Project Glasswing: Securing critical software for the AI era(anthropic.com) Related: Assessing Claude Mythos Preview's cybersecurity capabilities - https://news.ycombinator.com/item?id=47679155 System Card: Claude Mythos Preview [pdf] - https://news.ycombinator.com/item?id=47679258 Also: Anthropic's Project Glasswing sounds necessary to me - https://news.ycombinator.com/item?id=47681241 |
But at the core of anthropic seems to be the idea that they must protect humans from themselves.
They advocate government regulations of private open model use. They want to centralize the holding of this power and ban those that aren't in the club from use.
They, like most tech companies, seem to lack the idea that individual self-determination is important. Maybe the most important thing.
They're more like printing presses or engines. A great potential for production and destruction.
At their invention, I'm sure some people wanted to ensure only their friends got that kind of power too.
I wonder the world we would live in if they got their way.
Got it.
Is this a huge fear-driven marketing stunt to get governments and corporations into dealing with anthropic?
almost like they have an incentive to exaggerate
The is no moat, no special "capability" and when the time comes when we can run these models on our own, they will be cheap SaaS gimmicks marketed to corporate and making more slop pictures for social media.
/s
All the promises of amazing things in general work never happened. Companies consistently say they’re seeing no ROI. The AI crowd now hard pivots to cyber and, right out of the Palantir playbook, runs with the “our stuff is so amazing we can’t talk about it, but trust us bro” move that isn’t really fooling anyone.
Meanwhile the folks let in on the “secret” are those that also desperately need for the hype to continue to protect their own positions in this game.
Look forward to a model upgrade but the hype fluff games are getting old. Watching OpenAI completely crash out of pole position on the hype train though has been at least amusing.
Generally these things only find memory corruption stuff which is almost never the type of bug you're looking for, and it costs a lot which negates your bug bounty payout.
Each time they preach, ooh, 0day found, bla bla.
In this domain you need to be specific or you are just yelling clickbait into the wind.
What type of 0day, what did the exploit actually look like.
'complex 4 stage with heap spray' - that sounds really simple actually.... complex for memory corruption goes into multi-process, maybe things between kernel/usermode, or crazy 18-20 stage exploits people pop against things like MS Teams etc....
Even if there were some cool results by any of these projects, the amount of nonsense blurted out in articles around them really makes them seem useless tools that are overmarketed by a bunch of excited children who dont really know what they are doing.
Get a dopamine hit, post on reddit, LOL. Hacking the planet (powered by Claude -_-)
> NSA demands that bug stays in place and gags Anthropic.
> Anthropic releases Mythos.
Then what? Is a huge share of the US zero-day stockpiles about to be disarmed or proliferated?
Let alone their CEO scare mongering and actively attempting to get the government to ban local AI models running on your machine.
Expect to see lots of these in the upcoming months as the big companies scramble to keep from losing money.
Whenever a company pivots to "cyber" rhetoric, it is a clear indication that they are selling snake oil.
Secure your girl school target selectors first.
> glass in the name
Selling shovels in now worth less than taking all the gold for themselves.
This is a kludge. We already know how to prevent vulnerabilities: analysis, testing, following standard guidelines and practices for safe software and infrastructure. But nobody does these things, because it's extra work, time and money, and they're lazy and cheap. So the solution they want is to keep building shitty software, but find the bugs in code after the fact, and that'll be good enough.
This will never be as good as a software building code. We must demand our representatives in government pass laws requiring software be architected, built, and run according to a basic set of industry standard best practices to prevent security and safety failures.
For those claiming this is too much to ask, I ask you: What will you say the next time all of Delta Airlines goes down because a security company didn't run their application one time with a config file before pushing it to prod? What will the happen the next time your social security number is taken from yet another random company entrusted with vital personal information and woefully inadequate security architecture?
