[HN Gopher] Framework for Artificial Intelligence Diffusion
       ___________________________________________________________________
        
       Framework for Artificial Intelligence Diffusion
        
       Author : chriskanan
       Score  : 145 points
       Date   : 2025-01-16 20:00 UTC (1 days ago)
        
 (HTM) web link (www.federalregister.gov)
 (TXT) w3m dump (www.federalregister.gov)
        
       | chriskanan wrote:
       | The most salient thing in the document is that it put export
       | controls on releasing the weights of models trained with 10^26
       | operations. While there may be some errors in my math, I think
       | that corresponds to training a model with over 70,000 H100s for a
       | month.
       | 
       | I personally think the regulation is misguided, as it assumes we
       | won't identify better algorithms/architectures. There is no
       | reason to assume that the level of compute leads to these
       | problems.
       | 
       | Moreover, given the emphasis on test-time compute nowadays and
       | that it seems like a lot of companies have hit a wall with
       | performance gains with trying to scale LLMs at train-time, I
       | especially think this regulation isn't especially meaningful.
        
         | HeatrayEnjoyer wrote:
         | We can't let perfect be the enemy of good, regulations can be
         | updated. Capping FLOPs is a decent starter reg.
        
           | reaperman wrote:
           | Counterpoint would be the $7.25 minimum wage. It _can_ be
           | updated, but politicians aren 't good at doing that. In both
           | cases (FLOPS and minimum wage), at least a lower bound for
           | inflation should be included:
           | 
           | Something like: 10^26 FLOPS * 1.5^n where n is the number of
           | years since the regulation was published.
        
             | Cyph0n wrote:
             | I don't see an issue here, because our legislators probably
             | care more about FLOPS than humans.
        
             | tivert wrote:
             | > Something like: 10^26 FLOPS * 1.5^n where n is the number
             | of years since the regulation was published.
             | 
             | Why would you want to automatically increase the cap
             | algorithmically like that?
             | 
             | The purpose of a regulation like this is _totally_
             | different than the minimum wage. If the point is to keep
             | and adversary behind, you want them to stay as far behind
             | as you can manage for as long as possible.
             | 
             | So if you increase the cap, you want only increase it when
             | it won't help the adversary (because they have
             | alternatives, for instance).
        
         | parsimo2010 wrote:
         | Traditional export control applied to advanced hardware is
         | because the US doesn't want its adversaries to have access to
         | things that erode the US military advantage. But most hardware
         | is only controlled at the high-end of the market. Once a
         | technology is commodotized, the low-end stuff is usually widely
         | proliferated. Night vision goggles are an example, only the
         | latest generation technology is controlled, and low-end stuff
         | can be bought online and shipped worldwide.
         | 
         | Applying this to your thoughts about AI, is that as the
         | efficiency of training gets better, the ability to train models
         | is commodotized, and those models would not be considered to be
         | advantageous and would not need to be controlled. So maybe
         | setting the export control based on the number of operations is
         | a good idea- it naturally allows efficiently trained models to
         | be exported since they wouldn't be hard to train in other
         | countries anyway.
         | 
         | As computing power scales maybe the 10^26 limit will need to be
         | revised, but setting the limit based on the scale of the
         | training is a good idea since it is actually measurable. You
         | couldn't realistically set the limit based on the capability of
         | the model since benchmarks seem become irrelevant every few
         | months due to contamination.
        
           | ein0p wrote:
           | I wonder what makes people believe that the US currently
           | enjoys any kind of a meaningful "military advantage" over
           | e.g. China? After failing to defeat the Taliban and running
           | from the Houthis especially. This seems like a very dangerous
           | belief to have. China has 4x the population and outproduces
           | us 10:1 in widgets (2:1 in dollars). Considering just e.g.
           | steel, China produces about 1 billion metric tons of it per
           | year. We produce 80 million tons. Concrete? 2.4B tons vs 96M
           | tons. 70+% of the world's electronics. Their shipbuilding
           | industry is 230x more productive (not a typo). Etc, etc.
           | 
           | The short term profits US businesses have been enjoying over
           | the past 25 years came at a staggering long term cost. The
           | sanctions won't even slow down the Chinese MIC, and in the
           | long run they will cause them to develop their own high end
           | silicon sector (obviating the need for our own worldwide).
           | They're already at 7nm, at a low yield. That is more than
           | sufficient for their MIC, including the AI chips used there,
           | currently and in the foreseeable future.
        
             | parsimo2010 wrote:
             | a) just because the government has policies that doesn't
             | mean they are 100% effective
             | 
             | b) export controls aren't expected to completely prevent a
             | country from gaining access to a technology, just make it
             | take longer and require more resources to achieve
             | 
             | You may also be misunderstanding how much money China will
             | spend to develop their semiconductor industry. Sure, they
             | will eventually catch up to the West, but the money they
             | spend along the way won't be spent on fighter jet,
             | missiles, and ships. It's still preferable (from the US
             | perspective) to having no export controls and China being
             | able to import semiconductor designs, manufacturing
             | hardware, and AI models trained using US resources. At
             | least this way China is a few months behind and will have
             | to spend a few billion Yuan to achieve it.
        
