Statement on US government directive to suspend access to Fable 5 and Mythos 5

Just more corrupt behavior from the contemptible kakistocracy that's busy running things into the ground and enriching themselves while they're at it.

But it seems likely that they took this possibility into account - and that they now prominently and unremovable show the "Fable not avaiable" (link - government said so) is likely with the intention to make pressure on the US government.

https://iea.org.uk/yet-you-participate-in-society-in-defence...

They should have consulted their own models about the ramifications and unintended consequences; based on their actions over the past few months I think it is safe to say that the models are smarter than the decision-makers at anthropic, lol. I know the models are smarter than I am and even I could have told them that they were taking paths, FUD for example, that would lead to grief.

If you look outside HN, you'll see that people who interacted with Fable 5 overwhelmingly thought that it was a significant improvement, not simply an incremental one. Most reputable benchmarks show this as well.

I also think that’s a big clown show. People think that LLMs accidentally get good with security patterns. That is not the case, they included all of that in the training data. They could also have left out the knowledge.

Not only that, but also a golden opportunity to flex the muscle of anti-immigration.

It's easy to wave this aside as the current administration playing political games. But I don't think there is any reason to assume that the current era of open availability of models is going to continue indefinitely. Do you think that Chinese labs will continue to release open models forever, even why they get to the level that Mythos is at now, and beyond? And do you think that a competent US government would have no interest in regulating and restricting model access in 2 years time, assuming that model capabilities continue to improve? I think we bias towards thinking the status quo is the norm and will continue, but this news invites us to question that assumption and think about different ways the future could go.

This seems to be an administration that is dead set on antagonizing every single part of the world. Either enemy or ally.

So it's silly behavior, as typified by the last decade of American governance? Is there "serious" American leadership we should be expecting to see soon, e.g. 2029 AOC elected on a platform of unlimited 10GW datacenters and universal basic Mythos 8 models?

It may seem subjectively silly to you, but e.g. getting executed for refusing to point at a deer and call it a horse is pretty silly stuff as well, at least for those not living in the Qin Dynasty.

US voters deserve better.

Deserve's got nothing to do with it.

Yes, something better than a winner-take-all system.

We have now enough data about modern democracies to conclude that presidential and semi-presidential republics are flawed.

Winner-take-all mechanics are not democratic, period, they give voters very little choice to be represented (generally a handful of parties), just two in the US.

People aren't "conservative" or "liberal". They have a huge and diverse array on topics ranging from education, immigration, healthcare, privacy, civil rights, public spending, foreign policies, etc, etc, etc. Yet we bind people to choose among a handful of parties, which at best overlap with some of those opinions, just to elect a single person that is extremely hard to remove (both from a legal point of view and a from that person's rightful ability to claim popular mandate), that do not rely on confidence votes nor their own party support.

And that individual, in the end, really represents fully a very minority of the country (because of the points before) and can also do the opposite of what the promises were, unchecked.

And from a democracy safety itself, please not that every single country turning authoritarian in the last 60 years has been either presidential or semi presidential. Sri Lanka in the 70s is the only exception, there are no others. All others have been presidential.

"Silly" is a silly word for corruption.

Obviously nobody actually buys this (now 6 decade-old) ridiculous fiction that structured signals on a network (ie, information, thought, knowledge) are tantamount to a weapon that can be exported. I don't think many people believe it's the legitimate purvey of government to police what can be said on the internet at all.

But instead of the well-earned, "go to bed, grandpa, you're drunk again", companies act as though article II of the constitution specifically said, "vested in the executive branch is the power to violently repress proliferation of knowledge."

So yes, it's a silly idea by an outdated institution which is swirling the drain. But it's not being treated that way where it counts.

No - it's extremely effective.

Do you realize the difference between a 'few people using VPNs + fake IDs at 2-person companies ... vs companies all companies globally not allowed to use tech?

If 'Bank of Montreal' were caught using export controlled technology it could be devastating - so they're not going to be using it along with any little mom and pop shop.

We don't know what the Administration is doing other than 'This is Extremely Heavy Handed' and will have devastating consequences if it goes on.

10 years from now we’ll look back and laugh at how silly it all was.

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

I can hear alarm bells going off in less silly governments around the world as we speak. Genie's out of the bottle. The gears have been put in motion. Etc.

With due respect, this take is very deluded. US voters have very little to lose if the tech is not available to the rest of the world. US politicians and elite, regardless of political inclination, understand the enormous strategic potential of this technology and will ITAR the shit out of frontier models and/or use them as leverage for extracting concessions out of other countries.

