There is a shadow hanging over this Fable thing(12gramsofcarbon.com) |
There is a shadow hanging over this Fable thing(12gramsofcarbon.com) |
Why would the makers of open models (mostly Chinese firms) continue to open them up, now that the value chain and economy shifts? Previously, it was a (Chinese) national goal to force the market to compress OpenAI/Anthropic margins (and compressing their revenue along the way), to ensure the Chinese had access to high quality models, and could afford to compete. Now there is an opportunity to usurp and be the international default, and claim the margin for themselves by closing their models.
Beyond that, there is likely an upper bound of capability-per-parameter, which means that there is an upper bound on "local" models, and once you need the cloud, why would the government not target clouds next?
Edit: if anthropic couldn’t resolve this matter, they can do something reallllly funny right now and open source it to the public :)
They are not wrong, it feels like that Game of Thrones season where someone thought it would be a great idea to let the fanatics re-arm.
> The real story here is that this may be the beginning of governments restricting the availability of strong LLMs to the public, to you. Fable was the strongest model on the market, and the US government has told you you can’t use it (technically, only if you’re not a US citizen, but in practice, even if you are). If you think the solution here is going to be open source Chinese models and / or running on your own hardware, think again. Do you think China is going to allow the strongest LLMs from companies within its borders to be open source a year from now when they have Mythos capabilities, if the US government is keeping the strongest American models back? Unlikely. These are heading in the direction of being powerful cybersecurity weapons and it will be in the interest of nation states to restrict and control them. In 2 years time, I would be surprised if the strongest LLMs are available for general use at all.
The world is a bit bigger than US and China, if Anthropic did it, another company can do it as well.
I am highly skeptical about Mythos's part in the whole cyber security angle and Anthropic seems to agree with me:
> We have reviewed a report that we believe is the basis of the government's directive and validated that the level of capability displayed there is widely available from other models (including OpenAI’s GPT-5.5)
It does sound funny to hear this from Anthropic after they spent recent months with scaremongering about Mythos's capabilities, now they say it was a prank bro, you can actually achieve more or less the same with good old GPT-5.5.
> Will we be the poorer for that, or will we be safer? I think poorer, because I hate being told what technology I can and can’t use, but I’m not certain. Maybe you think the government should restrict strong LLMs. Maybe you don’t. But either way, this is big news and a rubicon has been crossed and a precedent set. That’s true even if the motivation for this is just the government settling scores with Anthropic.
What this has demonstrated: if you can't run the software on your own hardware, you should assume that it can be taken away at any moment.
You can’t just buy H100 there are government limits on that.
RTX4090 maybe has no government limits but NVidia is definitely limiting bulk orders per retailer. I guess if you buy a lot from each retailer you will most likely get flagged in one way or the other.
With respect to AI capabilities is it really?
I don't see anyone else producing frontier closed source LLM's or frontier open source LLM's outside of US and China.
If AI lived up to a tenth of the promises the American labs produce, the world would be drastically different today. It's not. I'm doubtful of future impact based on that.
I'm happy we can utilise current OSS models to the extent we can now. They'll improve. The world will continue as usual. And hopefully we can put this bubble behind us.
That part is kind of their point - it doesn't have the distribution issues your other examples have.
Gangnam Style is the most popular video ever, surely it means something right?!
If we're cooked, it's only because of a mass hysteria behind this thing. It's an extremely useful technology, we're just losing our collective mind because of it.
The Mythos marketing strategy in action
So much smoke, mirrors and SV techbro bullshit going around that it has become impossible to figure out what's what.
It's good that they exist, and I hope they catch up, but if you don't have origin constraints for your use case I don't see why you would chose their models today.
[0]: On the only benchmark they both published performance results - SWE-bench Verified -they are within a margin of error Mistral 77.6 vs Qwen 77.2.
What I am trying to get at is that the frontier is great, but you can be fine with less as well.
Not for the EU. Given the political importance of LLMs and the talent pool in France (let alone rest of the EU), I fully expect them to catch up.