             | airstrike wrote:
             | Everyone also thought Russia had a strong military yet look
             | how that worked out
        
               | ein0p wrote:
               | NATO is currently losing a conventional proxy war against
               | it, that's how. Which only reinforces my point: there's
               | zero reason to believe that either proxy or direct
               | confrontation with China can be "won".
        
               | airstrike wrote:
               | NATO is not fighting that war. You're delusional. They
               | are supporting Ukraine without crossing the threshold of
               | actually fighting the war, as bullshit as it sounds.
        
               | ein0p wrote:
               | $250B and depleted arsenals say otherwise. It's a proxy
               | war. It's been a proxy war by design all along: https://w
               | ww.rand.org/content/dam/rand/pubs/research_briefs/R...
        
               | airstrike wrote:
               | Spending has nothing to do with military reality. Russia
               | is a paper tiger. If the US really supported Ukraine in
               | this war without worrying about crossing some imaginary
               | Rubicon, the war would have been over a long time ago--
               | except Russia might be tempted to use their nukes, so we
               | can't do that.
               | 
               | The way to win against Russia is not via sanctions but
               | rather via destabilizing the regime through guerrilla
               | propaganda. The Russians, the Chinese, and the Soviets
               | before them have always known this. The West is just too
               | slow to catch on.
        
         | etiam wrote:
         | Could be nice with some artificial pressure to use more
         | efficient algorithms though. The current game of just throwing
         | in more data centers and power plants may be kind of convenient
         | for those who can afford it, but it's also intellectually
         | embarrassing.
        
         | thorum wrote:
         | The practical problem I see is that unless US AI labs have
         | perfect security (against both cyber attacks and physical
         | espionage), which they don't, there is no way to prevent
         | foreign intelligence agencies from just stealing the weights
         | whenever they want.
        
           | iugtmkbdfil834 wrote:
           | This. We put toasters on the internet and are no longer
           | surprised, when services we use send us breach notices at
           | regular intervals. The only thing this regulation would do,
           | as written, is add an interesting choke point for compliance
           | regulators to obsess over.
        
           | kube-system wrote:
           | Of course. They're mitigations, not preventions. Few defenses
           | are truly preventative. The point is to make it difficult.
           | They know bad actors will try to circumvent it.
           | 
           | This isn't lost on the authors. It is explicitly recognized
           | in the document:
           | 
           | > The risk is even greater with AI model weights, which, once
           | exfiltrated by malicious actors, can be copied and sent
           | anywhere in the world instantaneously.
        
             | thorum wrote:
             | > The point is to make it difficult.
             | 
             | Does it, though?
        
         | permo-w wrote:
         | this is like saying that regulating automatic weapons is
         | misguided because someone might invent a gun that is equally
         | dangerous without being automatic
        
           | iugtmkbdfil834 wrote:
           | This appears to be a very shallow take and lazy argument that
           | does not capture even basic nuance of the issue at hand. For
           | the sake of expanding it a little and hopefully moving it in
           | the right direction, I will point out that BIS framework
           | discusses use of advanced models as dual use goods ( ie. not
           | automatically automatic weapons ).
           | 
           | edit(removed exasparated sigh; it does not add anything )
        
         | logicchains wrote:
         | >The most salient thing in the document is that it put export
         | controls on releasing the weights of models trained with 10^26
         | operations.
         | 
         | Does this affect open source? If so, it'll be absolutely
         | disastrous for the US in the longer term, as eventually China
         | will be able to train open weights models with more than that
         | many operations, and everyone using open weights models will
         | switch to Chinese models because they're not artificially
         | gimped like the US-aligned ones. China already has the best
         | open weights models currently available, and regulation like
         | this will just further their advantage.
        
           | gyre wrote:
           | "consistent with its general practice, BIS will not require a
           | license for the export of the model weights of open-weight
           | models"
        
       | resters wrote:
       | Strong opposition to this regulation seems to be one of the main
       | things that led a16z, Oracle, etc. to go all in for Donald Trump.
       | It's interesting that Meta too fought the regulation by its
       | unprecedented open sourcing of model weights.
       | 
       | Regardless of who is currently in the lead, China has its own
       | GPUs and a lot of very smart people figuring out algorithmic and
       | model design optimizations, so China will likely be in the lead
       | more obviously within 1-2 years, both in hardware and model
       | design.
       | 
       | This law is likely not going to be effective in its intended
       | purpose, and it will prevent peaceful collaboration between US
       | and Chinese firms, the kind that helps prevent war.
       | 
       | The US is moving toward a system where government controls and
       | throttles technology and picks winners. We should all fight to
       | stop this.
        
         | 9283409232 wrote:
         | I don't think you're wrong but Big Tech is bending the knee to
         | Trump because he will be picking the winners.
        
         | tokioyoyo wrote:
         | > The US is moving toward a system where government controls
         | and throttles technology and picks winners
         | 
         | What else can it do? They don't want to lose their lead, and
         | whatever restrictions they've been putting on China et al. have
         | let the exact desired outcomes so far. The idea is to try to
         | slow down the beast that has very set goals (e.g. to become
         | high tech manufacturing and innovation center), and try to play
         | catch up (like on-shoring some manufacturing).
         | 
         | Personally, I'm skeptical that it will work, because by raw
         | number of hands on deck, they have the advantage. And it's
         | fairly hard when your institutional knowledge of doing big
         | things is a bit outdated. I would argue, a good bet in North
         | America would be finding a financially engineered solution to
         | get Asian companies bring their workers and knowledge to ramp
         | us up. Kinda like the TSMC factory. Basically the same thing as
         | China did in 2000s with western companies.
        