The main losers are Big AI labs, their investors, foreign employees and rest of the world.

Fwiw, China and other countries would’ve done the exact same thing. It’s perhaps the game theoretic optimal approach when your comparative advantage is so vast (capital, compute, talent, embedded knowledge) and keeps growing especially if RSI is real (making it nearly impossible for anyone else to catch up)

Can we stop with this bothsides-ing. The level of co-opting by this administration is unprecedented. There’s the strong-arming to get Intel equity stake, Nvidia/AMD revenue share, U.S. Steel golden share, Lithium Americas equity stake, Big Law pro bono pledges, TikTok forcible acquisition, Paramount-CBS-Skydance favor, it’s just unbelievable the stark use of power.

I suspect if OpenAI or Grok was operating at the same level they wouldn’t find themselves on the sharp end of the government stick

ChatGPT 2 was once too powerful to release.

AI has been moving faster than culture and thinking around it. Once we've adapted to what these models can do we'll relax a little, and then a new stepwise improvement will start it all over again. It always goes this way.

95% odds this gets reversed by Monday morning is why

Same as the new bridge between Windsor and Detroit can't open until some palms are greased.

Chaos is a ladder, gotta keep climbing

The US government believes that Fable/Mythos is a weapon that needs to be export-controlled, and limited to only US customers. Presumably OpenAI/xAI/Google would face the same constraints, for the same reasons.

OS/foreign models are unaffected - OS because they cannot control who runs them, and foreign because they are not controlled by the US government. We could assume that China will implement the same policy controls, but they see the world differently so might not.

So US AI companies are then limited to the US market, effectively, after about six months (the lag between the current frontier models and the OS models). They have much less incentive to push the envelope to create better models, because the US govt might also ban those completely.

The investor froth around the race to AGI dies, so valuations shrink (the current IPOs may be affected), and presumably the bubble bursts. None of the AI companies can afford to continue building data centres, so that all dies immediately. US GDP drops by ~5% because of that alone.

In a year's time, the US is in a major recession because it gambled so hard on AI. Europe less so, only because it was such a distant follower in that race. China is more-or-less unaffected. The best models are now OS/foreign, and AI is moving forward more slowly, but still moving forward.

Any other scenarios?

That's because if things keep advancing at this pace, in a couple of years, we will probably have Mythos-level, open weights models running on consumer or prosumer GPUs. Even if it takes a decade, that's no time in the grand scheme of things. And when the time comes, what will government do?

They will probably regulate who can buy GPUs and other tech that can run powerful AI, and what are people allowed to do with them. Probably using trusted computing and remote attestation and so on.

The public may not see more improvements but I'm sure they government will be forcing AI companies to continue improving them for their own use. Those schools aren't going to bomb themselves.

What an unbelievable claim. Especially since the vast majority of publicly available benchmarks disagree.

Efficiency will become the next thing to focus on. It was already emerging, but accelerating the focus on efficiency will lead to a ton of excess capacity and even some investments in data centers to go belly up. And ultimately the AI bubble bursting will look a lot like the dot com, with its surplus fiber.

Oh, and this will put gas on the fire that fighting AI and big tech is the next political rally cry. Along with “eat the rich” as they are seen as taking both jobs and money.

Curious to see where it’s all headed and how Trump’s call will impact it.

https://news.ycombinator.com/item?id=48510010

They’re falling back to Opus 4.8. Most people weren’t using Fable for everything anyway because it’s so expensive.

None of open weights models are even at Opus 4.8 levels. If someone was using Fable they don’t have any second best alternative outside of Anthropic.

It's tough, US technologies are everywhere but they are only liabilities. Microsoft is the stickiest of all.

Anthropic got the most rewarding hype ever in the history of mankind.

Imagine a private company invents a piece of technology soooo good that the US government has to issue a ban.

Did the government ban any models from Google or OpenAI? Nah, Russian/Chinese spies and ISIS are welcome to use those dumb models.

Anthropic will probably go for $2T IPO now.

"Mythos capabilities" is not some magic threshold. This is exactly the type of language that people used about GPT-4 in 2023. Today, I can run models far stronger than GPT-4 on my laptop at speeds better than GPT-4 offered.

Anthropic are quite good at coining sticky phrases like "Mythos-class models", but these are manipulative attempts to shape the discourse for business purposes and should be identified as such.

Ever since the original GPT-2 "it's too powerful to release!" I've realized that whatever is the current state of open models represents what we really have access to.

