while another country is
| short sighted
| Beijinger wrote:
| This is a few days old. https://www.techinasia.com/news/tencent-
| hunyuan-opensources-...
|
| Here is the GIT https://github.com/Tencent-Hunyuan/Hunyuan-A13B
| zeld4 wrote:
| The silicon valley should thank Huawei and Deepseek, as the only
| two reasons AI exists as an industry in US are (a) reduce labor
| cost (b) win over China.
|
| (a) alone sounds super negative to ordinary people, but with (b),
| justified by the existence and achievement of Huawei & Deepseek,
| the AI cause now sounds a legit jihad silicon valley is carrying.
| bgnn wrote:
| A close friend of mine is Chinese. He went back to China to join
| a HW start-up as a founding engineer 6 years ago. Then cane the
| sanctions. He said that was the best thing happened when I
| recently met him. Their company grew because Huawei and all the
| other Chinese manufacturers don't want to buy anything from a
| West-aligned country anymore. Nobody cares about the sanctions
| anymore apparently as they accepted it as a given, so their focus
| is self reliance.
| jjcc wrote:
| The impacts are different sector by sector. Those benefit the
| most are the small EDA software companies that barely survive
| before sanctions due to the huge technology gaps behind the
| large EDA companies like Synopsys. Now they have tones of new
| customers don't want to take risk of service interruption due
| to sanction.
|
| It is called hormesis.
___________________________________________________________________
(page generated 2025-07-02 23:00 UTC)