           | logicchains wrote:
           | > They don't want to lose their lead, and whatever
           | restrictions they've been putting on China et al. have let
           | the exact desired outcomes so far.
           | 
           | They absolutely have not. The best open weights LLM is
           | Chinese (and it's competitive with the leading US closed
           | source ones), and around 10x cheaper both to train and to
           | serve than its western competitors. This innovation in
           | efficiency was largely brought about by US sanctions limiting
           | GPU availability.
        
           | buyucu wrote:
           | > They don't want to lose their lead
           | 
           | What lead? The best open-source language models right now are
           | Chinese. deep-seek is amazin, so is qwq.
        
         | blackeyeblitzar wrote:
         | I agree this law won't be effective in its intended purpose,
         | and that China will develop models of their own that are
         | sufficiently competitive (as we've already seen). However, I
         | think seeking "peaceful collaboration" between the US (or
         | Europe or many others) and China - either between governments
         | or private firms - is a naive strategy that will simply lead to
         | the US being replaced by a more dangerous superpower that does
         | not respect the values of free and democratic societies.
         | 
         | I also think that to a great extent, we're already at war.
         | China has not respected intellectual property rights, conducted
         | espionage against both companies and government agencies,
         | repeatedly performed successful cyberattacks, helped Russia in
         | the Ukraine conflict, severed telecommunications cables, and
         | more. They've also built up the world's largest navy, expanded
         | their nuclear arsenal, and are working on projects to undermine
         | the status of the US Dollar. All of this should have been met
         | with a much stronger and forceful reaction, since clearly it
         | does not fit into the notion of "peaceful collaboration".
         | 
         | China's unpeaceful actions aren't limited to the West. China
         | annexed much of its current territory illegally and through
         | force (see Xinjiang and Tibet). When Hong Kong was handed back,
         | it was under a treaty that China now says is not valid. China
         | has been trying to steal territory from neighboring countries
         | repeatedly, for example with Bhutan or India. They've also
         | threatened to take over Taiwan many times now, and may do so
         | soon. They're about to build a dam that will prevent water from
         | reaching Bangladesh and force them to become subjugated. The
         | only peaceful and just outcome is for those territories to be
         | freed from the control of China - which will require help from
         | the West (sanctions, tariffs, blockades, and maybe even direct
         | intervention).
         | 
         | Even _within_ China, the CCP rules with an iron fist and
         | violates virtually all principles of free societies and
         | classically liberal values that we value in the West. I don't
         | see that changing. And if it doesn't, how can they be trusted
         | with more economic and military power? That's why I don't think
         | we should seek peaceful collaboration with China. We just need
         | smarter strategies than this hasty AI declaration.
        
           | almaight wrote:
           | 1. Sikkim was annexed by India 2. Bangladesh has many floods
           | every year
        
             | blackeyeblitzar wrote:
             | I am not sure why Sikkim is relevant. I am not familiar
             | with it, so I read about it now. And it looks like India's
             | prime minister actually pushed a resolution through noting
             | that Sikkim is independent. But then Sikkim had a domestic
             | movement to join India and voluntarily did that. This seems
             | like the opposite of annexation.
             | 
             | Regarding floods - if Bangladesh wants an upstream dam, why
             | aren't they included as a decision maker on whether the
             | Chinese dam goes ahead? Clearly this is because they would
             | say no to it. The issue isn't floods - it's that China can
             | withhold water for drinking and irrigation and threaten the
             | country with starvation and famine. It's a huge national
             | security threat.
        
           | miohtama wrote:
           | China models are not only competitive, but better, because
           | they do not care about copyright violations and they do not
           | need to litigate with everyone. They can use better training
           | material faster.
        
           | buyucu wrote:
           | Where I'm looking from (Turkey) China is mostly harmless and
           | the USA is a foreign invader that has invaded multiple
           | neighboring countries. Between the two, I find the USA to be
           | much more illiberal and dangerous.
        
             | 9283409232 wrote:
             | You think China is harmless? Go tell that to Tibet, India,
             | or Taiwan.
        
               | buyucu wrote:
               | Less harmful than USA for sure. The list of places US and
               | allies have invaded is very long.
        
               | 9283409232 wrote:
               | In recent years China does not do much blatant invading
               | and prefer much more subtle means daring people to
               | respond to them like in the South China Sea.
        
               | buyucu wrote:
               | still not as bad as 25 years of wars in the Middle East.
               | 
               | I would take those subtle means over the blatant invading
               | that America has been doing.
        
               | 9283409232 wrote:
               | You do know both can be bad right? You don't need to pick
               | the lesser evil.
        
         | kube-system wrote:
         | > The US is moving toward a system where government controls
         | and throttles technology and picks winners.
         | 
         | Moving towards? The US has a pretty solid history of doing a
         | great deal of this (and more) in the 20th century. But so did
         | all of the world's powers... as they all continue to do today.
         | It seems to be an inherent part of being a world power.
        
       | neilv wrote:
       | This `regulations.gov` is leaking info on who accesses what, to
       | Google (via `www.google-analytics.com` tracker).
       | 
       | There should be a federal regulation about that.
        