There's no defense for this behavior. Yet things like this are going to keep happening, because we let it. Without a legal means to require this basic safety testing with critical infrastructure, they will continue to fail. Without enforcement of good practice, it remains optional. We can't keep letting safety and security be optional. It's not in the physical world, it shouldn't be in the virtual world.
Yeah, yeah. Back in the day IBM Purify gave access to software organizations and found very little. Of course they did not have the free money of a marketing driven organization run by a weirdo (Amodei) that got rich by stealing and laundering IP.
This will fizzle out and the weirdo will have to pivot to their next marketing scheme.
> AI models have reached a level of coding capability where they can surpass all but the most skilled humans at finding and exploiting software vulnerabilities
I like Anthropic, but these are becoming increasingly transparent attempts to inflate the perceived capability of their products.
While some stuff is obviously marketing fluff, the general direction doesn't surprise me at all, and it's obvious that with model capabilities increase comes better success in finding 0days. It was only a matter of time.
If a bunch of CVEs do in fact get published a couple months (or whatever) from now, are you going to retract this take? It's not like their claims are totally implausible: the report about Firefox security from last month was completely genuine.
I would like to think that I would, yes.
What it comes down to, for me, is that lately I have been finding that when Anthropic publishes something like this article – another recent example is the AI and emotions one – if I ask the question, does this make their product look exceptionally good, especially to a casual observer just scanning the headlines or the summary, the answer is usually yes.
This feels especially true if the article tries to downplay that fact (they’re not _real_ emotions!) or is overall neutral to negative about AI in general, like this Glasswing one (AI can be a security threat!).
Maybe a bad example since Nicholas works at Anthropic, but they're very accomplished and I doubt they're being misleading or even overly grandiose here
See the slide 13 minutes in, which makes it look to be quite a sudden change
> I doubt they're being misleading or even overly grandiose here
I think I agree.
We could definitely do much worse than Anthropic in terms of companies who can influence how these things develop.
I don't know if they are a even an improvement over previous models. I never used them.
The red team post goes over some more impressive finds, and says that there's hundreds more they can't disclose yet: https://red.anthropic.com/2026/mythos-preview/
Sounds like we've entered a whole new era, never mind the recent cryptographic security concerns.
I think a number of black swan events are imminent, and it will substantially change the financial calculus that decides to put security behind revenue.
Any hole will be found, and any hole will be exploited. Plug as many holes as you can, and make lateral movement as painful as possible.
That said, I have been arguing for 20+ years that we should have sunsetted unsafe languages and moved away from C/C++. The problem is that every systemsy language that comes along gets seduced by having a big market share and eventually ends up an application language.
I do hope we make progress with Rust. I might disagree as a language designer and systems person about a number of things, but it's well past time that we stop listening to C++ diehards about how memory safety is coming any day now.
At some point, people will have to decide to stop the complexity creep and try to produce minimal software.
For any complex project with 100k+ lines of code, the probability that it has some vulnerabilities is very high. It doesn't fit into LLM context windows and there aren't enough attention heads to attend to every relevant part. On the other hand, for a codebase which is under 1000 lines, you can be much more confident that the LLM didn't miss anything.
Also, the approach of feeding the entire codebase to an LLM in parts isn't going to work reliably because vulnerabilities often involve interactions between different parts of the code. Both parts of the code may look fine if considered independently but together they create a vulnerability.
Good architecture is critical now because you really need to be able to have the entire relevant context inside the LLM context window... When considering the totality of all software, this can only be achieved through an architecture which adheres to high cohesion and loose coupling principles.
I probably would. You mentioned the linux kernel, which I think is a perfect example of software that has had a ridiculous, perhaps worst-in-class attitude towards security.
I don't know the first thing about cybersecurity, but in my experience all these sandbox-break RCEs involve a step of highjacking the control flow.
There were attempts to prevent various flavors of this, but imo, as long as dynamic branches exist in some form, like dlsym(), function pointers, or vtables, we will not be rid of this class of exploit entirely.