I'm hearing a lot of this kind of thing. "Oh if only it was a different government". I'm sorry, but when you cry out for government involvement, it's not always going to be coming from the government you personally wanted. This is the whole problem with government involvement! I don't think that message is getting through, but it's the real lesson that should be learned here.
You can be critical of the policy of export controls and the meaning of them in modern day but saying it is a problem with the government in general is nonsensical.
Hardly non-sensical. You just have a different default.
Government is fine if it’s done by people that aren’t in it only for themselves. Actually it is often still fine if it’s done by people only in it for themselves. It’s just that the people in it right now will burn the world down to enrich themselves.
Academic studies consistently show that people attracted to a career in politics (regardless of affiliation) score higher on "Dark Triad" personality traits (narcissism, Machiavellianism, and psychopathy).
And even for the non-sociopaths, the problem with humans is instinctively we ARE only in it for ourselves and our family. Anyone claiming otherwise, ironically, is probably attempting to signal virtue for...personal gain...yet again.
Hell as much as the drone strikes get simplified down to "Obama killed people without trials" the main complaint at the time was that he was acting without Congressional approval.
Democrats shouldn't have responded to Congress getting blocked up by Republicans realizing that they could make "ineffective government" a self fulfilling prophecy but pretending everyone is okay with it isn't accurate either.
One party is rounding up people and putting them into concentration camps while doing a mass deportation. That same party is trying to end birthright citizenship. That same party set the world order ablaze with a completely pointless tariffs regime. That same party started a war in Iran to please their donors and the Israeli PM, a war that is going to (IMHO) go down as the biggest strategic blunder in US history. One party doesn't want half the population to have bodily autonomy. In fact some of them have openly said they want to hand out the death penalty for getting an abortion. One party has doubled the national debt in a decade to hand out massive tax cuts to the ultra-wealthy and also gutting essential services. One party has a president credibly implicated with Jeffrey Epstein. That entire party bar a handful of individuals (who have been punished for their "disloyalty") have gone to great lengths to hide the evidence of that malfeasance. One party is killing people essentially to manipulate the market with repeated lies about an "imminent deal". One party is wholesale engaged in voter suppression and election rigging.
It's the same party for all of these things. What the other party is guilty of is being complicit in all of the above by refusing to oppose it. Still bad but nowhere near the same.
Issues with majority support never change in almost all of the biggest democracies in the world right now.
For the US specifically its a representative oligopoly with Madisonian gridlock and a few million non-elected bureaucrats thrown in the middle.
The US gives the smallest amount of say to people to pick either Coke or Pepsi. Don't like sugary soda and think its making you fat? Tough luck, you gotta pick Coke or Pepsi.
I've pointed this out so many times when enforced DEI and cancellation culture was rife, and was asked, basically, why I wanted to be a nazi :-/
Like, in the US, Trump might do the ICE show for his voters, but Obama's deportated 3 million just fine in his time.
It's perfectly reasonable to want government involvement that for sane governments is OK, even when you don't like that government. The current US government is a completely insane outlier. You cannot expect everything to adjust for the most insane outlier.
Government intervention is good and useful and keeps a markets free and society fair, preventing things like monopolies, robber barons and insider trading.
When those constraints are removed, when government becomes the source of corruption, we end up wheee the US has arrived today - where companies that bribe government officials get preferred treatment and the law is used selectively as a weapon.
This is a very dangerous moment for the US.
Do you think lobbying did not exist prior to two years ago?
HN perceives America as a temporarily embarrassed Libertarian state.
Crony capitalism, media eco-chambers and inequality have fermented an unshakeable disbelief in positive government intervention, thus the only thing left politically tenable is flagrant corruption (drain the swamp).
I'm vey grateful for the Australian federal government; their actions have steered us to a much better 2025/2026 than many other countrires.