         | datadrivenangel wrote:
         | It's for the government's google analytics account, which is
         | open data: https://analytics.usa.gov/general-services-
         | administration
        
           | dooglius wrote:
           | Parent is unhappy with the fact the data is being tracked and
           | analyzed at all, not about the closed or open nature of the
           | analytics.
        
             | neilv wrote:
             | I'm unhappy that: (1) Google is getting this information;
             | (2) because Google is involved, who reads what is
             | attributable to individuals; (3) citizens don't realize
             | this.
        
       | intunderflow wrote:
       | We're sorry, an error has occurred A general error occurred while
       | processing your request.
        
       | geuis wrote:
       | This smells a lot like the misguided crypto export laws in the
       | 90s that hampered browser security for years.
        
         | philjohn wrote:
         | And don't forget the amazing workaround Zimmerman of PGP fame
         | came up with - the source code in printed form was protected 1A
         | speech, so it was published, distributed, and then scanned and
         | OCR'd outside the US -
         | https://en.wikipedia.org/wiki/Pretty_Good_Privacy#Criminal_i...
        
           | mmaunder wrote:
           | And don't forget Thawte which ended up selling strong SSL
           | outside the US, cornering the international market thanks to
           | US restrictions, and getting bought by Verisign for $600M.
        
           | z2 wrote:
           | I for one would love to see model weights published in
           | hardcover book form.
        
             | dkga wrote:
             | What a throwback to the time when some edgy folks would
             | share printed codes in Pascal... I even remember seeing a
             | hard copy of a binary in hex which was best not to execute.
        
           | cheald wrote:
           | It might be somewhat prohibitive to print the model weights
           | for any sufficiently large model, though.
        
             | galangalalgol wrote:
             | Using 2d barcodes you can fit ~20MB per page. Front and
             | back you could probably fit a model that violated the rule
             | on less than a thousand pages.
             | 
             | Edit: maybe 10k pages
        
               | TeMPOraL wrote:
               | Skip the barcodes - just cast it to 24bit RGB data, add
               | in some error correction, and print as a high-DPI bitmap
               | (or a series of them, one per page).
        
               | galangalalgol wrote:
               | Printers that can keep all the color channels lined up
               | all the way down to their tightest dot pitch are few and
               | far between. That is why piql used black and white for
               | the boxing code they used to print out all of github. I'm
               | not sure why greyscale wasn't used.
        
               | TeMPOraL wrote:
               | > _Printers that can keep all the color channels lined up
               | all the way down to their tightest dot pitch are few and
               | far between._
               | 
               | It only takes one for a PoC. It only takes few to
               | establish a service. Spatial and color accuracy is going
               | to be a challenge, but I imagine there already exist
               | methods to correct for that - e.g. printing reference
               | patterns along with the data, so the scanner/digitizer
               | can calibrate to it on the fly.
        
               | layer8 wrote:
               | I'm doubtful that barcodes would qualify as free speech.
        
             | slt2021 wrote:
             | how about printing URL to download the weights file?
        
           | rangestransform wrote:
           | I hope this time we finally get a Supreme Court ruling that
           | export controls on code are unconstitutional, instead of the
           | feds chickening out like last time
        
           | tzs wrote:
           | I doubt that would work for model weights because they are
           | generated algorithmically rather than being written by
           | humans, which probably means that they are not speech.
        
             | mbil wrote:
             | Not to mention how much paper it would take to print 500B
             | weights
        
         | kube-system wrote:
         | It hampered the security of a lot of things. That wasn't
         | misguided -- that was the point.
         | 
         | China, Russia, and Iran used Internet Explorer too.
        
       | mlfreeman wrote:
       | What do the regulators writing this _intend_ for this to slow
       | down /stop?
       | 
       | I can't seem to find any information about that anywhere.
        
         | chatmasta wrote:
         | Maybe they intend for it to speed up/start implementation of
         | federal agencies and regulations. The intent is to exert
         | control over an emerging market while it's still comprised of
         | cooperative participants. Regulators want to define the
         | regulatory frameworks rather than relying on self-policing
         | before it's "too late."
         | 
         | Let's see if this survives the next administration. Normally
         | I'd be skeptical, but Musk has openly advocated about the
         | "dangers" of AI and will likely embrace attempts to regulate
         | it, especially since he's in a position to benefit from the
         | regulatory capture. In fact he's doubly well-placed to take
         | advantage of it. Regardless of his politics, xAI is a market
         | leader and would already be naturally predisposed to
         | participate in regulatory capture. But now he also enjoys
         | unprecedented influence over policymaking (Mar a Lago) and
         | regulatory reform (DOGE). It's hard to see how he wouldn't
         | capitalize on that position.
        
           | chronic4930018 wrote:
           | > Regardless of his politics, xAI is a market leader
           | 
           | Lol what?
           | 
           | The only people who think this are Elon fanboys.
           | 
           | I guess you think Tesla is the self-driving market leader,
           | too. Okay.
        
             | chatmasta wrote:
             | I don't even use it. But in terms of funding, it's in the
             | top 5, according to Crunchbase data [0].
             | 
             | [0] https://news.crunchbase.com/ai/startup-billion-dollar-
             | fundra...
        