It's shocking to me how many people on HN, who engage in long conversations about LLMs and AI, have never actually run a model on their own hardware.

All you need is a reasonably good macbook pro/studio or an RTX [3-5]090 and you can run useful models in the >= 30 tokens/second range (much higher if you choose the GPU path). The difference between what you can run on this hardware and what you can run on hardware that costs 2-5x is not that big. Don't be fooled by people on Twitter/X claiming you need some outrageous setup.

It's also increasingly clear that frontier models are nowhere near close to pushing the limits of efficiency. Quantization, MoE, and other techniques have dramatically improved even in the last year.

For work, of course use OpenAI/Anthropic models, but for anything personal, anyone who considers themselves a "real engineer" should be running local models, using open harnesses and seeing what they can accomplish with these.

Even if open releases slow down or even stop, we have the foundation, right now, for smart engineers to squeeze something quite useful out of. Hopefully we'll one day figure out how to train large models in a federated way. But either way: not your weights, not your inference.

The US government found a jailbreak that allowed the user to make Fable do bad things, this is so dangerous that this model must be held back in areas that are not the US...

If this is so dangerous why allow US nationals access to it? Are there no evil people in the US?

Going back to my perspective: let's say I control a big enterprise or a government body, how should I view this or US technology? Should I be like: yes, let's use US tech, they are a reliable partner and would never abruptly cut us off! Or should I be like: there are competent alternatives out there and if your work hinges on wether or not you had access to Fable 5, then your business is probably not going to survive for long.

Training best models is hella expensive. Anthropic spent fortune to train it, and it definitely plans to make a fortune with it either. This US decision, if not reversed, would cause Anthropic potentially tens of billions of dollars of revenue loss. When company heads to IPO, and burning cash faster than it generates it, such moment can change their entire trajectory, plans for the future, hiring, new models development, etc.

Alright, one might say that “US will fund it directly and LLMs will move from free market to controlled and funded by government assets”.

But even then. Training new models is expensive not only in terms of computer, and not only in terms of utilities, data centers, etc. But in terms of talent either. It is hard to retain top talents with you when they should just train special models for government. I am not sure we are in 1945 again and that top tier AI researches will agree to sit in silo and work for models which only privileged selected organizations might use. Whenever government steps into control, freedom and creativity is affected.

P.S. Where I agree, though, is that we witness the start of government censorship of AI models. Imagine soon Anthropic open back access to Fable. Can we know what they put inside and which capability limitations, derived based on ID/IP, they enforce? No, we can’t. Here I agree that at least the era of government censorship begins.

Banning foreign nationals from using your technology only makes sense if you don't rely on foreign nationals to build it in the first place.

Or are we so far along now we think we don't need them anymore.

I'm wondering if they might go for a restricted access model that goes beyond passport or citizenship, where people can still use it, but you have to be individually vetted, and put on a list to get security clearance.

That's separate from the question of whether Fable 5 and Mythos 5 are unsafe. I don't really know. Here's a few things that seem real, though: These models probably have some level of capability to assist with bioterrorism, Anthropic has self-admitted that their own safety measures are imperfect [1], so it should come as no surprise that jailbreaks seem far more possible than Anthropic is leading you to believe in this blog post [2].

[1] https://www.anthropic.com/news/fable-mythos-access: "We suspect that perfect jailbreak resistance is not currently possible for any model provider."

[2] https://x.com/elder_plinius/status/2064776322979676227

If Amazon sold a book that taught someone how to commit bioterrorism, would there be action against them to stop selling it? Its an imperfect analogy, but the parallels are there. LLMs don't get a free pass because they're also so good at writing typescript for beige CRUD apps and bedtime stories.

One thing I hope we align on: Synthetic safeguards (steering, rejections, etc) on top of models to block illegal/sensitive topics isn't good enough. Anthropic has self-admitted that it isn't good enough. We need the technology to lobotomize these capabilities the public deems too unsafe to allow out of the models at the most fundamental level. And, we need to align on what the scope of these forbidden fruit topics are. This is, actually, the only way open source continues to thrive. I want open source models to thrive, but they won't be allowed to thrive, nor should we want them to thrive, if they're teaching people how to engineer novel viruses and other horrible stuff.

Also foreign investments into Chinese AI labs have already been forbidden and asked to exit

That’s also the difference between being totalitarian government and laws and regulations based order.

But it is! How many times have OpenAI and Anthropic threatened the rest of humanity with extinction at the hands of their LLMs? Monthly I think.