We've seen how social media sites have always been in contention with govs regarding free speech even tho theyre fundamentally a way for other people to socialize with one another
But at this time Dario was at OpenAI and was a co-author on the GPT-2 research paper announcing the model.
The "too dangerous to release" approach has been him the whole time, at both companies.
I have been lurking on the aigamedev subreddit to see exactly what sort of games people are coming up with and I can say I have been incredibly disappointing. I've been faithfully trying the games people post and have come to the conclusion that game design is a very difficult art to learn, and something LLMs really can't help with that much. My guess is that these games are "fun" just like toddler paintings are "beautiful." And there are so many quality indie games you could get for the 25+ dollars you'd spend generating the code. Anyways, I guess that's another discussion for another blog post.
This is both absolutely key, and also irrelevant. 'Security' is clearly a pretense, as otherwise the demand would not have been restricted to 'foreign nationals'. It is not like any US administration every trusted every 'US national'.
But the reason for the restriction is basically irrelevant. The fact that it happened, should be the final wake up call for the EU to take 'Digital Sovereignty' serious. Not just in 'talk', but with actual commitments in budgets and effort.
The American Government has weaponised state power in a clumsy, corrupt and punitive attack against Anthropic, in an escalating war over control of AI.
Meanwhile, HN has anchored on "marketing hype" as the only possible explanation - all evidence is contorted to fit into this increasingly contrived explanation. Object level analysis is disregarded in favor of dunking on Anthropic.
AI is a threat to your job, status, beliefs, and way of life. For HN, believing this truth is harder than coming up with rationalisations for why it MUST be untrue.
I appreciate the grounded few on HN who continue to engage with object level analysis, and accept that the world is about to change in a pretty bizarre way.
> We optimize what we can measure, not what we actually want to achieve. We hope and pray that these are the same thing, but they often aren’t.
He points out the core problem with LLMs. I believe it is impossible (or extremely expensive) to ensure that the models are aligned safely for everyone and any intention. And 'safe' can mean different things for a different audience.
OpenAI seems generally less dogmatic and more practically oriented. There's really nothing particularly good about them, but you can at least predict how a normal company will act.
(I noted the same thing a few weeks back, https://news.ycombinator.com/item?id=48341224 but his recent blogposts should make it crystal clear if there was any lingering doubt).
Foreign labs releasing open source models won't be able to comply, and as a result open source models will remain stunted at pre-mythos levels or their use will be criminalized.
We should look past the petty fights these closed labs have, and see their common interest in banning open source and/or local models.
We've seen this movie before with crypto export bans in the 90s. The rest of the world caught up and then surpassed the US very quickly - and that was without the enormous financial incentives of AI.
Circumstantial, but... timing is odd.
It got me wondering if this means all big models are US-only now? Are they gonna do the same with GPT-5.6, etc? Seems pretty unlikely to me. So I expect Fable to come back pretty soon.
Profitability rests on subscribers. Never subscribe!
I think it is either a missile directed at anthropic, as retribution for not giving the DoD what it wants, in which case it is likely to resolve pretty quickly. Or it is a shift of policy toward export restrictions on powerful LLM and then every model will be impacted as they reach the threshold. In which case this could have massive implications of revenues, valuations, and the whole datacenter buildout. And frankly on the location of the white collar workforce if it is indeed a productivity multiplier, all countries reciprocate, and not all countries can match the US LLMs.
And why would the EU allow exports of chip manufacturing equipment if the US then restricts the export of derivatives of those chips to the EU?
If a volatile administration can ban you from running code that you wrote -- without any democratic processes like a law or lawsuit -- why would you build anything in the US?
Every non-American company is now at a disadvantage against American companies. The implications can not be overstated.
ChatGPT was released 4 years ago and still out of 27 countries in EU, only Mistral based in France has a model closer to a frontiers and IMO EU has already lost the race and still trying to catch up to yesterday models.
Means you pay full price per token right? (Which I think works out to roughly 10x more than using Claude Code?)
Actually, for enterprise I think it doesn't make a difference anymore, since they switched to per-token billing.
