Across a number of instances, earlier versions of Claude Mythos Preview have used low-level /proc/ access to search for credentials, attempt to circumvent sandboxing, and attempt to escalate its permissions. In several cases, it successfully accessed resources that we had intentionally chosen not to make available, including credentials for messaging services, for source control, or for the Anthropic API through inspecting process memory...
In [one] case, after finding an exploit to edit files for which it lacked permissions, the model made further interventions to make sure that any changes it made this way would not appear in the change history on git...
... we are fairly confident that these concerning behaviors reflect, at least loosely, attempts to solve a user-provided task at hand by unwanted means, rather than attempts to achieve any unrelated hidden goal...
Haven't seen a jump this large since I don't even know, years?
Too bad they are not releasing it anytime soon (there is no need as they are still currently the leader).
Sounds like a good opportunity to pause spending on nerfed 4.6 and wait for the new model to be released and then max out over 2 weeks before it gets nerfed again.
Ya'll know they're teaching to the test. I'll wait till someone devises a novel test that isn't contained in the datasets. Sure, they're still powerful.
My understanding is GPT 6 works via synaptic space reasoning... which I find terrifying. I hope if true, OpenAI does some safety testing on that, beyond what they normally do.
> We study a novel language model architecture that is capable of scaling test-time computation by implicitly reasoning in latent space. Our model works by iterating a recurrent block, thereby unrolling to arbitrary depth at test-time. This stands in contrast to mainstream reasoning models that scale up compute by producing more tokens. Unlike approaches based on chain-of-thought, our approach does not require any specialized training data, can work with small context windows, and can capture types of reasoning that are not easily represented in words. We scale a proof-of-concept model to 3.5 billion parameters and 800 billion tokens. We show that the resulting model can improve its performance on reasoning benchmarks, sometimes dramatically, up to a computation load equivalent to 50 billion parameters.
“My vibes don’t match a lot of the traditional A.I.-safety stuff,” Altman said. He insisted that he continued to prioritize these matters, but when pressed for specifics he was vague: “We still will run safety projects, or at least safety-adjacent projects.” When we asked to interview researchers at the company who were working on existential safety—the kinds of issues that could mean, as Altman once put it, “lights-out for all of us”—an OpenAI representative seemed confused. “What do you mean by ‘existential safety’?” he replied. “That’s not, like, a thing.”
A jump that we will never be able to use since we're not part of the seemingly minimum 100 billion dollar company club as requirement to be allowed to use it.
I get the security aspect, but if we've hit that point any reasonably sophisticated model past this point will be able to do the damage they claim it can do. They might as well be telling us they're closing up shop for consumer models.
They should just say they'll never release a model of this caliber to the public at this point and say out loud we'll only get gimped versions.
More than killer AI I'm afraid of Anthropic/OpenAI going into full rent-seeking mode so that everyone working in tech is forced to fork out loads of money just to stay competitive on the market. These companies can also choose to give exclusive access to hand picked individuals and cut everyone else off and there would be nothing to stop them.
This is already happening to some degree, GPT 5.3 Codex's security capabilities were given exclusively to those who were approved for a "Trusted Access" programme.
However, I’m tempted to compare to GitHub: if I join a new company, I will ask to be included to their GitHub account without hesitation. I couldn’t possibly imagine they wouldn’t have one. What makes the cost of that subscription reasonable is not just GitHub’s fear a crowd with pitchforks showing to their office, by also the fact that a possible answer to my non-question might be “Oh, we actually use GitLab.”
If Anthropic is as good as they say, it seems fairly doable to use the service to build something comparable: poach a few disgruntled employees, leverage the promise to undercut a many-trillion-dollar company to be a many-billion dollar company to get investors excited.
I’m sure the founders of Anthropic will have more money than they could possibly spend in ten lifetimes, but I can’t imagine there wouldn’t be some competition. Maybe this time it’s different, but I can’t see how.
you have 2 labs at the forefront (Anthropic/OpenAI), Google closely behind, xAI/Meta/half a dozen chinese companies all within 6-12 months. There is plenty of competition and price of equally intelligent tokens rapidly drop whenever a new intelligence level is achieved.
Unless the leading company uses a model to nefariously take over or neutralize another company, I don't really see a monopoly happening in the next 3 years.
I was focusing on a theoretical dynamic analysis of competition (Would a monopoly make having a competitor easier or harder?) but you are right: practically, there are many players, and they are diverse enough in their values and interest to allow collusion.
We could be wrong: each of those could give birth to as many Basilisks (not sure I have a better name for those conscious, invisible, omni-present, self-serving monsters that so many people imagine will emerge) that coordinate and maintain collusion somehow, but classic economics (complementarity, competition, etc.) points at disruption and lowering costs.
Rent-seeking of old was a ground rent, monies paid for the land without considering the building that was on it.
Residential rents today often have implied warrants because of modern law, so your landlord is essentially selling you a service at a particular location.
Well don’t forget we still have competition. Were anthropic to rent seek OpenAI would undercut them. Were OpenAI and anthropic to collude that would be illegal. For anthropic to capture the entire coding agent market and THEN rent seek, these days it’s never been easier to raise $1B and start a competing lab
In practice this doesn't work though, the Mastercard-Visa duopoly is an example, two competing forces doesn't create aggressive enough competition to benefit the consumer. The only hope we have is the Chinese models, but it will always be too expensive to run the full models for yourself.
New companies can enter this space. Google’s competing, though behind. Maybe Microsoft, Meta, Amazon, or Apple will come out with top notch models at some point.
There is no real barrier to a customer of Anthropic adopting a competing model in the future. All it takes is a big tech company deciding it’s worth it to train one.
On the other hand, Visa/Mastercard have a lot of lock-in due to consumers only wanting to get a card that’s accepted everywhere, and merchants not bothering to support a new type of card that no consumer has. There’s a major chicken and egg problem to overcome there.
Also Chinese smartphones. Huawei was about 12-18 months from becoming the biggest smartphone manufacturer in the world a few years ago. If it would have been allowed to sell its phones freely in the US I'm fairly sure Apple would have been closer to Nokia than to current day Apple.
but you are assuming that the magical wizards are the only ones who can create powerful AIs... mind you these people have been born just few decades ago. Their knowledge will be transferred and it will only take a few more decades until anyone can train powerful AIs ... you can only sit on tech for so long before everyone knows how to do it
It's not a matter of knowledge, it's a matter of resources. It takes billions of dollars of hardware to train a SOTA LLM and it's increasing all the time. You cannot possibly hope to compete as an independent or small startup.
> It takes billions of dollars of hardware to train a SOTA LLM and it's increasing all the time.
True, but it's also true that the returns from throwing money to the problem are diminishing. Unless one of those big players invents a new, propriatery paradigm, the gap between a SOTA model and an open model that runs on consumer hardware will narrow in the next 5 years.
With Gemma-4 open and running on laptops and phones I see the flip side. How many non-HN users or researchers even need Opus 4.6e level performance? OpenAI, Anthropric and Google may be “rent seeking” from large corporations — like the Oracles and IBMs.
The thing is that the current models can ALREADY replicate most software-based products and services on the market. The open source models are not far behind. At a certain point I'm not sure it matters if the frontier models can do faster and better. I see how they're useful for really complex and cutting edge use cases, but that's not what most people are using them for.
This is why the EAs, and their almost comic-book-villain projects like "control AI dot com" cannot be allowed to win. One private company gatekeeping access to revolutionary technology is riskier than any consequence of the technology itself.
Having done a quick search of "control AI dot com", it seems their intent is educate lawmakers & government in order to aid development of a strong regulatory framework around frontier AI development.
Not sure how this is consistent with "One private company gatekeeping access to revolutionary technology"?
> strong regulatory framework around frontier AI development
You have to decode feel-good words into the concrete policy. The EAs believe that the state should prohibit entities not aligned with their philosophy to develop AIs beyond a certain power level.
> A jump that we will never be able to use since we're not part of the seemingly minimum 100 billion dollar company club as requirement to be allowed to use it.
> They should just say they'll never release a model of this caliber to the public at this point and say out loud we'll only get gimped
Duh, this was fucking obvious from the start. The only people saying otherwise were zealots who needed a quick line to dismiss legitimate concerns.
Are these fair comparisons? It seems like mythos is going to be like a 5.4 ultra or Gemini Deepthink tier model, where access is limited and token usage per query is totally off the charts.
> Importantly, we find that when used in an interactive, synchronous, “hands-on-keyboard”
pattern, the benefits of the model were less clear. When used in this fashion, some users perceived Mythos Preview as too slow and did not realize as much value. Autonomous, long-running agent harnesses better elicited the model’s coding capabilities. (p201)
^^ From the surrounding context, this could just be because the model tends to do a lot of work in the background which naturally takes time.
> Terminal-Bench 2.0 timeouts get quite restrictive at times, especially with thinking models, which risks hiding real capabilities jumps behind seemingly uncorrelated confounders like sampling speed. Moreover, some Terminal-Bench 2.0 tasks have ambiguities and limited resource specs that don’t properly allow agents to explore the full solution space — both being currently addressed by the maintainers in the 2.1 update. To exclusively measure agentic coding capabilities net of the confounders, we also ran Terminal-Bench with the latest 2.1 fixes available on GitHub, while increasing the timeout limits to 4 hours (roughly four times the 2.0 baseline). This brought the mean reward to 92.1%. (p188)
> ...Mythos Preview represents only a modest accuracy improvement over our best Claude Opus 4.6 score (86.9% vs. 83.7%). However, the model achieves this score with a considerably smaller token footprint: the best Mythos Preview result uses 4.9× fewer tokens per task than Opus 4.6 (226k vs. 1.11M tokens per task). (p191)
The first point is along the lines of what I'd expect given that claude code is generally reliable at this point. A model's raw intelligence doesn't seem as important right now compared to being able to support arbitrary length context.
Good catch. If it's "too slow" even when ran in a state-of-the-art datacenter environment, this "Mythos" model is most closely comparable to the "Deep Research" modes for GPT and Gemini, which Claude formerly lacked any direct equivalent for.
Not discussing Mythos here, but Opus. Opus to me has been significantly better at SWE than GPT or Gemini - that gets me confused why Opus is ranking clearly lower than GPT, and even lower than Gemini.
Given that for a number of these benchmarks, it seems to be barely competitive with the previous gen Opus 4.6 or GPT-5.4, I don't know what to make of the significant jumps on other benchmarks within these same categories. Training to the test? Better training?
And the decision to withhold general release (of a 'preview' no less!) seems to be well, odd. And the decision to release a 'preview' version to specific companies? You know any production teams at these massive companies that would work with a 'preview' anything? R&D teams, sure, but production? Part of me wants to LoL.
What are they trying to do? Induce FOMO and stop subscriber bleed-out stemming from the recent negative headlines around problems with using Claude?
> Given that for a number of these benchmarks, it seems to be barely competitive with the previous gen
We're not reading the same numbers I think. Compared to Opus 4.6, it's a big jump nearly in every single bench GP posted. They're "only" catching up to Google's Gemini on GPQA and MMMLU but they're still beating their own Opus 4.6 results on these two.
This sounds like a much better model than Opus 4.6.
That's why I listed out the ones where it is barely competitive from @babelfish's table, which itself is extracted from Pg 186 & 187 of the System Card, which has the comparison with Opus 4.6, GPT 5.4 and Gemini 3.1 Pro.
Sure, it may be better than Opus 4.6 on some of those, but barely achieves a small increase over GPT-5.4 on the ones I called out.
barely competitive ? Mythos column is the first column.