They were hoping for a government supported monopoly. Careful what you wish for.

I remember something with either ChatGPT or Claude, way early on, where I had to upload my passport to use some level of it (maybe it was the OpenAI API).

Anyway, there's no way they just shut this completely down, the revenue from mythos is huge. So if they can't get the government to budge they'll find a way to be compliant without completely shutting down.

The fact that Anthropic are willing to dumb-down their own model responses to "Prevent foreign competitors from using the model to accelerate R & D and protect our leading position." [1] adds credence to this speculation. Anthropic are scared of their own model's power in the hands of competitors: it has nothing to do with security.

[1] https://eu.36kr.com/en/p/3848820681636481

The problem is that there are network effects at play - the more people you have using your models, the more training and fine-tuning data you're accumulating, so the faster you can develop the next frontier model. Not to mention the fact that more users means more revenue to fund your next-gen model training.

Perhaps the US administration is gambling that US citizens on their own provide enough of a training data and revenue flywheel for them to keep their AI development edge.

The next interesting question will be - will the US share this capability with her traditional strategic allies (e.g. five-eyes countries), or is it truly America First (or, 'America Alone')?

I don't necessarily disagree, but who is going to pay for improvements in models if they're not commercially available? Are AI companies going to become defence contractors with (US?) government paying for the training?

I don't know much it cost to go from AI model Foo 4 to Foo 5, but it's going to cost a pretty penny to eventually go to 6 and 7. These companies are doing so in the expectation of eventually charging customers to recoup the costs: if the only customer(s) is government(s) then the per-unit cost will be much higher. An analogy: you can get a 'civilian' toilet for USD 200, an FAA/EASA-approved toilet for airlines 2000, and a MILSPEC/NASA toilet for 20000.

Now this limited-customer model is certainly doable—tank manufacturers have a smaller base of customers than pickup truck manufacturers—but AI companies probably want regulatory clarity on what they're allowed to sell, and to whom, before they start developing new products/services.

The "Mythos capabilities" is a pretty arbitrary threshold. It could have happened any point before or after. China treats this technology in very different ways. And China's economy and foreign policies are very different, in such export-based economy export controls make much no sense. Unless something changes drastically in the global relations. I expect the opposite to happen, if the US leave a void globally, china will be even more incentivised to fill it.

This been said, it is likely that big models are soon no longer "open sourced" (qwen has already started), but because the chinese companies will prefer to sell access themselves, not because of government intervention.

In any case, what happens now is a huge thing, and not sure we grasp its consequences yet. It is definitely beyond any ai safety marketing stunt. Even if this is rolled back at some point, after "guarantees" are established, it has already moved things to change drastically.

I think this is bang on. The motives are kind of irrelevant, because now that the precedent has been set, I suspect they'll be much more likely to go here for future restrictions. It's very convenient (even if true) to just say "security reasons".

This logic is flawed. China had no incentive to release SOTA models to the world in the first place when OpenAI were milking everyone with closed source paid models. What changes now? Nothing. In fact, this is even more incentive for them to capture marketshare and dependence on Chinese models as the world will simply just use alternatives. Not bow down to restrictions. If your logic were correct that people would just comply, then the tons of VPN services wouldn't have a market in the first place.

So I doubt this action alone is enough to really stop other nations from getting access to state of the art AI. I think the US would have to go much further to really stop other nations from getting access to state of the art AI.

Thus they are far less likely to do something like this.

I can't agree more. This is a precedent not just in denial but possible vagueness. Judiciaries have 'vagueness doctrines' to counter such laws/directives but _these_ may be re-trumped by the deference given to national security.

If we don't get soon a framework by which models may be measured as 'too powerful' vs 'not too powerful' we supercharge the self-dealing (corruption) that this administration has brazenly adopted. Many fingers can be put on many scales; groks may be given a pass while others are held to higher "standards".

Will OpenAI now just asymptotically bump its versions to 5.99999999 to stay under a limit that nobody really understands?

I realize that this has all just happened and we might get some good rigorous clarification from our government.... sigh. We are living in a kakistocracy. Who am I kidding?

and OSS models gave way to running it with freedom and security.

So OSS models have always tried to catch up to the frontier and lag behind 3-6 months. For my use cases, I am happy with current OSS models especially so if you let frontier-ish models design the plan with your input

If I were to suppose that China created a frontier model so good and far ahead, then I can understand if they don't open-source it. Qwen does it already with their Max models being closed-source.

but if you are suggesting that China in whole will remove itself from AI race, then 3 (or 4) possibilities can occur.