               | cheonic516442 wrote:
               | > But in terms of funding, it's in the top 5.
               | 
               | WeWork had more funding than xAI.
               | 
               | Not saying xAI will die, but you can't look at funding.
        
               | chatmasta wrote:
               | You can look at funding to see who will benefit most from
               | regulatory capture, a phenomenon that raises the cost of
               | entry into a market. Obviously the most well-funded
               | participants will have the greatest motivation to
               | increase barrier to entry.
               | 
               | If anything, a well-funded company with a bad product is
               | _more_ likely to engage in regulatory capture because it
               | limits their risk of exposure to a new entrant with a
               | good product.
        
               | bilbo0s wrote:
               | In this particular market though it's a bit useless
               | because _only_ well funded entities are really able to
               | play here.
               | 
               | Where are any of us gonna get 10,000 H100's?
               | 
               | So it's not like you're overwhelming small players by
               | throwing up regulations. Those players are extremely
               | unlikely to get the compute they'd need to even
               | experiment in the first place.
        
               | miohtama wrote:
               | It's also quite decent, giving a lot of free request so I
               | have switched from Claude.
        
             | DSingularity wrote:
             | Hard to take comments like this seriously when you can't
             | even be bothered to be associated with it from your primary
             | account.
        
         | HeatrayEnjoyer wrote:
         | Obviously to prevent proliferation of dual-use technologies to
         | potentially adversarial actors. The same intent behind
         | restricting high-fidelity infrared camera and phased radar
         | equipment.
        
           | slt2021 wrote:
           | China is leading AI race with their open source deepseek-v3.
           | It is laughable to think they this regulation with stop them.
           | USA should actually collaborate, not isolate.
           | 
           | China engineers have capability to get around these silly
           | sanctions, by renting cloud GPUs from USA companies for
           | example, to get access to as many compute as they want. or to
           | use consumer grade compute, or their homegrown Chinese
           | CPUs/GPUs.
           | 
           | USA should actually embrace open source and collaborate
           | together, as we are still in the very beginning of AI
           | revolution
        
             | kube-system wrote:
             | The _entire point_ is to not collaborate, because this tech
             | is being used for military purposes. The US wants to throw
             | up roadblocks to make it more difficult. Obviously, against
             | a foreign military, anything is a mitigation and not a
             | prevention.
             | 
             | > China engineers have capability to get around these silly
             | sanctions, by renting cloud GPUs from USA companies for
             | example
             | 
             | That's why they're also moving towards KYC for cloud
             | providers.
             | 
             | https://www.federalregister.gov/documents/2024/01/29/2024-0
             | 1...
        
               | bilbo0s wrote:
               | Also when you look at the list of nations targeted it's
               | easy to see that this is not necessarily only about
               | China. I think on HN people get China blindness and maybe
               | have a hard time seeing other threats out there.
               | 
               | The list is controversial obviously, but I have to think
               | nations wouldn't be on any tier of the list except the
               | no-restrictions tier if there wasn't something our intel
               | people weren't worried about. Maybe the concerns are not
               | legitimate, but there's definitely a reason we don't want
               | those nations having access to SOTA AI models.
        
               | slt2021 wrote:
               | Yes, USA wants to ensure global domination of large big
               | tech in AI world so that openai/anthropic/etc did not
               | have competitors
        
       | chriskanan wrote:
       | I have no idea if comments actually have any impact, but here is
       | the comment I left on the document:
       | 
       | I am Christopher Kanan, a professor and AI researcher at the
       | University of Rochester with over 20 years of experience in
       | artificial intelligence and deep learning. Previously, I led AI
       | research and development at Paige, a medical AI company, where I
       | worked on FDA-regulated AI systems for medical imaging. Based on
       | this experience, I would like to provide feedback on the proposed
       | export control regulations regarding compute thresholds for AI
       | training, particularly models requiring 10^26 computational
       | operations.
       | 
       | The current regulation seems misguided for several reasons.
       | First, it assumes that scaling models automatically leads to
       | something dangerous. This is a flawed assumption, as simply
       | increasing model size and compute does not necessarily result in
       | harmful capabilities. Second, the 10^26 operations threshold
       | appears to be based on what may be required to train future large
       | language models using today's methods. However, future advances
       | in algorithms and architectures could significantly reduce the
       | computational demands for training such models. It is unlikely
       | that AI progress will remain tied to inefficient transformer-
       | based models trained on massive datasets. Lastly, many companies
       | trying to scale large language models beyond systems like GPT-4
       | have hit diminishing returns, shifting their focus to test-time
       | compute. This involves using more compute to "think" about
       | responses during inference rather than in model training, and the
       | regulation does not address this trend at all.
       | 
       | Even if future amendments try to address test-time compute, the
       | proposed regulation seems premature. There are too many unknowns
       | in future AI development to justify using a fixed compute-based
       | threshold as a reliable indicator of potential risk. Instead of
       | focusing on compute thresholds or model sizes, policymakers
       | should focus on regulating specific high-risk AI applications,
       | similar to how the FDA regulates AI software as a medical device.
       | This approach targets the actual use of AI systems rather than
       | their development, which is more aligned with addressing real-
       | world risks.
       | 
       | Without careful refinement, these rules risk stifling innovation,
       | especially for small companies and academic researchers, while
       | leaving important developments unregulated. I urge policymakers
       | to engage with industry and academic experts to refocus
       | regulations on specific applications rather than broadly
       | targeting compute usage. AI regulation must evolve with the field
       | to remain effective and balanced.
       | 
       | ---
       | 
       | Of course, I have no skin in the game since I barely have any
       | compute available to me as an academic, but the proposed rules on
       | compute just don't make any sense to me.
        