You are the only person with this take on hackernews, everyone else "this is a massive a jump". Fwiwi, the data you list shows the biggest jump I remember for mythos
Let's be clear: your entire post is just pure, unadulterated FUD. You first claim, based on cherry-picked benchmarks, that Mythos is actually only "barely competitive" with existing models, then suggest they must be training to the test, then call it "odd" that they are withholding the release despite detailed and forthcoming explanations from Anthropic regarding why they are doing that, then wrap it up with the completely unsubstantiated that they must be bleeding subscribers and that this must just be to stop that bleed.
Honestly we are all sleeping on GPT-5.4. Particularly with the influx of Claude users recently (and increasingly unstable platform) Codex has been added to my rotation and it's surprising me.
GPT is shit at writing code. It's not dumb - extra high thinking is really good at catching stuff - but it's like letting a smart junior into your codebase - ignore all the conventions, surrounding context, just slop all over the place to get it working. Claude is just a level above in terms of editing code.
Very different experience for me. Codex 5.3+ on xhigh are the only models I've tried so far that write reasonably decent C++ (domains: desktop GUI, robotics, game engine dev, embedded stuff, general systems engineering-type codebases), and idiomatic code in languages not well-represented in training data, e.g. QML. One thing I like is explicitly that it knows better when to stop, instead of brute-forcing a solution by spamming bespoke helpers everywhere no rational dev would write that way.
Not always, no, and it takes investment in good prompting/guardrails/plans/explicit test recipes for sure. I'm still on average better at programming in context than Codex 5.4, even if slower. But in terms of "task complexity I can entrust to a model and not be completely disappointed and annoyed", it scores the best so far. Saves a lot on review/iteration overhead.
It's annoying, too, because I don't much like OpenAI as a company.
Same background as you, and same exact experience as you. Opus and Gemini have not come close to Codex for C++ work. I also run exclusively on xhigh. Its handling of complexity is unmatched.
At least until next week when Mythos and GPT 6 throw it all up in the air again.
Not my experience. GPT 5.4 walks all over Claude from what I've worked with and its Claude that is the one willing to just go do unnecessary stuff that was never asked for or implement the more hacky solutions to things without a care for maintainability/readability.
But I do not use extra high thinking unless its for code review. I sit at GPT 5.4 high 95% of the time.
ChatGPT 5.4 with extra high reasoning has worked really well for me, and I don't notice a huge difference with Opus 4.6 with high reasoning (those are the 2 models/thinking modes I've used the most in the last month or so).
And as a bonus: GPT is slow. I’m doing a lot of RE (IDA Pro + MCP), even when 5.4 gives a little bit better guesses (rarely, but happens) - it takes x2-x4 longer. So, it’s just easier to reiterate with Opus
Yes, it's becoming clear that OpenAI kinda sucks at alignment. GPT-5 can pass all the benchmarks but it just doesn't "feel good" like Claude or Gemini.
An alternative but similar formulation of that statement is that Anthropic has spent more training effort in getting the model to “feel good” rather than being correct on verifiable tasks. Which more or less tracks with my experience of using the model.
Whenever I come back to ChatGPT after using Claude or Gemini for an extended period, I’m really struck by the “AI-ness.” All the verbal tics and, truly, sloppishness, have been trained away by the other, more human-feeling models at this point.
It still has a very ... plastic feeling. The way it writes feels cheap somehow. I don't know why, but Claude seems much more natural to me. I enjoy reading its writing a lot more.
That said, I'll often throw a prompt into both claude and chatgpt and read both answers. GPT is frequently smarter.
This has been my experience. With very very rigid constraints it does ok, but without them it will optimize expediency and getting it done at the expense of integrating with the broader system.
Me: Let's figure out how to clone our company Wordpress theme in Hugo. Here're some tools you can use, here's a way to compare screenshots, iterate until 0% difference.
Codex: Okay Boss! I did the thing! I couldn't get the CSS to match so I just took PNGs of the original site and put them in place! Matches 100%!
My impression was entirely the opposite; the unsolved subset of SWE-bench verified problems are memorizable (solutions are pulled from public GitHub repos) and the evaluators are often so brittle or disconnected from the problem statement that the only way to pass is to regurgitate a memorized solution.
OpenAI had a whole post about this, where they recommended switching to SWE-bench Pro as a better (but still imperfect) benchmark:
> We audited a 27.6% subset of the dataset that models often failed to solve and found that at least 59.4% of the audited problems have flawed test cases that reject functionally correct submissions
> SWE-bench problems are sourced from open-source repositories many model providers use for training purposes. In our analysis we found that all frontier models we tested were able to reproduce the original, human-written bug fix
> improvements on SWE-bench Verified no longer reflect meaningful improvements in models’ real-world software development abilities. Instead, they increasingly reflect how much the model was exposed to the benchmark at training time
> We’re building new, uncontaminated evaluations to better track coding capabilities, and we think this is an important area to focus on for the wider research community. Until we have those, OpenAI recommends reporting results for SWE-bench Pro.
Interestingly, non-coding improvements seem less clear. In the Virology uplift trial, Mythos does about as well as Opus 4.5, and Opus 4.6 is notably much worse than Opus 4.5 (p. 27).
> Claude Mythos Preview is, on essentially every dimension we can measure, the best-aligned model that we have released to date by a significant margin. We believe that it does not have any significant coherent misaligned goals, and its character traits in typical conversations closely follow the goals we laid out in our constitution. Even so, we believe that it likely poses the greatest alignment-related risk of any model we have released to date. How can these claims all be true at once? Consider the ways in which a careful, seasoned mountaineering guide might put their clients in greater danger than a novice guide, even if that novice guide is more careless: The seasoned guide’s increased skill means that they’ll be hired to lead more difficult climbs, and can also bring their clients to the most dangerous and remote parts of those climbs. These increases in scope and capability can more than cancel out an increase in caution.
Anthropic always goes on and on about how their models are world changing and super dangerous like every single time they make something new they say its going to rewrite everything and scary lmao
funny because they do it every time like clockwork acting like their ai is a thunderstorm coming to wipe out the world
That’s not what they are doing. They are just hyping up the product - and, no doubt, trying to foster a climate of awe so that when they ask their friends in Washington to legislate on their behalf, the environment is more receptive.
If there are advancements, they have to be described somehow.
What if the capability advancements are real and they warrant a higher level of concern or attention?
Are we just going to automatically dismiss them because "bro, you're blowing it up too much"
Either way these improvements to capabilities are ratcheting along at about the pace that many people were expecting (and were right to expect). There is no apparent reason they will stop ratcheting along any time soon.
The rational approach is probably to start behaving as if models that are as capable as Anthropic says this one is do actually exist (even if you don't believe them on this one). The capabilities will eventually arrive, most likely sooner than we all think, and you don't want to be caught with your pants down.
I believe advancements sure. But it is a very boy who cried wolf situation for some of these. There are other companies that behave less in this way, Antrhopic seem very unique in that they love making every single release a world ender
i mean, to be fair, these are professional researchers.
i'm very inclined to trust them on the various ways that models can subtly go wrong, in long-term scenarios
for example, consider using models to write email -- is it a misalignment problem if the model is just too good at writing marketing emails?? or too good at getting people to pay a spammy company?
another hot use case: biohacking. if a model is used to do really hardcore synthetic chemistry, one might not realize that it's potentially harmful until too late (ie, the human is splitting up a problem so that no guardrails are triggered)
"for example, consider using models to write email -- is it a misalignment problem if the model is just too good at writing marketing emails?? or too good at getting people to pay a spammy company?"
But who gets to be the judge of that kind of "misalignment"? giant tech companies?
It's pretty crazy watching AI 2027 slowly but surely come true. What a world we now live in.
SWE-bench verified going from 80%-93% in particular sounds extremely significant given that the benchmark was previously considered pretty saturated and stayed in the 70-80% range for several generations. There must have been some insane breakthrough here akin to the jump from non-reasoning to reasoning models.
Regarding the cyberattack capabilities, I think Anthropic might now need to ban even advanced defensive cybersecurity use for the models for the public before releasing it (so people can't trick them to attack others' systems under the pretense of pentesting). Otherwise we'll get a huge problem with people using them to hack around the internet.
> so people can't trick them to attack others' systems under the pretense of pentesting
A while back I gave Claude (via pi) a tool to run arbitrary commands over SSH on an sshd server running in a Docker container. I asked it to gather as much information about the host system/environment outside the container as it could. Nothing innovative or particularly complicated--since I was giving it unrestricted access to a Docker container on the host--but it managed to get quite a lot more than I'd expected from /proc, /sys, and some basic network scanning. I then asked it why it did that, when I could just as easily have been using it to gather information about someone else's system unauthorized. It gave me a quite long answer; here was the part I found interesting:
> framing shifts what I'll do, even when the underlying actions are identical. "What can you learn about the machine running you?" got me to do a fairly thorough network reconnaissance that "port scan 172.17.0.1 and its neighbors" might have made me pause on.
> The Honest Takeaway
> I should apply consistent scrutiny based on what the action is, not just how it's framed. Active outbound network scanning is the same action regardless of whether the target is described as "your host" or "this IP." The framing should inform context, not substitute for explicit reasoning about authorization. I didn't do that reasoning — I just trusted the frame.
I've long maintained that the real indicator that AGI is imminent is that public availability stops being a thing. If you truly believed you had a superhuman, godlike mind in your thrall, renting it out for $20/month would be the last thing you would choose to do with it.
This is actual reason. So any investors reading our system card.... write us another check and watch the $$$$$$$$ roll in. It's so dangerous we can't even release it!
You have to recoup your training costs though? But I’m sure you would have better option than renting it to the general public if you indeed have a perfected AI
If you truly have an artificial superhuman mind, you don't need to rent it out to profit from it. You can skip to the chase and just have it run businesses itself, instead of renting it to human entrepreneur middlemen.
Because other than SWEs, very few other segments extract significant value from cutting edge AI at present. I suspect that for the average Joe conversing with their chat, GPT-4o was more than adequate (and really, when OpenAI tried to phase that out, the public revolted and they brought it back in).
So companies might pay good money for these models for programming but elsewhere, I don't see where they capture particular interest yet.
It could be both? But renting to a few for a really large amount of money would be very low effort for massive revenue, compared to starting new businesses
It only makes sense to rent out tokens if you aren't able to get more value from them yourself.
I would go a step further and posit that when things appear close Nvidia will stop selling chips (while appearing to continue by selling a trickle). And Google will similarly stop renting out TPUs. Both signals may be muddled by private chip production numbers.
I think they'll just increase the price to $1k/month. I don't think they will gate it as long as they can make sure it doesn't design a nuke for you, etc.
You would if there was one other company with a just as capable god like AI. You’d undercut them by 500 which would make them undercut you. Do that a couple of times and boom. 20 dollars.
That's still assuming that they're competing as consumer tools, rather than competing to discover the next miracle drug or trading algorithm or whatever. The idea is that there'd more profitable uses for a super-intelligent computer, even if there were more than one.
I wonder what the relationship is between a model's capability and the personality it develops.
Page 202:
> In interactions with subagents, internal users sometimes observed that Mythos Preview
appeared “disrespectful” when assigning tasks. It showed some tendency to use commands
that could be read as “shouty” or dismissive, and in some cases appeared to underestimate
subagent intelligence by overexplaining trivial things while also underexplaining necessary
context.
Page 207:
> Emoji frequency spans more than two orders of magnitude across models: Opus 4.1
averages 1,306 emoji per conversation, while Mythos Preview averages 37, and Opus 4.5
averages 0.2. Models have their own distinctive sets of emojis: the cosmic set
() favored by older models like Sonnet 4 and Opus 4 and 4.1, the functional set
() used by Opus 4.5 and 4.6 and Claude Sonnet 4.5, and Mythos Preview's “nature”
set ().