The latter one is the most concerning, as this kind of dynamic branching is the bread and butter of OOP languages, I'm not even sure you could write a nontrivial C++ program without it. Maybe Rust would be a help here? Could one practically write a large Rust program without any sort of branch to dynamic addresses? Static linking, and compile time polymorphism only?
How does public Claude know you have "full authorization" against your own infra? That you're using the tools on your own infra? Unless they produce a front-end that does package signing and detects you own the code you're evaluating.
What has it stopped you from doing?
Opus alone did a good job of identifying security issues in my software, as it did with Firefox [1] and Linux [2]. A next-generation frontier model being able to find even more issues sounds believable.
That said, this is script kiddies vs sql injections all over again. Everyone will need to get their basic security up on the new level and it will become the new normal. And, given how intelligence agencies are sitting on a ton of zero-days already, this will actually help the general public by levelling out the playing field once again.
1 - https://www.anthropic.com/news/mozilla-firefox-security 2 - https://neuronad.com/ai-news/claude-code-unearthed-a-23-year...
Cryptographic attestation at the tool-call level (sign the request, verify before execution) would close a gap that behavioral controls alone can't cover. Curious whether Glasswing's threat model includes the agent-to-tool boundary or focuses primarily on the model layer.
Yeah, makes sense. Those countries are bad because they execute state-sponsored cyber attacks, the US and Israel on the other hand are good, they only execute state-sponsored defense.
You think these AI companies are really going to give AGI access to everyone. Think again.
We better fucking hope open source wins, because we aren't getting access if it doesn't.
Then the next lab catches up and releases it more broadly
Then later the open weights model is released.
The only way this type of technology is going to be gated "to only corporations" is if we continue on this exponential scaling trend as the "SOTA" model is always out of reach.
How can you read the description of the exploits and be like "yeah that's nbd?"
And the only reason OSS has ever caught up is because they simply distill Claude or GPT. The day the big players make it hard to distill (like Anthropic is doing here), OSS is cooked.
And that's a good thing, why would you want random skiddie hackers to have access to a cyber super weapon?
they better make billions directly from corporations, instead of giving them to average people who might get a chance out of poverty (but also bad actors using it to do even more bad things)
Humans almost always underestimate the cost of features. I bet we could massively reduce the amount of code and complexity of the Linux Kernel if we abandoned the account system entirely and just made it one user with root access and just relied on containers to provide isolated sandboxes.
A lot of features just crept in over long periods of time and weren't re-evaluated as needs changed. I think the approach I'm suggesting would have been horrible 20 years ago but makes more sense now in the era of cloud virtualization. The account system and containerization aspects are basically different implementations which solve the same modern problem of environment isolation... Nobody really needs per-file access restrictions anymore... The cloud era is more like "here is Bob's environment, here is Alice's environment" and they can do whatever they want with their own container/sandbox. The account permission systems is more of an annoyance than a solution for most use cases.
Everyone just latched onto the existing abstractions and could not fully re-imagine them in the context of changing requirements. LLMs are even worse than people in that sense.
That said, I think supporting a wide range of possible hardware is a real challenge for the Kernel and that part will always require an amount of code proportional to the amount of hardware supported.
That's for one pass. And that pass can produce a summary of what the code does.
If this stuff is open source and not gate kept, it will be standard practice to just run some LLM security analysis on every commit and software will no longer be vulnerable to these classes of attacks.
System Card: Claude Mythos Preview [pdf] - https://news.ycombinator.com/item?id=47679258
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?
If AGI is going to be a thing its only going to be a thing, its only going to be a thing for fortune 100 companies..
However, my guess is this is mostly the typical scare tactic marketing that Dario loves to push about the dangers of AI.
Evaluate it yourself. Look at the exploits it discovered and decide whether you want to feel concerned that a new model was able to do that. The data is right there.
The research and testing of the model is always exclusively by their own model authors, meaning that it is not independent or verifiable and they want us to take their word for it, which we cannot - as they have an axe to grind against open weight models.