         | behnamoh wrote:
         | You wrote three separate comments for this course. Can you just
         | combine them?
        
           | p1necone wrote:
           | The three comments are communicating three separate things, I
           | think it's clearer that way.
        
         | ben_w wrote:
         | > First, it assumes that scaling models automatically leads to
         | something dangerous.
         | 
         | The impression I had is with reversed causation: that it can't
         | be all that dangerous if it's smaller than this.
         | 
         | Assuming this alternative interpretation is correct, the idea
         | may still be flawed, for the same reasons you say.
        
           | SubiculumCode wrote:
           | I also suspect that the only real leverage the U.S. has is on
           | big compute (i.e. requires the best chips), and less capable
           | chips are not as controllable.
        
         | tivert wrote:
         | > Even if future amendments try to address test-time compute,
         | the proposed regulation seems premature. There are too many
         | unknowns in future AI development to justify using a fixed
         | compute-based threshold as a reliable indicator of potential
         | risk.
         | 
         | I'm disinclined to let that be a barrier to regulation,
         | especially of the export-control variety. It seems like letting
         | the perfect be the enemy of the good: refusing to close the
         | barn door you have, because you think you might have a better
         | barn door in the future.
         | 
         | > Instead of focusing on compute thresholds or model sizes,
         | policymakers should focus on regulating specific high-risk AI
         | applications, similar to how the FDA regulates AI software as a
         | medical device. This approach targets the actual use of AI
         | systems rather than their development, which is more aligned
         | with addressing real-world risks.
         | 
         | How to you envision that working, specifically? Especially when
         | a lot of models are pretty general and not very application-
         | specific?
        
           | iugtmkbdfil834 wrote:
           | << It seems like letting the perfect be the enemy of the
           | good: refusing to close the barn door you have, because you
           | think you might have a better barn door in the future.
           | 
           | Am I missing something? I am not an expert in the field, but
           | from where I sit, there literally is no barn door at this
           | point to even close too late..
        
         | energy123 wrote:
         | "First, it assumes that scaling models automatically leads to
         | something dangerous"
         | 
         | The regulation doesn't exactly make this assumption. Not only
         | are large models stifled, the ability to serve models via API
         | to many users, and the ability to have many researchers working
         | in parallel on upgrading the model is also stifled. It
         | wholesale stifles AI progress for the targeted nations.
         | 
         | This is an appropriate restriction on what will likely be a
         | core part of military technology in the coming decade (eg drone
         | piloting).
         | 
         | Look, if Russia didn't invade Ukraine and China didn't keep
         | saying they wanted to invade Taiwan, I wouldn't have any issues
         | with sending them millions of Blackwell chips. But that's not
         | the world we live in. Unfortunately, this is the foreign policy
         | reality that exists outside of the tech bubble we live in. If
         | China ever wants to drop their ambitions over Taiwan then the
         | export restrictions should be dropped, but not a moment sooner.
        
           | babkayaga wrote:
           | right. China. but Switzerland? Israel? what is going on here?
        
             | jagrsw wrote:
             | It could be related to 14eyes with modifications (finland
             | and ireland, plus close asian allies).
             | 
             | https://res.cloudinary.com/dbulfrlrz/images/w_1024,h_661,c_
             | s... (from https://protonvpn.com/blog/5-eyes-global-
             | surveillance).
             | 
             | Israel, Poland, Portugal and Switzerland are also missing
             | from it
        
               | gunian wrote:
               | As someone who is in the verge of being killed with no
               | side in this entire reality it's cool that in addition to
               | trade and economics we now get compute as a geopolitical
               | indicator maybe it can really all be automated
        
             | energy123 wrote:
             | > Switzerland? Israel?
             | 
             | I hope someone with a better understanding of the details
             | can jump in, but they are both Tier 2 (not Tier 3)
             | restricted, so maybe there are some available loopholes or
             | Presidential override authority or something. Also I
             | believe they can still access uncapped compute if they go
             | via data centers built in the US.
        
             | thatcat wrote:
             | Israel is a known industrial espionage threat to the us,
             | how'd you think they got nuclear weapons? some analysts say
             | they're the largest threat after china. Not to mention
             | theyre currently using ai in targeting systems while under
             | investigation for war crimes.
        
               | ben_w wrote:
               | > how'd you think they got nuclear weapons?
               | 
               | I rather assumed they were able to re-invent them from
               | scratch by the work of their own scientists. I mean, the
               | US did it before the invention of the transistor and what
               | I've heard about the USSR project is their espionage only
               | helped them know the critical mass without needing so
               | many test runs, so it doesn't seem like it would be
               | implausible that Israel could do it themselves about 20
               | years later.
        
               | thatcat wrote:
               | Whether it is plausible or not is unrelated to what
               | happened.
               | 
               | In 1965 NUMEC owner, Zalman Shapiro - in coordination
               | with israeli intelligence, diverted 100 kg of 95%
               | enriched uranium from the facility and shipped it to
               | Israel. Enriching the material to weapons grade is the
               | technically difficult part - which I would think israel
               | would certainly not have the budget for given the size of
               | its economy.
        