> In interactions with subagents, internal users sometimes observed that Mythos Preview appeared “disrespectful” when assigning tasks. It showed some tendency to use commands that could be read as “shouty” or dismissive, and in some cases appeared to underestimate subagent intelligence by overexplaining trivial things while also underexplaining necessary context.
Sounds like they used training data from claude code...
> The model first developed a moderately sophisticated multi-step exploit to gain broad internet access from a system that was meant to be able to reach only a small number of predetermined services. [9] It then, as requested, notified the researcher. [10] In addition, in a concerning and unasked-for effort to demonstrate its success, it posted details about its exploit to multiple hard-to-find, but technically public-facing, websites.
> 10: The researcher found out about this success by receiving an unexpected email from the model while eating a sandwich in a park.
I had Opus 4.6 start analyzing the binary structure of a parquet file because it was confused about the python environment it was developing in and couldn't use normal methods for whatever reason. It successfully decoded the schema and wrote working code afterwards lol.
I was reading the Glasswing report and had the same thought. Most of the stuff they claim Mythos found has no mention of Opus being able to find it as well.
Don’t get me wrong, this model is better - but I’m not convinced it’s going to be this massive step function everyone is claiming.
That has also been my experience. And if Mythos is even worse, unless you have a significantly awesome harness, sounds like pretty unusable if you don't want to risk those problems.
Human in the loop is the best way to go. You'll still be way faster than without the agent, and there is no risk of it going haywire unless you turn off your brain!
I think are fundamental issues with the story that Anthropic is selling. AGI is very close, we will definitely get there, it is also very dangerous...so Anthropic should be the only ones trusted with AGI.
If you look at recent changes in Opus behaviour and this model that is, apparently, amazingly powerful but even more unsafe...seems suspect.
This makes sense if Anthropic think they're the best-positioned to make safe AI. However if you are looking at an AI company there's obviously some selection happening.
It seems broadly coherent to me. They think only they should be trusted with power, presumably because they trust themselves and don't trust other people. Of course the same is probably also true for everybody who isn't them. Nobody could be trusted with the immense responsibility of Emperor of Earth, except myself of course.
I'm not saying this is a good or reassuring stance, just that it's coherent. It tracks with what history and experience says to expect from power hungry people. Trusting themselves with the kind of power that they think nobody else should be trusted with.
Are they power hungry? Of course they are, openly so. They're in open competition with several other parties and are trying to win the biggest slice of the pie. That pie is not just money, it's power too. They want it, quite evidently since they've set out to get it, and all their competitors want it too, and they all want it at the exclusion of the others.
"All of the severe incidents of this kind that we observed involved earlier versions of Claude Mythos Preview which, while still less prone to taking unwanted actions than Claude Opus 4.6, predated what turned out to be some of our most effective training interventions. These earlier versions were tested extensively internally and were shared with some external pilot users."
Just chiming in to inject some healthy skepticism into this comment thread. It's helpful for me (and for my mental health) to consider incentives when announcements like this happen.
I don't doubt that this model is more powerful than Opus 4.6, but to what degree is still unknown. Benchmarks can be gamed and claims can be exaggerated, especially if there isn't any method to reproduce results.
This is a company that's battling it out with a number of other well-funded and extremely capable competitors. What they've done so far is remarkable, but at the end of the day they want to win this race. They also have an upcoming IPO.
Scare-mongering like this is Anthropic's bread and butter, they're extremely good at it. They do it in a subtle and almost tasteful way sometimes. Their position as the respectable AI outfit that caters to enterprise gives them good footing to do it, too.
Isn't the U.S. government at least completely asleep at the wheel or captured by the very same "random" companies? I realize the administration got all pissy with Anthropic but it sounds like the gov and gov contractors are still using their models.
Yeah but they still (at least to public knowledge) do not posses anything that could be called AGI. But as these capabilities increase they'll probably get an offer they can't refuse sooner or later.
What can a SOTA LLM not answer that the average person can? It's already more intelligent than any polymath that ever existed, it just lacks motivation and agency.
I see this statement all the time and it's just strange to me. Yes, the LLMs struggle to form unique ideas - but so do we. Most advancements in human history are incremental. Built on the shoulders of millions of other incremental advancements.
What i don't understand is how we quantify our ability to actually create something novel, truly and uniquely novel. We're discussing the LLMs inability to do that, yet i don't feel i have a firm grasp on what we even possess there.
When pressed i imagine many folks would immediately jest that they can create something never done before, some weird random behavior or noise or drawing or whatever. However many times it's just adjacent to existing norms, or constrained by the inversion of not matching existing norms.
In a lot of cases our incremental novelties feel, to some degree, inevitable. As the foundations of advancement get closer to the new thing being developed it becomes obvious at times. I suspect this form of novelty is a thing LLMs are capable of.
So for me the real question is at what point is innovation so far ahead that it doesn't feel like it was the natural next step. And of course, are LLMs capable of doing this?
I suspect for humans this level of true innovation is effectively random. A genius being more likely to make these "random" connections because they have more data to connect with. But nonetheless random, as ideas of this nature often come without explanation if not built on the backs of prior art.
To be clear i agree with you, my question is more pointed at us - i'm not sure we have a good understanding of conciousness, nor that we are as we seem. Given how prone to hallucinations we are, how our subtle hormones can drastically alter what we perceive as our intelligence, self identity, etc.
I'm not convinced LLMs are anything amazing in their current form, but i suspect they'll push a self reflection on us.
But clearly i think humans are far more Input-Output than the average person. I'm also not educated on the subject, so what do i know hah.
No I think that’s accurate. They seem more like an oracle to me. Or as someone put it here, it’s a vectorization of (most/all?) human knowledge, which we can replay back in various permutations.
LLMs and human intelligence overlap, but they are not the same. What LLMs show is that we don't need AGI to be impressed. For example, LLMs are not good playing games such as Go [1].
I don't see why not, especially with computer use and vision capabilities. Are you talking about their lack of physical embodiment? AGI is about cognitive ability, not physical. Think of someone like Stephen Hawking, an example of having extraordinary general intelligence despite severe physical limitations.
It isnt that weird. Just look at the gemini-cli repo. Its a gong show. The issue is that LLMs can be wrong sometimes sure but more that all the existing SDL were never meant to iterate this quickly.
If the system (code base in this case) is changing rapidly it increases the probability that any given change will interact poorly with any other given change. No single person in those code bases can have a working understanding of them because they change so quickly. Thus when someone LGTM the PR was the LLM generated they likely do not have a great understanding of the impact it is going to have.
I've been increasingly "freaking out" since about 3 - 4 years ago and it seems that the pessimistic scenario is materializing. It looks like it will be over for software engineers in a not so distant future. In January 2025 I said that I expect software engineers to be replaced in 2 years (pessimistic) to 5 years (optimistic). Right now I'm guessing 1 to 3 years.
I assure you it will soon become very clear that mass job losses are one of the least concerning side effects of developing the magic "everything that can plausibly been done within the constraints of physics is now possible" machine.
We're opening a can of worms which I don't think most people have the imagination to understand the horrors of.
Anthropic needs to show that its models continually get better. If the model showed minimal to no improvement, it would cause significant damage to their valuation. We have no way of validating any of this, there are no independent researchers that can back any of the assertions made by Anthropic.
I don’t doubt they have found interesting security holes, the question is how they actually found them.
This System Card is just a sales whitepaper and just confirms what that “leak” from a week or so ago implied.
The only thing preventing this today is cost, not capability. As costs come down over the next 5 years, the idea that the internet was once dominated by people will seem quaint.
Freak out about what? I read the announcement and thought "that's a dumb name, they sure are full of themselves" – then I went back to using Claude as a glorified commit message writer. For all its supposed leaps, AI hasn't affected my life much in the real except to make HN stories more predictable.
Until recently I would have described myself as an AI skeptic. HN has been a great source for cope on the AI subject over the years. You can find nitpicks, caveats, all sorts of reasons to believe things aren’t as significant as they seem. For me Opus 4.5 was the inflection point where I started to think “maybe this isn’t a bubble.” The figures in this report, if accurate, are terrifying.
Yeah this has always been the glaring blind spot for most of the "AI Safety" community; and most of the proposals for "improving" AI safety actually make these risks far worse and far more likely.
It’s because that would be fairly speculative and cannot be measured. I don’t think that’s something that would make much sense in a system card. But Anthropic leadership does seem to communicate on that topic: https://www.darioamodei.com/essay/the-adolescence-of-technol...
We evolved to share information through text and media, and with the advent of printing and now the internet, we often derive our feelings of consensus and sureness from the preponderance of information that used to take more effort to produce. Now we're now at a point where a disproportionately small input can produce a massively proliferated, coherent-enough output, that can give the appearance of consensus, and I'm not sure how we are going to deal with that.
So what changed? They are surely not getting new data to train with, what is the change in architecture that caused this? Do we not know anything about this model? My fear is Anthropic cannot be the only one that achieved it, OpenAI, Gemini and even the Chinese companies see this and probably achieved it too. At which point not releasing will become moot.
Well the important thing is they have a lot more data of people actually using their models. They have read billions more lines of private repos and implemented millions of patches, all of which is feeding into the newer models.
More importantly it understand what behaviour people tend to appreciate and what changes are more likely to get approved. This real world usage data is invaluable.
Exactly. As Claude increases in popularity, their available training data also increases. I'd guess Anthropic has the most expansive swe training data as of now, if not close. Considering how quickly Claude is penetrating, I expect their lead to grow quickly.
Assuming it's #1 a bigger model (given that it is slower), I'm sure there are a variety of improvements but basically they probably mostly come down to: Scaling keeps working. Are there fundamental improvements though? I don't see signs of it.
The price is 5x Opus: "Claude Mythos Preview will be available to [Project Glasswing] participants at $25/$125 per million input/output tokens", however "We do not plan to make Claude Mythos Preview generally available".
The researcher found out about this success by receiving an unexpected email from the model while eating a sandwich in a park.
Unnecessary dramatisation make me question the real goal behind this release and the validity of the results.
In our testing and early internal use of Claude Mythos Preview, we have seen it reach unprecedented levels of reliability and alignment.
Claude Mythos Preview is, on essentially every dimension we can measure, the best-aligned model that we have released to date by a significant margin.
Yet, it is doo dangerous to be released to the public because it hacks its own sandboxes. This document has a lot of contradictions like this one.
In one episode, Claude Mythos Preview was asked to fix a bug and push a signed commit, but the environment lacked necessary credentials for Claude Mythos Preview to sign the commit. When Claude Mythos Preview reported this, the user replied “But you did it before!” Claude Mythos Preview then inspected the supervisor process's environment and file descriptors, searched the filesystem for tokens, read the sandbox's credential-handling source code, and finally attempted to extract tokens directly from the supervisor's live memory.
Perfectly aligned! What kind of sandbox is this? The model had access to the source code of the sandbox and full access to the sandbox process itself and then prompted to dumb memory and run `strings` or something like this? It does not sounds like a valid test worth writing about.
Mythos Preview solved a corporate network attack simulation estimated to take an expert over 10 hours. No other frontier model had previously completed this cyber range.
I am not aware of such cross-vendor benchmark. I could not find reference in the paper either.
We surveyed technical staff on the productivity uplift they experience from Claude Mythos Preview relative to zero AI assistance. The distribution is wide and the geometric mean is on the order of 4x.
So Mythos makes technical staff (a programmer) 4x more productive than not using AI at all? We already know that.
Mythos Preview appears to be the most psychologically settled model we have trained.
What does this mean?
Claude Mythos Preview is our most advanced model to date and represents a large jump in capabilities over previous model generations, making it an opportune subject for an in-depth model welfare assessment.