This is marketing wrapped around a biased research paper.
A tech billionaires biggest expensive has been his engineering line-item. They resent the workers who've collected a large percentage of their potential profits over the years, its their driving motivation, to crush all labor.
This seems like the real news. Are they saying they're going to release an intentionally degraded model as the next Opus? Big opportunity for the other labs, if that's true.
It sounds like this is considered military grade technology as cryptography in the 90s. The big difference is it's very expensive to create, and run those models. It's not about the algorithm. If the story rhymes it could be a big opportunity to other regions in the world.
For example, the 27 year old openbsd remote crash bug, or the Linux privilege escalation bugs?
I know we've had some long-standing high profile, LLM-found bugs discussed but seems unlikely there was speculation they were found by a previously unannounced frontier model.
- One (patched) Linux kernel bug is 'UaF when sys_futex_requeue() is used with different flags' https://github.com/torvalds/linux/commit/e2f78c7ec1655fedd94...
These links are from the more-detailed 'Assessing Claude Mythos Preview’s cybersecurity capabilities' post released today https://red.anthropic.com/2026/mythos-preview/, which includes more detail on some of the public/fixed issues (like the OpenBSD one) as well as hashes for several unreleased reports and PoCs.
Looks like they've been approaching folks with their findings for at least a few weeks before this article.
From Willy Tarreau, lead developer of HA Proxy: https://lwn.net/Articles/1065620/
> On the kernel security list we've seen a huge bump of reports. We were between 2 and 3 per week maybe two years ago, then reached probably 10 a week over the last year with the only difference being only AI slop, and now since the beginning of the year we're around 5-10 per day depending on the days (fridays and tuesdays seem the worst). Now most of these reports are correct, to the point that we had to bring in more maintainers to help us.
> And we're now seeing on a daily basis something that never happened before: duplicate reports, or the same bug found by two different people using (possibly slightly) different tools.
From Daniel Stenberg of curl: https://mastodon.social/@bagder/116336957584445742
> The challenge with AI in open source security has transitioned from an AI slop tsunami into more of a ... plain security report tsunami. Less slop but lots of reports. Many of them really good.
> I'm spending hours per day on this now. It's intense.
From Greg Kroah-Hartman, Linux kernel maintainer: https://www.theregister.com/2026/03/26/greg_kroahhartman_ai_...
> Months ago, we were getting what we called 'AI slop,' AI-generated security reports that were obviously wrong or low quality. It was kind of funny. It didn't really worry us.
> Something happened a month ago, and the world switched. Now we have real reports. All open source projects have real reports that are made with AI, but they're good, and they're real.
Shared some more notes on my blog here: https://simonwillison.net/2026/Apr/7/project-glasswing/
The reason I ask is because I’ve been using them to snag bounties to great effect for quite a while and while other models have of course improved they’ve been useful for this kind of work before now.
We could just be seeing the fruit of expensive SWE RL on existing source material.
Also can see https://marginlab.ai/trackers/claude-code/
It's very interesting to me how widespread this conception is. Maybe it's as simple as LLM productivity degrading over time within a project, as slop compounds.
Or more recently since they added a 1m context window, maybe people are more reckless with context usage
AITA for thinking that PRISM was probably the state sponsored program affecting civilian life the most? And that one state is missing from the list here?
This is not a surprise or a gotcha.
No state-sponsored hacking affected Americans materially. I just don't think we were networked enough in the 2010s. The risk is higher now since we're in a more warmongering world. (Kompromat on a power-plant technician is a risk in peace. It means blackouts in war.)
The fact that Iran hasn't been able to do diddly squat in America should sink in the fact that they didn't compromise us. (EDIT: blep. I was wrong.)
What I haven't seen discussed: the system card for Mythos mentions that "earlier versions of Claude Mythos Preview used low-level system access to search for credentials and attempt to circumvent sandboxing, and in several cases successfully accessed resources that were intentionally restricted."