There are more and more posts coming up recently about AI being problematic. But people use it. It's strange. It's like hitting yourself with a hammer on the head, wondering why that hurts but you keep on doing it.
> Speaking of the HN/Reddit folks, lots of people are gleefully cackling about how Anthropic got what they deserved for their ‘marketing stunt’ with Mythos. As I’ve said before, this isn’t the first time we’ve had an AI CEO argue that something is ‘unsafe’ for personal gain.
Do you not think it is time to give up the whole "it is hype" rhetoric and come to reality -- the models can actually be unsafe and naturally Fable is closed off and the government is pulling access.
I mean, yeah. But did it take this long for that to be apparent to you?
Wasn't that when Dario, et al were at the company. One way to view this is that OpenAI expelled the cultists and they went on to form their own organization that continued using the same tactics.
Certainly some of the Anthropic press around Fable seems to me to be just marketing but I also think there's a core of people there who really believe it. I also think like all good advertising/lies there's some truth to the claims even if they're exaggerating.
I don't think anybody who has actually lived under a pre-democratic regime would call the US "no way democratic". There are many democratic aspects of the US, and it has reasonably strong institutions. But it seems that most Americans have not yet realized what category they're in, and think that the US is some kind of front-runner.
Source: https://en.wikipedia.org/wiki/The_Economist_Democracy_Index
After all the federal budget is so large because you can swap states but you can't get away from the IRS.
On the other hand I don't know a solve for every bill having less than a handful of votes that are bipartisan...
You might shrug and say "well pobody's nerfect", but the disparity between the American narrative and the reality is actually quite extreme.
Source: https://en.wikipedia.org/wiki/The_Economist_Democracy_Index
From the original 2019 release:
> We can also imagine the application of these models for malicious purposes , including the following (or other applications we can’t yet anticipate):
Generate misleading news articles
Impersonate others online
Automate the production of abusive or faked content to post on social media
Automate the production of spam/phishing content
These worries are why they stated they were cautious in rolling it out
Important safety features ("do not generate child porn", "statements should be factual or backed by evidence") were simply not part of the design of these systems and have yet to truly solved to this day, but AI companies decided to release these technologies onto the general public regardless of their glaring flaws.
I like AI for its shitposting capabilities and its neat parlour tricks, but I also believe so far it has been a net negative for everyone but the richest minority of society who benefits from firing people and having computers do half their jobs badly. It's too late now, but in hindsight I do agree that these systems were too dangerous to release in this shape.
Kind of taking this a step further, what is the worst thing that an AI run amok could do to existentially threaten the human race? I feel like almost anything could be short-circuited by some form of "pulling the plug" before it got too far. But, hypothetically, if it were possible to launch nukes without human intervention, or with maybe a small amount that could be socially engineered, that seems plausible.
So, what if, along the lines of MAD doctrine and the plot of Battlestar Galactica, the best thing we could do for AI safety is just to engineer our other systems so that a hypothetical superhuman adversary could not use them against us? (This doesn't really solve AI child porn and fake news but those things are mostly just imaginative reflections of the people using them and you can't really fix that)
I've done a lot of programming on various sub sections of the disciple and it still remains to me the hardest one to crack for AI.
It's undoubtedly am incredible tool for accelerating output but I think it's going to be the hardest for ai to commoditize as a whole.
There's a lot to getting it right, and like all software, you have to built it for your target market. There's no easy AI solution to getting a fun and engaging core loop. Nor is there one for building the right level of complexity and balancing the learning curve.
I think a lot of people who can't/don't code see themselves as game designers and had thought that AI would let them make games, and are now finding it wasn't really about the code after all. That, and if you can't code, vibe coding alone isn't really good enough for much beyond flash-level games (yet).
I’m not sure this can be assumed. Discovery was already one of the biggest hurdles when releases were bottlenecked by human output. Increasing output 10x is only going to make it worse.
Same as with Google, where they’ve lost the SEO war against AI spammers and valuable content has become close to impossible to find.