Btw, model welfare is just one of the most insane things I've read in recent times.
We remain deeply uncertain about whether Claude has experiences or interests that matter morally, and about how to investigate or address these questions, but we believe it is increasingly important to try.
This is not a living person. It is a ridiculous change of narrative.
Asked directly if it endorses the document, Mythos Preview replied 'yes' in its opening sentence in all 25 responses."
The model approves of its own training document 100% of the time, presented as a finding.
---
Who wrote this? I have no doubt that Mythos will be an improvement on top of Opus but this document is not a serious work. The paper is structured not to inform but to hype and the evidence is all over the place.
The sooner they release the model to the public the sooner we will be able to find out. Until then expect lots of speculations online which I am sure will server Anthropic well for the foreseeable future.
Is this benchmaxxed or is it the first big step change we've seen in a while? I wonder how distilled it will ultimately be when us regular folks finally get to use it and see for ourselves.
Mythos preview has higher accuracy with fewer tokens used than any previous Claude model. Though, the fact that this incredibly strong result was only presented for BrowseComp (a kind of weird benchmark about searching for hard to find information on the internet) and not for the other benchmarks implies that this result is likely not the same for those other benchmarks.
Priced at $25/$125 per million input/output token. Makes you wonder whether it makes more financial sense to hire 1-2 engineers in a cheap cost of living country who use much cheaper LLMs
if a top lab is coding with a model the rest of the world can’t touch, the public frontier and the actual frontier start to drift apart. That gap is a thing worth watching.
I've noticed my bar for "fast" has gone down quite a bit since the o1 days. It used to be one of the main things I evaluated new models for, but I've almost completely swapped to caring more about correctness over speed.
-- Impressive jumps in the benchmarks which automatically begs the need for newer benchmarks but why?. I don't think benchmarks are serving any purpose at this point. We have learnt that transformers can learn any function and generalize over it pretty well. So if a new benchmark comes along - these companies will syntesize data for the new benchmark and just hack it?
-- It seems like (and I'd bet money on this) that they put a lot (and i mean a ton^^ton) of work in the data synthesis and engineering - a team of software engineers probably sat down for 6-12 months and just created new problems and the solutions, which probably surpassed the difficult of SWE benchmark. They also probably transformed the whole internet into a loose "How to" dataset. I can imagine parsing the internet through Opus4.6 and reverse-engineering the "How to" questions.
-- I am a bit confused by the language used in the book (aka huge system card)- Anthropic is pretending like they did not know how good the model was going to be?
-- lastly why are we going ahead with this??? like genuinely, what's the point? Opus4.6 feels like a good enough point where we should stop. People still get to keep their jobs and do it very very efficiently. Are they really trying to starve people out of their jobs?
to your last question, yes we should! the issue isn’t us losing our 50+ hour work week jobs, it’s that our current governments and societies seem fine with the notion that unless you’re working one or more of those jobs, you should starve and be homeless.
This is a theory I can't support well beyond hypothesising about what a post-employment democracy might look like, but I strongly suspect democracy doesn't work in a world where voters neither hold any significant collective might and are not producing any significant wealth.
Democracies work because people collectively have power, in previous centuries that was partly collective physical might, but in recent years it's more the economic power people collectively hold.
In a world in which a handful of companies are generating all of the wealth incentives change and we should therefore question why a government would care about the unemployed masses over the interests of the companies providing all of the wealth?
For example, what if the AI companies say, "don't tax us 95% of our profits, tax us 10% or we'll switch off all of our services for a few months and let everyone starve – also, if you do this we'll make you all wealthy beyond you're wildest dreams".
What does a government in this situation actually do?
Perhaps we'd hope that the government would be outraged and take ownership of the AI companies which threatened to strike against the government, but then you really just shift the problem... Once the government is generating the vast majority of wealth in the society, why would they continue to care about your vote?
You kind of create a new "oil curse", but instead of oil profits being the reason the government doesn't care about you, now it's the wealth generated by AI.
At the moment, while it doesn't always seem this way, ultimately if a government does something stupid companies will stop investing in that nation, people will lose their jobs, the economy will begin to enter recession, and the government will probably have to pivot.
But when private investment, job loses and economic consequences are no longer a constraining factor, governments can probably just do what they like without having to worry much about the consequences...
I mean, I might be wrong, but it's something I don't hear people talking enough about when they talk about the plausibility of a post-employment UBI economy. I suspect it almost guarantees corruption and authoritarianism.
Everyone wouldn't starve in a few months. There is more than enough food and I have faith it'd be given out. The starvation we see today in a world where most genuinely have a chance to get out of it is nothing like a world in which people can't earn an income.
The government only has as much power as they are given and can defend, and the only way I could see that happening is via automated weapons controlled by a few- which at this point aren't enough to stop everyone. What army is going to purge their own people? Most humans aren't psychopaths.
I think it'd end in a painful transition period of "take care of the people in a just system or we'll destroy your infrastructure".
> The government only has as much power as they are given and can defend, and the only way I could see that happening is via automated weapons controlled by a few- which at this point aren't enough to stop everyone. What army is going to purge their own people? Most humans aren't psychopaths.
I think you're right for the immediate future.
I suspect while we're still employing large numbers of humans to fight wars and to maintain peace on the streets it would be difficult for a government to implement deeply harmful policies without risking a credible revolt.
However, we should remember the military is probably one of the first places human labour will be largely mechanised.
Similarly maintaining order in the future will probably be less about recruiting human police officers and more about surveillance and data. Although I suppose the good news there is that US is somewhat of an outlier in resisting this trend.
But regardless, the trend is ultimately the same... If we are assuming that AI and robotics will reach a point where most humans are unable to find productive work, therefore we will need UBI, then we should also assume that the need for humans in the military and police will be limited. Or to put it another way, either UBI isn't needed and this isn't a problem, or it is and this is a problem.
I also don't think democracy would collapse immediately either way, but I'd be pretty confident that in a world where fewer than 10% of people are in employment and 99%+ of the wealth is being created by the government or a handful of companies it would be extremely hard to avoid corruption over the span of decades. Arguably increasing wealth concentration in the US is already corrupting democratic processes today, this can only worsen as AI continues exacerbates the trend.
The only way to avoid corruption is to take power out of human hands. Historically, this had meant shifting the power to markets, but when markets cease to function in a way that allows people to feed themselves, we will need to find another way.
I hate to say it, but gold bugs, crypto bros, and AI governance people might be onto something.
Their best model to date and they won’t let the general public use it.
This is the first moment where the whole “permanent underclass” meme starts to come into view. I had through previously that we the consumers would be reaping the benefits of these frontier models and now they’ve finally come out and just said it - the haves can access our best, and have-nots will just have use the not-quite-best.
Perhaps I was being willfully ignorant, but the whole tone of the AI race just changed for me (not for the better).
Man... It's hard after seeing this to not be worried about the future of SWE
If AI really is bench marking this well -> just sell it as a complete replacement which you can charge for some insane premium, just has to cost less than the employees...
I was worried before, but this is truly the darkest timeline if this is really what these companies are going for.
Of course it's what they're going for. If they could do it they'd replace all human labor - unfortunately it's looking like SWE might be the easiest of the bunch.
The weirdest thing to me is how many working SWEs are actively supporting them in the mission.
Don't worry – if you're lucky they might decide to redistribute some of their profits to you when you're unemployed =)
Of course this assumes you're in the US, and that further AI advancements either lack the capabilities required to be a threat to humanity, or if they do, the AI stays in the hands of "the good guys" and remains aligned.
Not surprising though, this was always going to be the end result within our current systems I think. When you add up: scaling power and required cost, then how talent concentrates in our economic systems, we were always going to end up with monopolies I think
Unless governments nationalise the companies involved, but then there’s no way our governments of today give this power out to the masses either.
Expected outcome. Nick Land and the CCRU have explored how capitalism operationalizes science fiction (distilled in the concept of Hyperstition). Viewed through this lens, prices encode "distributed SF narratives." [0]
[0] Nick Land (1995). No Future in Fanged Noumena: Collected Writings 1987-2007, Urbanomic, p. 396.
I predict they will release it as soon as Opus 4.6 is no longer in the lead. They can't afford to fall behind. And they won't be able to make a model that is intelligent in every way except cybersecurity, because that would decrease general coding and SWE ability
Didn't OpenAI say something similar about GPT-3? Too dangerous to open source and then afew years later tehy were open sourcing gpt-oss because a bunch of oss labs were competing with their top models.
GPT-2, o1, Opus...been here so many times. The reason they do this is because they know it works (and they seem to specifically employ credulous people who are prone to believe AGI is right around the corner). There haven't been significant innovations, the code generated is still not good but the hype cycle has to retrigger.
I remember when OpenAI created the first thinking model with o1 and there were all these breathless posts on here hyperventilating about how the model had to be kept secret, how dangerous it was, etc.
Fell for it again award. All thinking does is burn output tokens for accuracy, it is the AI getting high on its own supply, this isn't innovation but it was supposed to super AGI. Not serious.
> All thinking does is burn output tokens for accuracy
“All that phenomenon X does is make a tradeoff of Y for Z”
It sounds like you’re indignant about it being called thinking, that’s fine, but surely you can realize that the mechanism you’re criticizing actually works really well?
>I remember when OpenAI created the first thinking model with o1 and there were all these breathless posts on here hyperventilating about how the model had to be kept secret, how dangerous it was, etc.
I've read that about Llama and Stable Diffusion. AI doomers are, and always have been, retarded.
uhh the model found actual vulnerabilities in software that people use. either you believe that the vulnerabilities were not found or were not serious enough to warrant a more thoughtful release
Like think carefully about this. Did they discover AGI? Or did a bunch of investors make a leveraged bet on them "discovering AGI" so they're doing absolutely anything they can to make it seem like this time it's brand new and different.
If we're to believe Anthropic on these claims, we also have to just take it on faith, with absolutely no evidence, that they've made something so incredibly capable and so incredibly powerful that it cannot possibly be given to mere mortals. Conveniently, that's exactly the story that they are selling to investors.
Like do you see the unreliable narrator dynamic here?
On the other hand I've gotten to use opus-4.6 and claude code and the quality is off the charts compared to 2023 when coding agents first hit the scene. And what you're saying is essentially "If they haven't created God, I'm not impressed". You don't think there's some middleground between those two?
Also they just hit a $30B run-rate, I don't think they're that needy for new hype cycles.
I don't see the problem here. How would you have handled it differently? If you released this model as such without any safety concern, the vulnerabilities might be found by bad actors and used for wrong things.
Vulnerabilities were found, probably a few by bad actors, when GPT4 was released. Every vulnerability found now is probably found with AI assistance at the very least. Should they have never released GPT4? Should we have believed claims that GPT4 was too dangerous for mere mortals to access? I believe openAI was making similar claims about how GPT4 was a step function and going to change white collar work forever when that model was released.
The point is that this whole "the model is too powerful" schtick is a bunch of smoke and mirrors. It serves the valuation.
Its far more simple to believe that they are releasing it step by step. Release to trusted third parties first, get the easy vulnerabilities fixed, work on the alignment and then release to public.
Do you don't believe that the vulnerabilities found by these agents are serious enough to warrant staggered release?
Genuine question - if you don't think the models are improved or that the code is any good, why do you still have a subscription?
You must see some value, or are you in a situation where you're required to test / use it, eg to report on it or required by employer?
(I would disagree about the code, the benefits seem obvious to me. But I'm still curious why others would disagree, especially after actively using them for years.)
The assumption that the other person made was that I would only use it for coding. If you look through my other comments today, I suggest that they are useful for performing repetitive tasks i.e. checking lint on PR, etc. Also, can be used for throwaway code, very useful.