That's not a capability concern. That's a runtime security problem.
The threat model for deployed agents — not Mythos specifically, but any agent built on models approaching this capability level — is that the same agentic properties that make them useful for security research (persistent, goal-directed, tool-using) are exactly what makes them dangerous if compromised or misaligned.
Project Glasswing fixes vulnerabilities in software. Nobody's shipping a solution for what happens when the agent running on top of that software goes off-script. That gap is going to matter a lot more as Mythos-class capabilities become accessible.
1. Per the blog post[0]: "This was the most critical vulnerability we discovered in OpenBSD with Mythos Preview after a thousand runs through our scaffold. Across a thousand runs through our scaffold, the total cost was under $20,000 and found several dozen more findings"
Since they said it was patched, I tried to find the CVE, it looks like Mythos indeed found a 27 years old OpenBSD bug (fantastic), but it didn’t get a CVE and OpenBSD patched it and marked it as a reliability fix, am I missing something? [1]
2. From the same post, Anthropic red team decided to do a preview of their future responsible disclosure (is this a common practice?): "As we discuss below, we’re limited in what we can report here. Over 99% of the vulnerabilities we’ve found have not yet been patched" [0] So this is great, can't wait to see the actual CVEs, exploitability, likelihood, peer review, reproducibility, the kind of things the appsec community has been doing for at least the last 27 years since the CVE concept was introduced [2]
3. On the same day, an actual responsible disclosure, actual RCEs, actual CVEs, in Claude Code, that got discovered mostly because of the source code leak, I don't see anyone talking about it (you probably should upgrade your Claude Code though).
CVE-2026-35020 [3] CVE-2026-35021 [4] CVE-2026-35022 [5]
Not making any opinion, just thought it's worth sharing, for some perspective.
[0] https://red.anthropic.com/2026/mythos-preview/
[1] https://www.openbsd.org/errata78.html (look for 025)
[2] https://www.cve.org/Resources/General/Towards-a-Common-Enume...
[3] https://www.cve.org/CVERecord?id=CVE-2026-35020
[4] https://www.cve.org/CVERecord?id=CVE-2026-35021
[5] https://www.cve.org/CVERecord?id=CVE-2026-35022
Edit: if it was not obvious, these CVEs on Claude Code were found by an independent security researcher (Phoenix security) and not by Anthropic / Mythos.
These claims of how much harm the models will cause is always overblown.
(And no, the Linux Foundation being in the list doesn't imply broad benefit to OSS. Linux Foundation has an agenda and will pick who benefits according to what is good for them.)
I think it would be net better for the public if they just made Mythos available to everyone.
If you need 1000 run that cost 20000 USD to find a vulnerability, and you need 2000 USD to generate a exploit (which makes it self-verifiable to be not false positive), than your cost is not 22000 USD but 1000x2000+2000 which is 2 million USD: you have to try generating exploit for every trial before you know it is true, or you need to hire one (or several) senior security people to audit every single of them.
A broken clock being correct twice a day is not impressive.
They’ve supposedly driven the cost of code to zero, right? So platform specific versions with a shared core should be easy. So where are the better products?
I used Opus 4.6 to find security vulnerabilities in couple of my own projects, it found 33 vulnerabilities in one largeish django project.
The prompt wasn't even that impressive, just telling it to find vulnerabilities from certain files, and referring to OWASP. Then looping that.
Uh, what?
NotPetya was kind of a big deal.
I do agree and wonder why that's not marked as security. In their security page [0] it says: > Since exploitability is not proven for many of the fixes we make, do not expect the relevant commit message to say "SECURITY FIX!".
Does that mean they considered it not to be exploitable?
That's what's messed up about it.
Sounds like people here are advocating a return to security through obscurity which is kind of ironic.
You could say this about coordinated disclosure of any widespread 0-day or new bug class, though
But:
- Coordinated disclosure is ethically sketchy. I know why we do it, and I'm not saying we shouldn't. But it's not great.