1. Some chinese companies might stop the production of OSS models if their names are known (z.ai etc.) but there are multiple other companies who are fighting with their research labs as well. They might create a decent model and OSS it to get known within world and China.

2. The whole Chinese economy (well similar to America, but to an even more extreme level from my understanding) depends on AI and is a bet on AI. They are funneling state and all bank money into these companies. From point 1, they wouldn't wish to be silent with frontier models and then lag behind and wait for other countries to catch up (point 3)

3. Europe(MistralAI)/India(SarvanAI? Kinda recent) will jump on the opportunity. (My point is that these two regions are trying to create their own models. How much they lack from the frontier is another thing but if China were to remove itself from the race, then they will have much more time to figure out how to make better models)

My point is that america and china are in arms race of closed source vs open source models. If china were to close source its models, they might simply lag behind and other countries will catch up.

4. Either that or you are right and we will have the current frontier OSS models and some more. IMO they are reasonably good as well and I used to wonder what would happen if say it would have been net good if AI was stuck at a similar level to sonnet 4.5 (IMO it was sweet spot), so I don't think that I am reasonably worried about it all. If absolutely need be, you can have an frontier model direct a plan and have OSS models do the grunt work.

For hobby use, only simple models will be allowed. The models of the future will be so large that the huge data centers being built right now will be necessary for training them. Open sourcing will be a strategic nation state level decision. Regular people won't have it. We will have toys, we will have infinite AI generated entertainment and the rest isn't exactly planned out I guess.

That would also significantly dampen the commercial incentives to develop such strong models, given the high costs involved.

On the other hand, such a future would probably save white-collar jobs.

based on Anthropic's own self promotion. no reason to think that Chinese models are not just as good or better. the key thing here is training on machine code and dis-assembled binaries and the Chinese have a complete data set of pirated software, with no limitations on how they use it. I seriously doubt they are actually behind.

> only if you're not a US citizen, but in practice, even if you are

the issue here is that Anthropic needs a legal opinion that their mechanisms for detecting foreign users in the US are compliant, which is technically hard to do, and a complex intersection of technical details and national security law, so getting a legal opinion can't happen overnight. it will be back.

Honestly, and I don't say it lightly, long term this may have bigger impact on humanity as a whole than Iran war and its varying outcomes ( and consequences ). Separately, note how much this news was not really reported much today. Granted, a lot was happening, but it is telling.

China has flaws, plenty of them, but there's no real evidence to believe their motivations or mechanisms of pursuing motivations are that similar to that of the United States.

That sounds so great.

> Will we be the poorer for that, or will we be safer?

We will be not just safer but richer. These LLMs are like drugs that should absolutely not be cast freely into the highways and byways. My main worry is that this action will be a haphazard one-off and not part of a coherent plan of curtailing LLM propagation.

I’ve been trying to track LLM code generation adoption in the critical infrastructure world - as far as I can tell, it’s nill. Zero. Nada. Nobody is relying on these models to write secure code for anything where failure is catastrophic. Planes falling out of the sky. Nuclear reactors going into meltdown. Electrical grids loosing synchronicity. Lots of these BS claims from the marketing and investment crowd, but - it’s just a useful tool for non-critical areas. That’s all it is.

That part is up to Anthropic. KYC[0] is not exotic, it's just a pain in the butt: if Fable is that good, they can do the KYC.

I don't think this is the right move from the government, but we shouldn't pretend that "citizens only" is an insurmountable hurdle for a company that just got a $65B capital infusion.

[0] https://en.wikipedia.org/wiki/Know_your_customer

It doesn't have to be free, we have the means to transact in a p2p fashion electronically as well.

Despite the damage it’s better to build up an immunity through ongoing exposure, unless you want to end up like the previous American civilisations.

Anthropic gets into argument with US government over model usage -> Release a model calling it too advanced for safe use -> release the model to public knowing well that this admin has thinnest of skins and will do something

Regulatory capture in roundabout way. Now it is going to take crying wolf over other companies/countries developing “Mythos grade model” to kick off action especially in next two years of this admin.

Companies will keep improving models because AI is not yet fully there. But it is incredibly naive to think governments were ever going to allow state of the art technology to be released to public or do things this publicly. Every company wants to show off and get publicly restricted because it shows off their strength.

I can only say well played Anthropic.

That's a bold prediction considering that's true today...

Government always restricted data, tools, technology. In France for instance you're not allowed to have a gun, but policemen have.