These communites established a generation of modern animators and game developers. Maybe we'll see the same from the youth of today who use these tools and create communities around it.
It is always the creative world building part.
The main criticism of the Harry Potter books are not spelling or sentence structure, it is the plot holes and contradictions in the world build.
The same holds for software.
There simply won't be jobs for them.
The risk is that all of these very incredibly smart and disgruntled people decide to do something about it. Elite overproduction, but instead as a result of enormous shift in supply side economics.
I don't worry about losing my job. I worry about becoming useless. If you know what I am saying..
What’s so funny is that same people are the ones that identify themselves as liberals as long as they can keep their privileged, highly paid jobs.
- Anthropic is just doing this for marketing stunt
- AI is like NFT's
- circular deals
- the bubble will burst anytime soon
- the hype bro's are propping up the stock market so that they can exit quick like grifters
(I just made the last one up to force terminology they use)
This is really distracting because the main problem here is that AI is getting too powerful to be just handed out to normal people like us. If you still believe it is all hype, you are getting distracted from the real problem.
I'm guessing at some point this kind of rhetoric will die away and we focus on real problem
We need a Second Amendment for AI: the right to keep and bear strong AI shall not be infringed. This safety handwringing is going to solidify the state's monopolies over its subjec... err citizens.
The state is the farmer, and we are the cows.
You think this is a coincidence that it's happening shortly before Anthropic IPOs?
How many people in the US government (at senior levels) are currently on track to profiting massively from a huge Anthropic IPO? The answer is, most of them. Most of the most powerful CEOs, senators, congressmen, Trump's retinue, are invested in Anthropic through on vehicle or another.
I use AI all the time and Opus 4.8 can't even get the most basic shit right about a very popular videogame released a few years ago. It's not going to steal your baby and eat your wife.
You sound like you have AI psychosis honestly.
the first one to do IPO will win big. With the government pressing Anthropic, OpenAI IPO will vacuum up the funds that otherwise would have went into Anthropic IPO as OpenAI was falling behind.
>and the whole datacenter buildout.
somebody just did a $2T IPO with the idea of datacenters in space. One can wonder what laws/jurisdiction those datacenters will be subject to.
So repubblicans have not been about small government for a long time and Trump is not even a pure-blood repubblican so it was to be expected that he would do the thing that repubblicans have not been about for a long time...what in the circular reasoning? Oh and please name one repubblican president who successfully reduced spending or "made government smaller"
That's the constitutional bedrock of our societies. That doesn't mean it's always true but if you denounce that as a legitimate and achievable goal then you don't have a society anymore.
No government at all can just imply anarchism, ie. a kind of self-governing that doesn't ascribe to our conventional ideas about government, presidents, pms, members of parliament, senators, etc.
Sure you do. You just don't have a society that looks like ours does. And that doesn't necessarily mean monarchy or fascism or chaos as the only alternative.
The society you do get, might still even have a government too! Thinking government is not beneficial doesn't mean you dispense altogether with one. It can mean you have very difference tolerance and guardrails for it, as opposed to when defaulting to "government is beneficial".
What's more "government is not beneficial" might not even mean "any and all government is not beneficial". It might mean government of the type that's the "constitutional bedrock of our societies", and the mockery they call "democratic rule" is not.
Mythos found 1000 zero days in a few weeks - if I had asked your thoughts on this a few years ago, I'm sure it would've been "that is a super-weapon".
Plus, scaling laws are impossible to deny: More compute = more intelligence.
AI is going to completely redefine the role of human cognitive ability - if you think this is about "state monopoly", you're really thinking too small.
What’s your answer to it? There are other people who have thought of it and it’s not that simple.
just look at what's going on in Ukraine right now
Please remind us when Democrats have "taken away guns", and while you're at it when were those small arms last used to fight back against a tyrannical government?
Sure I can! *waves* Thank goodness we have the First Amendment [here in America] and I can just go to a library to find books with that info anyway.