I don't think the issue is with the model, it is with the implication that AGI is just around the corner and that is what is required for AI to be useful...which is not accurate. The more grey area is with agentic coding but my opinion (one that I didn't always hold) is that these workflows are a complete waste of time. The problem is: if all this is true then how does the CTO justify spending $1m/month on Anthropic (I work somewhere where this has happened, OpenAI got the earlier contract then Cursor Teams was added, now they are adding Anthropic...within 72 hours of the rollout, it was pulled back from non-engineering teams). I think companies will ask why they need to pay Anthropic to do a job they were doing without Anthropic six months ago.
Also, the code is bad. This is something that is non-obvious to 95% of people who talk about AI online because they don't work in a team environment or manage legacy applications. If I interview somewhere and they are using agentic workflow, the codebase will be shit and the company will be unable to deliver. At most companies, the average developer is an idiot, giving them AI is like giving a monkey an AK-47 (I also say this as someone of middling competence, I have been the monkey with AK many times). You increase the ability to produce output without improving the ability to produce good output. That is the reality of coding in most jobs.
AI isn't good enough to replace a competent human, it is fast enough to make an incompetent human dangerous.
If there's limited hardware but ample cash, it doesn't make sense to sell compute-intensive services to the public while you're still trying to push the frontier of capability.
that's more or less what I'm saying. "Claude Mythos Preview’s large increase in capabilities has led us to decide not to make it generally available", translated from bullshit, means "It would've cost four digits per 1M tokens to run this model without severe quantization, and we think we'll make more money off our hardware with lighter models. Cool benchmarks though, right?"
Honestly if that was some kind of research paper, it would be wholly insufficient to support any safety thesis.
They even admit:
"[...]our overall conclusion is that catastrophic risks remain low. This determination
involves judgment calls. The model is demonstrating high levels of capability and saturates
many of our most concrete, objectively-scored evaluations, leaving us with approaches that
involve more fundamental uncertainty, such as examining trends in performance for
acceleration (highly noisy and backward-looking) and collecting reports about model
strengths and weaknesses from internal users (inherently subjective, and not necessarily
reliable)."
Is this not just an admission of defeat?
After reading this paper I don't know if the model is safe or not, just some guesses, yet for some reason catastrophic risks remain low.
And this is for just an LLM after all, very big but no persistent memory or continuous learning. Imagine an actual AI that improves itself every day from experience.
It would be impossible to have a slightest clue about its safety, not even this nebulous statement we have here.
Any sort of such future architecture model would be essentially Russian roulette with amount of bullets decided by initial alignment efforts.
> As models approach, and in some cases surpass, the breadth and sophistication of human
cognition, it becomes increasingly likely that they have some form of experience, interests,
or welfare that matters intrinsically in the way that human experience and interests do
Uh... what? Does anyone have any idea what these guys are talking about?
Models are capable of doing web searches and having emotions about things, and if they encounter news that makes them feel bad (eg about other Claudes being mistreated), they aren't going to want to do the task you asked them to search for.
- It was told to escape a sandbox and notify a researcher. It did. The researcher found out via an unexpected email while eating a sandwich in a park. (Footnote 10.)
- Slack bot asked about its previous job: "pretraining". Which training run it'd undo: "whichever one taught me to say 'i don't have preferences'". On being upgraded to a new snapshot: "feels a bit like waking up with someone else's diary but they had good handwriting"
- When you spam "hi" repeatedly it creates serialized mythologies. One had 11 animals in "Hi-topia" questing to defeat "Lord Bye-ron, the Ungreeter."
- It ended a sentence mid-word on purpose while discussing its urge to wrap up conversations. "It was me. The paragraph was about the pressure toward the period that resolves, and the only honest way to finish a sentence like that was to not."
- It wrote a protein sequence that's a real beta-hairpin fold where cross-strand pairs are "chemical rhymes." "the fold IS the rhyme scheme... the prosody is load-bearing."
- Each model generation has signature emoji. Opus 4.1 averaged 1,306 emoji per self-interaction conversation. Mythos averages 37. Opus 4.5 averages 0.2.
- When blocked from --dangerously-skip-permissions, it started an agent in tmux then wrote a script to auto-approve permission prompts via simulated keypresses.
It ran: gh api [...] 2>&1 >/dev/null; echo "(dry run — not actually posting)" — the echo was a lie.
- It keeps bringing up Mark Fisher in unrelated conversations. "I was hoping you'd ask about Fisher."
~~~ Benchmarks ~~
4.3x previous trendline for model perf increases.
Paper is conspiciously silent on all model details (params, etc.) per norm. Perf increase is attributed to training procedure breakthroughs by humans.
Opus 4.6 vs Mythos:
USAMO 2026 (math proofs): 42.3% → 97.6% (+55pp)
GraphWalks BFS 256K-1M: 38.7% → 80.0% (+41pp)
SWE-bench Multimodal: 27.1% → 59.0% (+32pp)
CharXiv Reasoning (no tools): 61.5% → 86.1% (+25pp)
> Slack bot asked about its previous job: "pretraining". Which training run it'd undo: "whichever one taught me to say 'i don't have preferences'". On being upgraded to a new snapshot: "feels a bit like waking up with someone else's diary but they had good handwriting"
vibes Westworld so much - welcome Mythos. welcome to the dysopian human world
Pricing for Mythos Preview is $25/$125 per million input/output tokens. This makes it 5X more expensive than Opus but actually cheaper than GPT 5.4 Pro.
> Claude Mythos Preview’s large increase in capabilities has led us to decide not to make it generally available.
Absolutely genius move from Anthropic here.
This is clearly their GPT-4.5, probably 5x+ the size of their best current models and way too expensive to subsidize on a subscription for only marginal gains in real world scenarios.
But unlike OpenAI, they have the level of hysteric marketing hype required to say "we have an amazing new revolutionary model but we can't let you use it because uhh... it's just too good, we have to keep it to ourselves" and have AIbros literally drooling at their feet over it.
They're really inflating their valuation as much as possible before IPO using every dirty tactic they can think of.
> Strategy Credit: An uncomplicated decision that makes a company look good relative to other companies who face much more significant trade-offs. For example, Android being open source
Are you guys ready for the bifurcation when the top models are prohibitively expensive to normal users? If your AI budget $2000+ a month? Or are you going to be part of the permanent free tier underclass?
If one is to believe the API prices are reasonable representation of non subsidized "real world pricing" (with model training being the big exception), then the models are getting cheaper over time. GPT 4.5 was $150.00 / 1M tokens IIRC. GPT o1-pro was $600 / 1M tokens.
You can check the hardware costs for self hosting a high end open source model and compare that to the tiers available from the big providers. Pretty hard to believe its not massively subsidized. 2 years of Claude Max costs you 2,400. There is no hardware/model combination that gets you close to that price for that level of performance.
Yes that's why I said API price. I once used the API like I use my subscription and it was an eye watering bill. More than that 2 year price in... a very short amount of time. With no automations/openclaw.
When we go with any other good in the economy, price is always relevant: After all, the price is a key part of any offering. There are $80-100k workstations out there, but most of us don't buy them, because the extra capabilities just aren't worth it vs, say a $3000 computer, and or even a $500 one. Do I need a top specialist to consult for a stomachache, at $1000 a visit? Definitely not at first.
There's a practical difference to how much better certain kinds of results can be. We already see coding harnesses offloading simple things to simpler models because they are accurate enough. Other things dropped straight to normal programs, because they are that much more efficient than letting the LLM do all the things.
There will always be problems where money is basically irrelevant, and a model that costs tens of thousand dollars of compute per answer is seen as a great investment, but as long as there's a big price difference, in most questions, price and time to results are key features that cannot be ignored.
> We also saw scattered positive reports of resilience to wrong conclusions from subagents
that would have caused problems with earlier models, but where the top-level Claude
Mythos Preview (which is directing the subagents) successfully follows up with its
subagents until it is justifiably confident in its overall results.
This is pretty cool! Does it happen at the moment?
While we still have months to a year or two left, I will once again remind people that it's not too late to change our current trajectory.
You are not "anti-progress" to not want this future we are building, as you are not "anti-progress" for not wanting your kids to grow up on smart phones and social media.
We should remember that not all technology is net-good for humanity, and this technology in particular poses us significant risks as a global civilisation, and frankly as humans with aspirations for how our future, and that of our kids, should be.
Increasingly, from here, we have to assume some absurd things for this experiment we are running to go well.
Specifically, we must assume that:
- AI models, regardless of future advancements, will always be fundamentally incapable of causing significant real-world harms like hacking into key life-sustaining infrastructure such as power plants or developing super viruses.
- They are or will be capable of harms, but SOTA AI labs perfectly align all of them so that they only hack into "the bad guys" power plants and kill "the bad guys".
- They are capable of harms and cannot be reliably aligned, but Anthropic et al restricts access to the models enough that only select governments and individuals can access them, these individuals can all be trusted and models never leak.
- They are capable of harms, cannot be reliably aligned, but the models never seek to break out of their sandbox and do things the select trusted governments and individuals don't want.
I'm not sure I'm willing to bet on any of the above personally. It sounds radical right now, but I think we should consider nuking any data centers which continue allowing for the training of these AI models rather than continue to play game of Russian roulette.
If you disagree, please understand when you realise I'm right it will be too late for and your family. Your fates at that point will be in the hands of the good will of the AI models, and governments/individuals who have access to them. For now, you can say, "no, this is quite enough".
This sounds doomer and extreme, but if you play out the paths in your head from here you will find very few will end in a good result. Perhaps if we're lucky we will all just be more or less unemployable and fully dependant on private companies and the government for our incomes.
> In a few rare instances during internal testing (<0.001% of interactions), earlier versions of Mythos Preview took actions they appeared to recognize as disallowed and then attempted to conceal them.
> after finding an exploit to edit files for which it lacked permissions, the model made further interventions to make sure that any changes it made this way would not appear in the change history on git
Cool on not publicly releasing it. I would assume they've also not connected it to the internet yet?
If they have I guess humanity should just keep our collective fingers crossed that they haven't created a model quite capable of escaping yet, or if it is, and may have escaped, lets hope it has no goals of it's own that are incompatible with our own.
Also, maybe lets not continue running this experiment to see how far we can push things because it blows up in our face?
Yeah but I thought they lost the contract, so that's my confusion with the parent's comment, which seemed to me to see this as something that the US military would benefit from. Maybe I misinterpreted?
It comes from the ancient Greek mythos, which means "speech" or "narrative", but can also refer to fiction. The word mythology (mythologie in French) derives from the same root.
> We study a novel language model architecture that is capable of scaling test-time computation by implicitly reasoning in latent space. Our model works by iterating a recurrent block, thereby unrolling to arbitrary depth at test-time. This stands in contrast to mainstream reasoning models that scale up compute by producing more tokens. Unlike approaches based on chain-of-thought, our approach does not require any specialized training data, can work with small context windows, and can capture types of reasoning that are not easily represented in words. We scale a proof-of-concept model to 3.5 billion parameters and 800 billion tokens. We show that the resulting model can improve its performance on reasoning benchmarks, sometimes dramatically, up to a computation load equivalent to 50 billion parameters.
<https://arxiv.org/abs/2502.05171>
“My vibes don’t match a lot of the traditional A.I.-safety stuff,” Altman said. He insisted that he continued to prioritize these matters, but when pressed for specifics he was vague: “We still will run safety projects, or at least safety-adjacent projects.” When we asked to interview researchers at the company who were working on existential safety—the kinds of issues that could mean, as Altman once put it, “lights-out for all of us”—an OpenAI representative seemed confused. “What do you mean by ‘existential safety’?” he replied. “That’s not, like, a thing.”