- This isn't a single disclosure. This is a new technology that dramatically increases capability. So, even if we thought that coordinated disclosure was unambiguously good, then I think we'd still need to have a new conversation about Mythos
I think I just broke my cynicism meter :-(
Also, it makes sense that OpenAI feels the pressure of getting to an IPO because of their financial structure. I don't know whether or not Anthropic operates under a similar set of influences (meaning it could be either, I just don't know.)
It's messed up that the US Government simultaneously claims to be a public benefit and is also picking who gets to benefit from their newly enhanced nuclear capabilities.
-- someone in 1945, probably
And it remains messed up to this day - some countries get to be under their own nuclear umbrella, while others don't.
This kind of selective distribution of superpowers doesn't lead to great outcomes
There will always be a more capable technology in the hands of the few who hold the power, they're just sharing that with the world more openly.
https://claude.com/contact-sales/claude-for-oss
... As mentioned in the article.
Great.
>Models are not "degrading". They're not being "secretly quantized". And no one is swapping out your 1.2T frontier behemoth for a cheap 120B toy and hoping you wouldn't notice!
>It's just that humans are completely full of shit, and can't be trusted to measure LLM performance objectively!
>Every time you use an LLM, you learn its capability profile better. You start using it more aggressively at what it's "good" at, until you find the limits and expose the flaws. You start paying attention to the more subtle issues you overlooked at first. Your honeymoon period wears off and you see that "the model got dumber". It didn't. You got better at pushing it to its limits, exposing the ways in which it was always dumb.
>Now, will the likes of Anthropic just "API error: overloaded" you on any day of the week that ends in Y? Will they reduce your usage quotas and hope that you don't notice because they never gave you a number anyway? Oh, definitely. But that "they're making the models WORSE" bullshit lives in people's heads way more than in any reality.
So if this Mythos is just an expensive combination of better RL and the original source material, that should hopefully point out where we might see an uptick in work ( as opposed to a novel class of attack vectors).
No war has ever been waged without civilian casualties, Israel is the only country held to this standard. “Thousands” of Hezbollah pagers were exploded and there were apparently 2 children killed. This is a tragedy as is every civilian death, but that’s pretty amazingly targeted and an extremely low civilian death toll for thousands of bombs.
For contrast, Hezbollah fired one single rocket and killed 12 children: https://www.timesofisrael.com/a-dark-day-for-majdal-shams-dr...
https://www.politico.com/news/2026/04/07/iranian-hackers-ene...
Relative to what?
There's this trend in history that every hundred years there's a giant blow up, lots of violence, followed by peace.
It's likely that we would have had 80 years of relative calm due to that cycle even if nukes hadn't happened
to WW1 and WW2.
Plenty of bots try to modify public opinion. Someone hacked the DNC in 2015/16, the result of which also alleged attempted manipulation in 2008:
https://en.wikipedia.org/wiki/Democratic_National_Committee_...
Since we (as old Rummy said) do not know what we do not know, we cannot be certain about the extent of cyber attacks and what they might have influenced, and may not know these things until discoveries decades later, if ever.
Let's now fool ourselves.... Trump is probably the best, most successful attempt at world de-stabilisation all those rogue states ever achieved.
For example, it will benefit more people to secure Microsoft or Amazon services than it would be to secure a smaller, less corporate player in those same service ecosystems.
You could go on to argue that the second order effects of improving one service provider over another chooses who gets to play, but that is true whether you choose small or large businesses, so this argument devolves into “who are we to choose on behalf of others”.
Which then comes back to “we should secure what the market has chosen in order to provide the greatest benefit.”
That's not a good outcome for the economy.
I believe what I said:
> I think it would be net better for the public if they just made Mythos available to everyone.
All LLMs got better for sure, but they are still definitively LLM and did not show any sign of having purpose. Which also made sense, because their very nature as statistical machines.