What's the difference ?

Imo china, and deepseek will keep its open source model because they invest in long term. At some point they could do something similar, but not now.

USA government is just hurting AI development in their country, and that's good news to me.

But who do you trust more to make these decisions? A democratically elected government or a private company?

But it is a computer program and it won’t be long before the dam breaks to open source and anyone can use Mythos level AI at home for anything.

There is no stopping that unless you would set up a Police state more strict than China

It may be a de facto weapon and for better or worse everyone will be able to use it sooner or later.

I predict it will give birth to many great things and many equally terrible. Milions of people will die and milions will be saved. Such is nature of humanity. The good always comes with the bad. That was true for every invention.

Cars are used to kill by terrorists, rockets are used to bomb kids but also to go into space. It’s all same, old story.

You can’t stop LLMs the way you cannot control fire. Everyone can pour gasoline in the forest and cause terrible damage. And any excuse that we shouldn’t have matches must be viewed as what it is - an authoritarian, futile, desire for control

I think this also misses the point. The precedent here almost surely implies that it will be illegal to use these frontier models as well.

I can see a future where weights are distributed on the darkweb or bittorrent, or people are trying to use small fly by night hosts of models.

But if this says these models are dangerous and the companies and people can't be trusted with them, then I don't see why that wouldn't also apply to open weight models.

I mean, maybe in principle, but if the object is just hobbling Anthropic you might still get OpenAI's latest model without that much trouble.

The whole idea is a lie and a marketing stunt to prop up the US stock market.

A pathogen with a very long incubation time and a high fatality rate would be about as bad as nuclear war. Maybe we need to figure out how to possibly defend against one person doing this before making it easy for anyone to do it.

The government is full of stupidity and this is indeed a big moment, but Anthropic has been begging for this outcome in their public messaging. If their fear-mongering was genuine, then great, they got their pause. If not, then what exactly did they want to happen?

Those of us who are negative about AI for political reasons have been saying from the start that the biggest problem with AI is power. People can’t now all of a sudden be thinking that huh nation states have power (along with Big Tech and the rest of the power brokers).

But this is in fact quite a tortured fear, all wrapped up in the usual hyping—though this part is expected of LLM Euphorics. The usual story of simply making human labor less valuable and concentrating hardware for compute is just, you know, this rotten economic system working as it is intended. No need for weapons, subterfuge, three-letter agencies, much more straightforward, and just a natural evolution of X-CLASS CAPABILITIES.

At first people said this was tin-foil hat territory. But ANYONE who publicly pisses this guy off, mysteriously get a government takedown weeks later. They're not even pretending anymore.

When Trump attacked them before because he wanted anthropic to decide who lives and dies, they said no. (That's probably a lie, I'm sure it has to do with money - but I digress).

So exec ban happened. Problem is - everyone uses claude. Microsoft is going through the same thing now. They find a way to use it anyway.

So they reverted it. What better way to go around circumvention than to just outright ban it.

Funniest thing - his own law makers are the ones who run on freedom and a nanny state. They are literally preventing us from using a tech "for our own safety" - can't get more nanny than that.

It would be too naive to suppose that the strongest LLMs are available to plebs now.

First I want to see them play video games at a high skill level, preferably without any access to game state beyond the same visual output that humans have access to, like a raster frame X number of times per second. One LLM model played Factorio, albeit at a very, very poor level, which can be seen if you slow the video to 0.25 playback speed and pause frequently.

https://old.reddit.com/r/factorio/comments/1u1blr6/claude_fa...

There have been streams of other games, where LLMs and AIs have likewise performed very poorly.

I recognize that LLMs might be better at language processing than these sorts of tasks. But being able to play video games is part of general capability. And this kind of hardcore video game playing, with no access to game state, is also a general task where feigning skill can be harder. If LLMs excel at pretending to be competent without actually being competent, like this AI training approach is arguably about

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

Then some AIs might be trained and designed for deceiving humans instead of actually being competent and capable. And thus, one response is that they should be met with more difficult tests.

Basically, make tests that AIs or LLMs will not have an easy time cheating. Hopefully, that will engender research in greater LLM/AI competence, not in greater ability to cheat or deceive, neither for LLM/AI researchers and companies, nor for LLMs/AIs themselves.

I would be surprised if the public ever had anything close to the strongest LLM. It’s not like nuclear bombs were created by the private sector, then the government started the Manhattan Project and seized them all for itself.

They probably had Fable-quality models in 2016.