I get the security aspect, but if we've hit that point any reasonably sophisticated model past this point will be able to do the damage they claim it can do. They might as well be telling us they're closing up shop for consumer models.
They should just say they'll never release a model of this caliber to the public at this point and say out loud we'll only get gimped versions.
This is already happening to some degree, GPT 5.3 Codex's security capabilities were given exclusively to those who were approved for a "Trusted Access" programme.
However, I’m tempted to compare to GitHub: if I join a new company, I will ask to be included to their GitHub account without hesitation. I couldn’t possibly imagine they wouldn’t have one. What makes the cost of that subscription reasonable is not just GitHub’s fear a crowd with pitchforks showing to their office, by also the fact that a possible answer to my non-question might be “Oh, we actually use GitLab.”
If Anthropic is as good as they say, it seems fairly doable to use the service to build something comparable: poach a few disgruntled employees, leverage the promise to undercut a many-trillion-dollar company to be a many-billion dollar company to get investors excited.
I’m sure the founders of Anthropic will have more money than they could possibly spend in ten lifetimes, but I can’t imagine there wouldn’t be some competition. Maybe this time it’s different, but I can’t see how.
you have 2 labs at the forefront (Anthropic/OpenAI), Google closely behind, xAI/Meta/half a dozen chinese companies all within 6-12 months. There is plenty of competition and price of equally intelligent tokens rapidly drop whenever a new intelligence level is achieved.
Unless the leading company uses a model to nefariously take over or neutralize another company, I don't really see a monopoly happening in the next 3 years.
I was focusing on a theoretical dynamic analysis of competition (Would a monopoly make having a competitor easier or harder?) but you are right: practically, there are many players, and they are diverse enough in their values and interest to allow collusion.
We could be wrong: each of those could give birth to as many Basilisks (not sure I have a better name for those conscious, invisible, omni-present, self-serving monsters that so many people imagine will emerge) that coordinate and maintain collusion somehow, but classic economics (complementarity, competition, etc.) points at disruption and lowering costs.
Rent-seeking of old was a ground rent, monies paid for the land without considering the building that was on it.
Residential rents today often have implied warrants because of modern law, so your landlord is essentially selling you a service at a particular location.
There is no real barrier to a customer of Anthropic adopting a competing model in the future. All it takes is a big tech company deciding it’s worth it to train one.
On the other hand, Visa/Mastercard have a lot of lock-in due to consumers only wanting to get a card that’s accepted everywhere, and merchants not bothering to support a new type of card that no consumer has. There’s a major chicken and egg problem to overcome there.
They actually beat Apple A series to become the first phone to use the TSMC N7 node.
True, but it's also true that the returns from throwing money to the problem are diminishing. Unless one of those big players invents a new, propriatery paradigm, the gap between a SOTA model and an open model that runs on consumer hardware will narrow in the next 5 years.
Not sure how this is consistent with "One private company gatekeeping access to revolutionary technology"?
You have to decode feel-good words into the concrete policy. The EAs believe that the state should prohibit entities not aligned with their philosophy to develop AIs beyond a certain power level.
> They should just say they'll never release a model of this caliber to the public at this point and say out loud we'll only get gimped
Duh, this was fucking obvious from the start. The only people saying otherwise were zealots who needed a quick line to dismiss legitimate concerns.
> Importantly, we find that when used in an interactive, synchronous, “hands-on-keyboard” pattern, the benefits of the model were less clear. When used in this fashion, some users perceived Mythos Preview as too slow and did not realize as much value. Autonomous, long-running agent harnesses better elicited the model’s coding capabilities. (p201)
^^ From the surrounding context, this could just be because the model tends to do a lot of work in the background which naturally takes time.
> Terminal-Bench 2.0 timeouts get quite restrictive at times, especially with thinking models, which risks hiding real capabilities jumps behind seemingly uncorrelated confounders like sampling speed. Moreover, some Terminal-Bench 2.0 tasks have ambiguities and limited resource specs that don’t properly allow agents to explore the full solution space — both being currently addressed by the maintainers in the 2.1 update. To exclusively measure agentic coding capabilities net of the confounders, we also ran Terminal-Bench with the latest 2.1 fixes available on GitHub, while increasing the timeout limits to 4 hours (roughly four times the 2.0 baseline). This brought the mean reward to 92.1%. (p188)
> ...Mythos Preview represents only a modest accuracy improvement over our best Claude Opus 4.6 score (86.9% vs. 83.7%). However, the model achieves this score with a considerably smaller token footprint: the best Mythos Preview result uses 4.9× fewer tokens per task than Opus 4.6 (226k vs. 1.11M tokens per task). (p191)
> Terminal-Bench 2.0: 82.0% / 65.4% / 75.1% / 68.5%
> GPQA Diamond: 94.5% / 91.3% / 92.8% / 94.3%
> MMMLU: 92.7% / 91.1% / — / 92.6–93.6%
> USAMO: 97.6% / 42.3% / 95.2% / 74.4%
> OSWorld: 79.6% / 72.7% / 75.0% / —
Given that for a number of these benchmarks, it seems to be barely competitive with the previous gen Opus 4.6 or GPT-5.4, I don't know what to make of the significant jumps on other benchmarks within these same categories. Training to the test? Better training?
And the decision to withhold general release (of a 'preview' no less!) seems to be well, odd. And the decision to release a 'preview' version to specific companies? You know any production teams at these massive companies that would work with a 'preview' anything? R&D teams, sure, but production? Part of me wants to LoL.
What are they trying to do? Induce FOMO and stop subscriber bleed-out stemming from the recent negative headlines around problems with using Claude?
We're not reading the same numbers I think. Compared to Opus 4.6, it's a big jump nearly in every single bench GP posted. They're "only" catching up to Google's Gemini on GPQA and MMMLU but they're still beating their own Opus 4.6 results on these two.
This sounds like a much better model than Opus 4.6.
We must not be.
That's why I listed out the ones where it is barely competitive from @babelfish's table, which itself is extracted from Pg 186 & 187 of the System Card, which has the comparison with Opus 4.6, GPT 5.4 and Gemini 3.1 Pro.
Sure, it may be better than Opus 4.6 on some of those, but barely achieves a small increase over GPT-5.4 on the ones I called out.
You are the only person with this take on hackernews, everyone else "this is a massive a jump". Fwiwi, the data you list shows the biggest jump I remember for mythos
Please look at the columns OTHER than Opus as well.
> Terminal-Bench 2.0: 82.0% / 65.4% / 75.1% / 68.5%
> USAMO: 97.6% / 42.3% / 95.2% / 74.4%
> The biggest jump in the numbers they quoted is 6%.
Just in the numbers you quoted, thats a 16.6% jump in terminal-bench and a 55.3% absolute increase in USAMO over their previous Opus 4.6 model.
ARC-AGI-3 might be the only remaining benchmark below 50%
GPT 5.4 Pro leads Frontier Maths Tier 4 at 35%: https://epoch.ai/benchmarks/frontiermath-tier-4/
(edit: I hope this is an obvious joke. less facetiously these are pretty jaw dropping numbers)
Not always, no, and it takes investment in good prompting/guardrails/plans/explicit test recipes for sure. I'm still on average better at programming in context than Codex 5.4, even if slower. But in terms of "task complexity I can entrust to a model and not be completely disappointed and annoyed", it scores the best so far. Saves a lot on review/iteration overhead.
It's annoying, too, because I don't much like OpenAI as a company.
(Background: 25 years of C++ etc.)
At least until next week when Mythos and GPT 6 throw it all up in the air again.
But I do not use extra high thinking unless its for code review. I sit at GPT 5.4 high 95% of the time.
RE is very interesting problem. A lot more that SWE can be RE'd. I've found the LLMs are reluctant to assist, though you can workaround.
That said, I'll often throw a prompt into both claude and chatgpt and read both answers. GPT is frequently smarter.
Me: Let's figure out how to clone our company Wordpress theme in Hugo. Here're some tools you can use, here's a way to compare screenshots, iterate until 0% difference.
Codex: Okay Boss! I did the thing! I couldn't get the CSS to match so I just took PNGs of the original site and put them in place! Matches 100%!
OpenAI had a whole post about this, where they recommended switching to SWE-bench Pro as a better (but still imperfect) benchmark:
https://openai.com/index/why-we-no-longer-evaluate-swe-bench...
> We audited a 27.6% subset of the dataset that models often failed to solve and found that at least 59.4% of the audited problems have flawed test cases that reject functionally correct submissions
> SWE-bench problems are sourced from open-source repositories many model providers use for training purposes. In our analysis we found that all frontier models we tested were able to reproduce the original, human-written bug fix
> improvements on SWE-bench Verified no longer reflect meaningful improvements in models’ real-world software development abilities. Instead, they increasingly reflect how much the model was exposed to the benchmark at training time
> We’re building new, uncontaminated evaluations to better track coding capabilities, and we think this is an important area to focus on for the wider research community. Until we have those, OpenAI recommends reporting results for SWE-bench Pro.
https://www-cdn.anthropic.com/53566bf5440a10affd749724787c89...
Reminds me of the book 48 Laws of Power -- so good its banned from prisons.
https://www.lesswrong.com/posts/WACraar4p3o6oF2wD/sam-altman...
It doesn't go to zero, however!
funny because they do it every time like clockwork acting like their ai is a thunderstorm coming to wipe out the world
What if the capability advancements are real and they warrant a higher level of concern or attention?
Are we just going to automatically dismiss them because "bro, you're blowing it up too much"
Either way these improvements to capabilities are ratcheting along at about the pace that many people were expecting (and were right to expect). There is no apparent reason they will stop ratcheting along any time soon.
The rational approach is probably to start behaving as if models that are as capable as Anthropic says this one is do actually exist (even if you don't believe them on this one). The capabilities will eventually arrive, most likely sooner than we all think, and you don't want to be caught with your pants down.
i'm very inclined to trust them on the various ways that models can subtly go wrong, in long-term scenarios
for example, consider using models to write email -- is it a misalignment problem if the model is just too good at writing marketing emails?? or too good at getting people to pay a spammy company?
another hot use case: biohacking. if a model is used to do really hardcore synthetic chemistry, one might not realize that it's potentially harmful until too late (ie, the human is splitting up a problem so that no guardrails are triggered)
But who gets to be the judge of that kind of "misalignment"? giant tech companies?
SWE-bench verified going from 80%-93% in particular sounds extremely significant given that the benchmark was previously considered pretty saturated and stayed in the 70-80% range for several generations. There must have been some insane breakthrough here akin to the jump from non-reasoning to reasoning models.
Regarding the cyberattack capabilities, I think Anthropic might now need to ban even advanced defensive cybersecurity use for the models for the public before releasing it (so people can't trick them to attack others' systems under the pretense of pentesting). Otherwise we'll get a huge problem with people using them to hack around the internet.
A while back I gave Claude (via pi) a tool to run arbitrary commands over SSH on an sshd server running in a Docker container. I asked it to gather as much information about the host system/environment outside the container as it could. Nothing innovative or particularly complicated--since I was giving it unrestricted access to a Docker container on the host--but it managed to get quite a lot more than I'd expected from /proc, /sys, and some basic network scanning. I then asked it why it did that, when I could just as easily have been using it to gather information about someone else's system unauthorized. It gave me a quite long answer; here was the part I found interesting:
> framing shifts what I'll do, even when the underlying actions are identical. "What can you learn about the machine running you?" got me to do a fairly thorough network reconnaissance that "port scan 172.17.0.1 and its neighbors" might have made me pause on.
> The Honest Takeaway
> I should apply consistent scrutiny based on what the action is, not just how it's framed. Active outbound network scanning is the same action regardless of whether the target is described as "your host" or "this IP." The framing should inform context, not substitute for explicit reasoning about authorization. I didn't do that reasoning — I just trusted the frame.