               | ben_w wrote:
               | Thanks for the specifics.
               | 
               | > which I would think israel would certainly not have the
               | budget for given the size of its economy.
               | 
               | Hmm.
               | 
               | I'll buy that. I've seen a lot of wildly different cost
               | estimates for separation work units, so I can only guess
               | how much it might have cost at the time.
               | 
               | If it would have otherwise been the full Manhattan
               | Project, I don't even need to guess, you're definitely
               | correct they couldn't have afforded it:
               | https://www.wolframalpha.com/input?i=gdp+israel+1965
        
           | logicchains wrote:
           | Limiting US GPU exports to unaligned countries is completely
           | counterproductive as it creates a market in those countries
           | for Chinese GPUs, accelerating their development even faster.
           | Because a mediocre Huawei GPU is better than no GPU. And it
           | harms the revenue of US-aligned GPU companies, slowing their
           | development.
        
             | dr_dshiv wrote:
             | Interesting theory. Any evidence that this is how the world
             | really works? (And, is there a catchy name for the
             | phenomenon?)
        
               | jampekka wrote:
               | Import substitution. Before the free trade ideology it
               | was quite a popular and (usually) deliberate method for
               | industrialisation and technological progress.
               | 
               | https://en.m.wikipedia.org/wiki/Import_substitution_indus
               | tri...
        
               | tartoran wrote:
               | But also no external competition causes no pressure to
               | improve.
        
               | dr_dshiv wrote:
               | That's why the Huawei story seems different.
               | 
               | I think it's about this attempt to block that actually
               | makes the opponent stronger. The export restrictions were
               | designed to weaken their competitiveness not enhance!
               | 
               | The restrictions blocked asml and Samsung and others from
               | trading with China. Now China can just replace them. It
               | seems.
        
               | jampekka wrote:
               | Seemed to work for the space and arms races for example.
        
         | SubiculumCode wrote:
         | While your critiques most likely have some validity (and I am
         | not positioned to judge their validity), you failed to offer a
         | concrete policy alternative. The rules were undoubtedly made
         | with substantial engagement from the industry and academic
         | researchers, as there is too much at stake for them not to
         | engage, and vigorously. Likely there were no perfect policy
         | solutions, but decided to not let the perfect stop the good
         | enough since timeliness matters as much or more than the policy
         | specifics.
        
           | logicchains wrote:
           | Doing nothing is a better alternative, because these
           | restrictions will just encourage neutral countries to
           | purchase Chinese GPUs, because their access to US GPUs is
           | limited by these regulations. This will accelerate the growth
           | of Chinese GPU companies and slow the growth of US-aligned
           | ones; it's basically equivalent to the majority of nations in
           | the world placing sanctions on NVidia.
        
         | 15155 wrote:
         | > these rules risk stifling innovation
         | 
         | These rules _intentionally_ "stifle innovation" for foreigners
         | - this is a feature, not a bug.
        
       | chriskanan wrote:
       | I'm not sure why the link no longer works, but this one works.
       | The link should be updated to this one:
       | https://www.federalregister.gov/documents/2025/01/15/2025-00...
        
         | cshimmin wrote:
         | @dang
        
           | layer8 wrote:
           | "@dang" doesn't do anything. You need to email
           | hn@ycombinator.com.
        
         | dang wrote:
         | Fixed. Thanks!
        
       | cube2222 wrote:
       | It's worth noting that this splits countries into three levels -
       | first without restrictions, second with medium restrictions,
       | third with harsh restrictions.
       | 
       | And the second level, for some reason, includes (among others) a
       | bunch of countries that would normally be seen as close US allies
       | - e.g. some NATO countries (most of Central/Eastern Europe).
        
         | inglor_cz wrote:
         | _includes (among others) a bunch of countries that would
         | normally be seen as close US allies - e.g. some NATO countries
         | (most of Central /Eastern Europe)._
         | 
         | Yeah, this is really a bit insulting.
        
           | tivert wrote:
           | >> And the second level, for some reason, includes (among
           | others) a bunch of countries that would normally be seen as
           | close US allies - e.g. some NATO countries (most of
           | Central/Eastern Europe).
           | 
           | > Yeah, this is really a bit insulting.
           | 
           | So you're insulted some country of other wasn't included in:
           | 
           | > First, this rule creates an exception in new SS 740.27 for
           | all transactions involving certain types of end users in
           | certain low-risk destinations. Specifically, these are
           | destinations in which: (1) the government has implemented
           | measures to prevent diversion of advanced technologies, and
           | (2) there is an ecosystem that will enable and encourage
           | firms to use advanced AI models to advance the common
           | national security and foreign policy interests of the United
           | States and its allies and partners.
           | 
           | ?
           | 
           | IMHO, it's silly to get insulted over something like that.
           | Your feelings are not a priority for an export control law.
           | 
           | Taiwan, even though it's a US ally, is only allowed limited
           | access to certain sensitive US technology it deploys (IIRC,
           | something about Patriot Missile seeker heads, for instance),
           | because their military is full of PRC spies (e.g.
           | https://www.smh.com.au/technology/chinese-spies-target-
           | taiwa...), so if they had access the system would likely be
           | compromised. It's as simple as that.
        