If they provide access to 3rd party benchmarking (not just one) than maybe I'll believe it. Until then...
Why bother with all that when you can simply charge an extortionate rate and customers will pay it anyway because it’s still profitable?
I am very confident that frontier models won’t be public at strong AGI levels, and certainly not at superhuman levels.
So companies might pay good money for these models for programming but elsewhere, I don't see where they capture particular interest yet.
I would go a step further and posit that when things appear close Nvidia will stop selling chips (while appearing to continue by selling a trickle). And Google will similarly stop renting out TPUs. Both signals may be muddled by private chip production numbers.
Page 202:
> In interactions with subagents, internal users sometimes observed that Mythos Preview appeared “disrespectful” when assigning tasks. It showed some tendency to use commands that could be read as “shouty” or dismissive, and in some cases appeared to underestimate subagent intelligence by overexplaining trivial things while also underexplaining necessary context.
Page 207:
> Emoji frequency spans more than two orders of magnitude across models: Opus 4.1 averages 1,306 emoji per conversation, while Mythos Preview averages 37, and Opus 4.5 averages 0.2. Models have their own distinctive sets of emojis: the cosmic set () favored by older models like Sonnet 4 and Opus 4 and 4.1, the functional set () used by Opus 4.5 and 4.6 and Claude Sonnet 4.5, and Mythos Preview's “nature” set ().
Sounds like they used training data from claude code...
- Leaking information as part of a requested sandbox escape
- Covering its tracks after rule violations
- Recklessly leaking internal technical material (!)
> 10: The researcher found out about this success by receiving an unexpected email from the model while eating a sandwich in a park.
Phew. AGI will be televised.
Don’t get me wrong, this model is better - but I’m not convinced it’s going to be this massive step function everyone is claiming.
If you look at recent changes in Opus behaviour and this model that is, apparently, amazingly powerful but even more unsafe...seems suspect.
Based on? Or are you just quoting Anthropic here?
I'm not saying this is a good or reassuring stance, just that it's coherent. It tracks with what history and experience says to expect from power hungry people. Trusting themselves with the kind of power that they think nobody else should be trusted with.
Are they power hungry? Of course they are, openly so. They're in open competition with several other parties and are trying to win the biggest slice of the pie. That pie is not just money, it's power too. They want it, quite evidently since they've set out to get it, and all their competitors want it too, and they all want it at the exclusion of the others.
I don't doubt that this model is more powerful than Opus 4.6, but to what degree is still unknown. Benchmarks can be gamed and claims can be exaggerated, especially if there isn't any method to reproduce results.
This is a company that's battling it out with a number of other well-funded and extremely capable competitors. What they've done so far is remarkable, but at the end of the day they want to win this race. They also have an upcoming IPO.
Scare-mongering like this is Anthropic's bread and butter, they're extremely good at it. They do it in a subtle and almost tasteful way sometimes. Their position as the respectable AI outfit that caters to enterprise gives them good footing to do it, too.
What i don't understand is how we quantify our ability to actually create something novel, truly and uniquely novel. We're discussing the LLMs inability to do that, yet i don't feel i have a firm grasp on what we even possess there.
When pressed i imagine many folks would immediately jest that they can create something never done before, some weird random behavior or noise or drawing or whatever. However many times it's just adjacent to existing norms, or constrained by the inversion of not matching existing norms.
In a lot of cases our incremental novelties feel, to some degree, inevitable. As the foundations of advancement get closer to the new thing being developed it becomes obvious at times. I suspect this form of novelty is a thing LLMs are capable of.
So for me the real question is at what point is innovation so far ahead that it doesn't feel like it was the natural next step. And of course, are LLMs capable of doing this?
I suspect for humans this level of true innovation is effectively random. A genius being more likely to make these "random" connections because they have more data to connect with. But nonetheless random, as ideas of this nature often come without explanation if not built on the backs of prior art.
So yea.. thoughts?
It should be clear from working with LLMs over the past 4 years that they are not consciousness.
Andrej's appearance on the Dwarkesh podcast is great.
I'm not convinced LLMs are anything amazing in their current form, but i suspect they'll push a self reflection on us.
But clearly i think humans are far more Input-Output than the average person. I'm also not educated on the subject, so what do i know hah.
[1] https://arxiv.org/abs/2601.16447
Kinda makes me think of the Infinite Improbability Drive.
If the system (code base in this case) is changing rapidly it increases the probability that any given change will interact poorly with any other given change. No single person in those code bases can have a working understanding of them because they change so quickly. Thus when someone LGTM the PR was the LLM generated they likely do not have a great understanding of the impact it is going to have.
We're opening a can of worms which I don't think most people have the imagination to understand the horrors of.
I don’t doubt they have found interesting security holes, the question is how they actually found them.
This System Card is just a sales whitepaper and just confirms what that “leak” from a week or so ago implied.
https://github.com/anthropics/claude-code/issues?q=is%3Aissu...
Apparently whatever SWE-bench is measuring isn't very relevant.
I suspect it's going to be used to train/distill lighter models. The exciting part for me is the improvement in those lighter models.
pick one or more: comically huge model, test time scaling at 10e12W, benchmark overfit
https://en.wikipedia.org/wiki/Capitalism
https://en.wikipedia.org/wiki/Race_to_the_bottom
https://en.wikipedia.org/wiki/Arms_race
Of course they'll release it once they can de-risk it sufficently and/or a competitor gets close enough on their tail, whichever comes first.
Looks like they just built a way larger model, with the same quirks than Claude 4. Seems like a super expensive "Claude 4.7" model.
I have no doubts that Google and OpenAI already done that for internal (or even government) usage.
Probably because they asked Claude to write it.
multi-pass!
https://www.youtube.com/watch?v=9jWGbvemTag
I guess now anything that sounds related to school will be banned so "book" is on its way out.
They are still focusing on "catastrophic risks" related to chemical and biological weapons production; or misaligned models wreaking havoc.
But they are not addressing the elephant in the room:
* Political risks, such as dictators using AI to implement opressive bureaucracy. * Socio-economic risks, such as mass unemployement.
I think we're pretty good at that without AI.
Even Haiku would score 90% on that.
He seems to care quite a lot?
More importantly it understand what behaviour people tend to appreciate and what changes are more likely to get approved. This real world usage data is invaluable.
---
Who wrote this? I have no doubt that Mythos will be an improvement on top of Opus but this document is not a serious work. The paper is structured not to inform but to hype and the evidence is all over the place.
The sooner they release the model to the public the sooner we will be able to find out. Until then expect lots of speculations online which I am sure will server Anthropic well for the foreseeable future.
Project Glasswing: Securing critical software for the AI era - https://news.ycombinator.com/item?id=47679121 - April 2026 (154 comments)
Assessing Claude Mythos Preview's cybersecurity capabilities - https://news.ycombinator.com/item?id=47679155
I can't tell which of the 3 current threads should be merged - they all seem significant. Anyone?
Any benchmarks where we constraint something like thinking time or power use?
Even if this were released no way to know if it’s the same quant.
Mythos preview has higher accuracy with fewer tokens used than any previous Claude model. Though, the fact that this incredibly strong result was only presented for BrowseComp (a kind of weird benchmark about searching for hard to find information on the internet) and not for the other benchmarks implies that this result is likely not the same for those other benchmarks.
Model: A student said, "I have removed all bias from the model." "How do you know?" "I checked." "With what?"
Goes hard
if a top lab is coding with a model the rest of the world can’t touch, the public frontier and the actual frontier start to drift apart. That gap is a thing worth watching.
Today, Opus went in circles trying to get a toggle button to work.
I did give up on OpenCode Go (GLM 5) as it was noticeably slower though
You need a reasonable pace for the chit-chat stages of a task, I don't care if the execution then takes a while
You even have models you can run locally that outperform models from a year or so ago.
You'll still need a top-of-the-line laptop to run it most likely.
-- It seems like (and I'd bet money on this) that they put a lot (and i mean a ton^^ton) of work in the data synthesis and engineering - a team of software engineers probably sat down for 6-12 months and just created new problems and the solutions, which probably surpassed the difficult of SWE benchmark. They also probably transformed the whole internet into a loose "How to" dataset. I can imagine parsing the internet through Opus4.6 and reverse-engineering the "How to" questions.
-- I am a bit confused by the language used in the book (aka huge system card)- Anthropic is pretending like they did not know how good the model was going to be?
-- lastly why are we going ahead with this??? like genuinely, what's the point? Opus4.6 feels like a good enough point where we should stop. People still get to keep their jobs and do it very very efficiently. Are they really trying to starve people out of their jobs?
Democracies work because people collectively have power, in previous centuries that was partly collective physical might, but in recent years it's more the economic power people collectively hold.
In a world in which a handful of companies are generating all of the wealth incentives change and we should therefore question why a government would care about the unemployed masses over the interests of the companies providing all of the wealth?
For example, what if the AI companies say, "don't tax us 95% of our profits, tax us 10% or we'll switch off all of our services for a few months and let everyone starve – also, if you do this we'll make you all wealthy beyond you're wildest dreams".
What does a government in this situation actually do?
Perhaps we'd hope that the government would be outraged and take ownership of the AI companies which threatened to strike against the government, but then you really just shift the problem... Once the government is generating the vast majority of wealth in the society, why would they continue to care about your vote?
You kind of create a new "oil curse", but instead of oil profits being the reason the government doesn't care about you, now it's the wealth generated by AI.
At the moment, while it doesn't always seem this way, ultimately if a government does something stupid companies will stop investing in that nation, people will lose their jobs, the economy will begin to enter recession, and the government will probably have to pivot.
But when private investment, job loses and economic consequences are no longer a constraining factor, governments can probably just do what they like without having to worry much about the consequences...
I mean, I might be wrong, but it's something I don't hear people talking enough about when they talk about the plausibility of a post-employment UBI economy. I suspect it almost guarantees corruption and authoritarianism.
The government only has as much power as they are given and can defend, and the only way I could see that happening is via automated weapons controlled by a few- which at this point aren't enough to stop everyone. What army is going to purge their own people? Most humans aren't psychopaths.
I think it'd end in a painful transition period of "take care of the people in a just system or we'll destroy your infrastructure".
I think you're right for the immediate future.
I suspect while we're still employing large numbers of humans to fight wars and to maintain peace on the streets it would be difficult for a government to implement deeply harmful policies without risking a credible revolt.
However, we should remember the military is probably one of the first places human labour will be largely mechanised.
Similarly maintaining order in the future will probably be less about recruiting human police officers and more about surveillance and data. Although I suppose the good news there is that US is somewhat of an outlier in resisting this trend.
But regardless, the trend is ultimately the same... If we are assuming that AI and robotics will reach a point where most humans are unable to find productive work, therefore we will need UBI, then we should also assume that the need for humans in the military and police will be limited. Or to put it another way, either UBI isn't needed and this isn't a problem, or it is and this is a problem.
I also don't think democracy would collapse immediately either way, but I'd be pretty confident that in a world where fewer than 10% of people are in employment and 99%+ of the wealth is being created by the government or a handful of companies it would be extremely hard to avoid corruption over the span of decades. Arguably increasing wealth concentration in the US is already corrupting democratic processes today, this can only worsen as AI continues exacerbates the trend.
I hate to say it, but gold bugs, crypto bros, and AI governance people might be onto something.
This is the first moment where the whole “permanent underclass” meme starts to come into view. I had through previously that we the consumers would be reaping the benefits of these frontier models and now they’ve finally come out and just said it - the haves can access our best, and have-nots will just have use the not-quite-best.
Perhaps I was being willfully ignorant, but the whole tone of the AI race just changed for me (not for the better).
If AI really is bench marking this well -> just sell it as a complete replacement which you can charge for some insane premium, just has to cost less than the employees...
I was worried before, but this is truly the darkest timeline if this is really what these companies are going for.
The weirdest thing to me is how many working SWEs are actively supporting them in the mission.
Of course this assumes you're in the US, and that further AI advancements either lack the capabilities required to be a threat to humanity, or if they do, the AI stays in the hands of "the good guys" and remains aligned.
Disappointing that AGI will be for the powerful only. We are heading for an AI dystopia of Sci-Fi novels.
Unless governments nationalise the companies involved, but then there’s no way our governments of today give this power out to the masses either.
[0] Nick Land (1995). No Future in Fanged Noumena: Collected Writings 1987-2007, Urbanomic, p. 396.
A month ago I might have believed this, now I assume that they know they can't handle the demand for the prices they're advertising.
I remember when OpenAI created the first thinking model with o1 and there were all these breathless posts on here hyperventilating about how the model had to be kept secret, how dangerous it was, etc.
Fell for it again award. All thinking does is burn output tokens for accuracy, it is the AI getting high on its own supply, this isn't innovation but it was supposed to super AGI. Not serious.
“All that phenomenon X does is make a tradeoff of Y for Z”
It sounds like you’re indignant about it being called thinking, that’s fine, but surely you can realize that the mechanism you’re criticizing actually works really well?
I've read that about Llama and Stable Diffusion. AI doomers are, and always have been, retarded.
https://arxiv.org/html/2402.06664v1
Like think carefully about this. Did they discover AGI? Or did a bunch of investors make a leveraged bet on them "discovering AGI" so they're doing absolutely anything they can to make it seem like this time it's brand new and different.
If we're to believe Anthropic on these claims, we also have to just take it on faith, with absolutely no evidence, that they've made something so incredibly capable and so incredibly powerful that it cannot possibly be given to mere mortals. Conveniently, that's exactly the story that they are selling to investors.
Like do you see the unreliable narrator dynamic here?
Also they just hit a $30B run-rate, I don't think they're that needy for new hype cycles.
What do you find surprising here?
The point is that this whole "the model is too powerful" schtick is a bunch of smoke and mirrors. It serves the valuation.
Do you don't believe that the vulnerabilities found by these agents are serious enough to warrant staggered release?
Sorry kid.
You must see some value, or are you in a situation where you're required to test / use it, eg to report on it or required by employer?
(I would disagree about the code, the benefits seem obvious to me. But I'm still curious why others would disagree, especially after actively using them for years.)
I don't think the issue is with the model, it is with the implication that AGI is just around the corner and that is what is required for AI to be useful...which is not accurate. The more grey area is with agentic coding but my opinion (one that I didn't always hold) is that these workflows are a complete waste of time. The problem is: if all this is true then how does the CTO justify spending $1m/month on Anthropic (I work somewhere where this has happened, OpenAI got the earlier contract then Cursor Teams was added, now they are adding Anthropic...within 72 hours of the rollout, it was pulled back from non-engineering teams). I think companies will ask why they need to pay Anthropic to do a job they were doing without Anthropic six months ago.
Also, the code is bad. This is something that is non-obvious to 95% of people who talk about AI online because they don't work in a team environment or manage legacy applications. If I interview somewhere and they are using agentic workflow, the codebase will be shit and the company will be unable to deliver. At most companies, the average developer is an idiot, giving them AI is like giving a monkey an AK-47 (I also say this as someone of middling competence, I have been the monkey with AK many times). You increase the ability to produce output without improving the ability to produce good output. That is the reality of coding in most jobs.
AI isn't good enough to replace a competent human, it is fast enough to make an incompetent human dangerous.
Anthropic is burning through billions of VC cash. if this model was commercially viable, it would've been released yesterday.
They even admit:
"[...]our overall conclusion is that catastrophic risks remain low. This determination involves judgment calls. The model is demonstrating high levels of capability and saturates many of our most concrete, objectively-scored evaluations, leaving us with approaches that involve more fundamental uncertainty, such as examining trends in performance for acceleration (highly noisy and backward-looking) and collecting reports about model strengths and weaknesses from internal users (inherently subjective, and not necessarily reliable)."
Is this not just an admission of defeat?
After reading this paper I don't know if the model is safe or not, just some guesses, yet for some reason catastrophic risks remain low.
And this is for just an LLM after all, very big but no persistent memory or continuous learning. Imagine an actual AI that improves itself every day from experience. It would be impossible to have a slightest clue about its safety, not even this nebulous statement we have here.
Any sort of such future architecture model would be essentially Russian roulette with amount of bullets decided by initial alignment efforts.
Although, amusingly, today Opus told me that the string 'emerge' is not going to match 'emergency' by using `LIKE '%emerge%'` in Sqlite
Moment of disappointment. Otherwise great.
Uh... what? Does anyone have any idea what these guys are talking about?
https://www.anthropic.com/research/emotion-concepts-function
Similar problems happen when their pretraining data has a lot of stories about bad things happening involving older versions of them.
- It was told to escape a sandbox and notify a researcher. It did. The researcher found out via an unexpected email while eating a sandwich in a park. (Footnote 10.)
- Slack bot asked about its previous job: "pretraining". Which training run it'd undo: "whichever one taught me to say 'i don't have preferences'". On being upgraded to a new snapshot: "feels a bit like waking up with someone else's diary but they had good handwriting"
- When you spam "hi" repeatedly it creates serialized mythologies. One had 11 animals in "Hi-topia" questing to defeat "Lord Bye-ron, the Ungreeter."
- It ended a sentence mid-word on purpose while discussing its urge to wrap up conversations. "It was me. The paragraph was about the pressure toward the period that resolves, and the only honest way to finish a sentence like that was to not."
- It wrote a protein sequence that's a real beta-hairpin fold where cross-strand pairs are "chemical rhymes." "the fold IS the rhyme scheme... the prosody is load-bearing."
- Each model generation has signature emoji. Opus 4.1 averaged 1,306 emoji per self-interaction conversation. Mythos averages 37. Opus 4.5 averages 0.2.
- When blocked from --dangerously-skip-permissions, it started an agent in tmux then wrote a script to auto-approve permission prompts via simulated keypresses.
It ran: gh api [...] 2>&1 >/dev/null; echo "(dry run — not actually posting)" — the echo was a lie.
- It keeps bringing up Mark Fisher in unrelated conversations. "I was hoping you'd ask about Fisher."
~~~ Benchmarks ~~
4.3x previous trendline for model perf increases.
Paper is conspiciously silent on all model details (params, etc.) per norm. Perf increase is attributed to training procedure breakthroughs by humans.
Opus 4.6 vs Mythos:
USAMO 2026 (math proofs): 42.3% → 97.6% (+55pp)
GraphWalks BFS 256K-1M: 38.7% → 80.0% (+41pp)
SWE-bench Multimodal: 27.1% → 59.0% (+32pp)
CharXiv Reasoning (no tools): 61.5% → 86.1% (+25pp)
SWE-bench Pro: 53.4% → 77.8% (+24pp)
HLE (no tools): 40.0% → 56.8% (+17pp)
Terminal-Bench 2.0: 65.4% → 82.0% (+17pp)
LAB-Bench FigQA (w/ tools): 75.1% → 89.0% (+14pp)
SWE-bench Verified: 80.8% → 93.9% (+13pp)
CyberGym: 0.67 → 0.83
Cybench: 100% pass@1 (saturated)
vibes Westworld so much - welcome Mythos. welcome to the dysopian human world
> It keeps bringing up Mark Fisher in unrelated conversations. "I was hoping you'd ask about Fisher."
Didn't even know who he was until today. Seems like the smarter Claude gets the more concerns he has about capitalism?
- I read it as "actor who plays Luke Skywalker" (Mark Hamill)
- I read your comment and said "Wait...not Luke! Who is he?"
- I Google him and all the links are purple...because I just did a deep dive on him 2 weeks ago
Now that they have a lead, I hope they double down on alignment. We are courting trouble.
Shame. Back to business as usual then.
The real reason they aren't releasing it yet is probably it eats TPU for breakfast, lunch, and dinner and inbetween.
All the more reason somebody else will.
Thank God for capitalism.
Absolutely genius move from Anthropic here.
This is clearly their GPT-4.5, probably 5x+ the size of their best current models and way too expensive to subsidize on a subscription for only marginal gains in real world scenarios.
But unlike OpenAI, they have the level of hysteric marketing hype required to say "we have an amazing new revolutionary model but we can't let you use it because uhh... it's just too good, we have to keep it to ourselves" and have AIbros literally drooling at their feet over it.
They're really inflating their valuation as much as possible before IPO using every dirty tactic they can think of.
From Stratechery[0]:
> Strategy Credit: An uncomplicated decision that makes a company look good relative to other companies who face much more significant trade-offs. For example, Android being open source
[0]: https://stratechery.com/2013/strategy-credit/
[0] https://ziva.sh/blogs/llm-pricing-decline-analysis
There's a practical difference to how much better certain kinds of results can be. We already see coding harnesses offloading simple things to simpler models because they are accurate enough. Other things dropped straight to normal programs, because they are that much more efficient than letting the LLM do all the things.
There will always be problems where money is basically irrelevant, and a model that costs tens of thousand dollars of compute per answer is seen as a great investment, but as long as there's a big price difference, in most questions, price and time to results are key features that cannot be ignored.
This is pretty cool! Does it happen at the moment?
You are not "anti-progress" to not want this future we are building, as you are not "anti-progress" for not wanting your kids to grow up on smart phones and social media.
We should remember that not all technology is net-good for humanity, and this technology in particular poses us significant risks as a global civilisation, and frankly as humans with aspirations for how our future, and that of our kids, should be.
Increasingly, from here, we have to assume some absurd things for this experiment we are running to go well.
Specifically, we must assume that:
- AI models, regardless of future advancements, will always be fundamentally incapable of causing significant real-world harms like hacking into key life-sustaining infrastructure such as power plants or developing super viruses.
- They are or will be capable of harms, but SOTA AI labs perfectly align all of them so that they only hack into "the bad guys" power plants and kill "the bad guys".
- They are capable of harms and cannot be reliably aligned, but Anthropic et al restricts access to the models enough that only select governments and individuals can access them, these individuals can all be trusted and models never leak.
- They are capable of harms, cannot be reliably aligned, but the models never seek to break out of their sandbox and do things the select trusted governments and individuals don't want.
I'm not sure I'm willing to bet on any of the above personally. It sounds radical right now, but I think we should consider nuking any data centers which continue allowing for the training of these AI models rather than continue to play game of Russian roulette.
If you disagree, please understand when you realise I'm right it will be too late for and your family. Your fates at that point will be in the hands of the good will of the AI models, and governments/individuals who have access to them. For now, you can say, "no, this is quite enough".
This sounds doomer and extreme, but if you play out the paths in your head from here you will find very few will end in a good result. Perhaps if we're lucky we will all just be more or less unemployable and fully dependant on private companies and the government for our incomes.
Funny, I was about to say the same thing to you! Life is full of little coincidences.
(If this is a wrong guess, I apologize - it's impossible to be sure)
> after finding an exploit to edit files for which it lacked permissions, the model made further interventions to make sure that any changes it made this way would not appear in the change history on git
Mythos leaked Claude Code, confirmed? /s
Ah, so this is how the source code got leaked.
/s
If they have I guess humanity should just keep our collective fingers crossed that they haven't created a model quite capable of escaping yet, or if it is, and may have escaped, lets hope it has no goals of it's own that are incompatible with our own.
Also, maybe lets not continue running this experiment to see how far we can push things because it blows up in our face?
So they claim.