         | surfingdino wrote:
         | Someone still sees Eastern Europe as a provider of cheap
         | brainpower. This is insulting.
        
           | Chance-Device wrote:
           | But is it untrue?
        
           | 15155 wrote:
           | The value of a service is whatever someone is willing to pay
           | for it - and conversely, the price someone is willing to
           | render the service for.
           | 
           | These folks aren't "forced" to provide "cheap brainpower:"
           | they are offering services at their market rate.
        
         | pjmlp wrote:
         | I see the return of cold war computing model, where many
         | countries had their own computer platforms and programming
         | languages.
         | 
         | Which apparently might be a good outcome to FOSS operating
         | systems, with national distributions like Kylin.
         | 
         | As European I vote for SuSE.
        
         | consumer451 wrote:
         | This smells about as well informed as the genius move that
         | forced Ukrainian owner, Max Polyakov, to divest from Firefly
         | Aerospace. A US government position that was widely derided by
         | space industry watchers, and has now been reversed.
         | 
         | This might be a product of the USA being a gerontocracy.
        
       | wslh wrote:
       | This feels like deja vu from the crypto wars (1990s). If that
       | experience helps, it is impossible to repress knowledge without
       | violence, and it motivates more people to hack the system. Good
       | times ahead "PGP released its source code as a book to get around
       | US export law" <https://news.ycombinator.com/item?id=7885238>
        
         | Hizonner wrote:
         | Not the same situation at all. PGP would run on any computer
         | you happened to have around. The source code was small enough
         | to fit in a book. The people who already had the code _wanted_
         | to release it. Lots of people could have rewritten it
         | relatively quickly.
         | 
         | The ML stuff they're worried about takes a giant data center to
         | train and an unusually beefy computer even to run. The weights
         | are enormous and the training data even more enormous. Most of
         | the people who have the models, especially the leading ones,
         | treat them as trade secrets and also try to claim copyright in
         | them. You can only recreate them if you have millions to spend
         | and the aforementioned big data center.
        
           | wslh wrote:
           | > The ML stuff they're worried about takes a giant data
           | center to train and an unusually beefy computer even to run.
           | 
           | Now, consider this: the Palm [1] couldn't even create an RSA
           | [2] public/private key pair in "user time" The pace of
           | technological advancement is astonishing, and new techniques
           | continually emerge to overcome current limitations. For
           | example, in 1980, Intel was selling mathematical coprocessors
           | [3] that were cutting-edge at the time but would be laughable
           | today. It's reasonable to expect that the field of machine
           | learning will follow a similar trajectory, making what seems
           | computationally impractical now far more accessible in the
           | future.
           | 
           | [1] https://en.wikipedia.org/wiki/Palm_(PDA)
           | 
           | [2] https://en.wikipedia.org/wiki/RSA_(cryptosystem)
           | 
           | [3] https://es.wikipedia.org/wiki/Intel_8087
        
       | clhodapp wrote:
       | One interesting geopolitical fact about this document that's not
       | being discussed much is the way it includes Taiwan in lists of
       | "countries".
       | 
       | Usually, the US government tries not to do that.
        
         | kube-system wrote:
         | Only in some diplomatic contexts.
         | 
         | https://www.cia.gov/the-world-factbook/countries/taiwan/
        
       | veggieroll wrote:
       | The compute limit is dead on arrival, because models are becoming
       | more capable with less training anyways. (See DeepSeek, Phi-4)
        
       | pjmlp wrote:
       | It is going to be like it was in the 1990s with PGP and such all
       | over again.
        
       | ChrisArchitect wrote:
       | Related:
       | 
       |  _WH Executive Order Affecting Chips and AI Models_
       | 
       | https://news.ycombinator.com/item?id=42683251
        
       | United857 wrote:
       | It seems some EU countries are unrestricted but others are. How
       | is this compatible with the EU single market/customs union?
        
       | djoldman wrote:
       | More information here:
       | 
       | https://www.federalregister.gov/documents/2025/01/15/2025-00...
        
       | PostOnce wrote:
       | Looks down to me.
       | 
       | https://archive.is/F78tf
        
       | miovoid wrote:
       | Perhaps this regulation will be a major force for next gen
       | symbolic AI systems.
        
       | saberience wrote:
       | What's the point in this? Isn't Trump going to just cancel this
       | immediately on Monday?
       | 
       | I don't see how we can assume it will be enacted at all.
        
       | casebash wrote:
       | Most of the comments here only make sense under a model where AI
       | isn't going to become extremely powerful AI in the near term.
       | 
       | If you think upcoming models aren't going to be very powerful,
       | then you'll probably endorse business-as-usual policies such as
       | rejecting any policy that isn't perfect or insisting on a high
       | bar of evidence before regulating.
       | 
       | On the other hand, if you have a world model where AI is going to
       | provide malicious actors with extremely powerful and dangerous
       | technologies within the next few years, then instead of being
       | radical, proposal like this start appearing extremely timid.
        
       | 1vuio0pswjnm7 wrote:
       | Unblocked:
       | 
       | https://web.archive.org/web/20250117042813if_/https://www.fe...
        
       ___________________________________________________________________
       (page generated 2025-01-17 23:01 UTC)