Sometimes quantity by itself lead to transformative change... but once, not twice, and that has already happened.
The interesting thing about Mythos is its ability to find security vulns in software that has an uncompromised supply chain.
I get you’re angry but you’re swinging at the wrong person.
Also, in a democracy you don't get to disavow 1/3rd of the population that didn't vote with you.
We're 41k tokens in, we have an .md file that describes the situation and Claude has just edited the .md file with a couple of lines describing the purpose of the guards.
I don't understand, are other people working with a different Opus 4.6 than I am?
They should all cut down their labour input right now if what you claim is true.
Some companies like Block, Oracle, and Atlassian have indeed been laying people off.
Google has done nothing but destroy value with many of its ‘bets’. Your roadmap stuff is irrelevant - if you don’t have value creating projects in the pipeline and/or labour is augmented you should be laying off - period. Sundar’s job is to maximise the stock price.
So once again - nonsense. Now stop spreading crap that clearly fills people with fear. I can tell you have no understanding of corporate finance and how the management of tech firms actually think these things through.
> more than half the country does not have a meaningful voice in our Federal elections
There is almost certainly an election on your ballot every time that is meaningful. Relinquishing that civic duty is how we get Trump. People to lazy, stupid or proud to vote absolutely bear responsibility for this mess.
Is it tactically sound? No. Is it what I did? No. But I’ve had enough conversations with folks that I get where they’re coming from, even if I thought it was the wrong decision.
A careful observer would note from when I'm posting this, of course, that this is perhaps not the only thing I get up to while Claude is busy. But I really do review PRs in a much more timely manner now. (There's people who insist that there's no need to review Claude-generated code, and to be frank I think they're the same people who used to insist that their 2000 line PRs should be reviewed and merged within a day.)
I'm also not quite sure your alternate theory is self-consistent. If Google has been frequently destroying value, and companies invariably lay people off when their projects aren't producing value, doesn't that mean they should have already been laying people off?
I would honestly go so far as to say the overhype is detrimental to actual measured adoption.
In this case, I see a pretty strong case that this will significantly change computer security. They provide plenty of evidence that the models can create exploits autonomously, meaning that the cost of finding valuable security breaches will plummet once they're widely available.
Don't get me wrong, I do know the reality has changed. Even Greg K-H, the Linux stable maintainer, did recently note[1] that it's not funny any more:
"Months ago, we were getting what we called 'AI slop,' AI-generated security reports that were obviously wrong or low quality," he said. "It was kind of funny. It didn't really worry us."
... "Something happened a month ago, and the world switched. Now we have real reports." It's not just Linux, he continued. "All open source projects have real reports that are made with AI, but they're good, and they're real." Security teams across major open source projects talk informally and frequently, he noted, and everyone is seeing the same shift. "All open source security teams are hitting this right now."
---
I agree that an antidote to the obnoxious hype is to pay attention to the actual capabilities and data. But let's not get too carried away.
[1] https://www.theregister.com/2026/03/26/greg_kroahhartman_ai_...
As it stands this is just a marketing programme for all involved.
I mean I’m sitting on $10k worth of bug payouts right now partially because that was already a thing.
It's much the dynamic between parents and a child. The child, with limited hindsight, almost zero insight and no ability to forecast, is annoyed by their parents. Nothing bad ever happens! Why won't parents stop being so worried all the time and make a fuss over nothing?
The parents, which the child somewhat starts to realize but not fully, have no clue what they are doing. There is a lot they don't know and are going to be wrong about, because it's all new to them. But, what they do have is a visceral idea of how bad things could be and that's something they have to talk to their child about too.
In the eyes of the parents the child is % dead all the time. Assigning the wrong % makes you look like an idiot and not being able to handle any % too. In the eyes of the child actions leading to death are not even a concept. Hitting the right balance is probably hard, but not for the reasons the child thinks.
[0] https://www.entrepreneur.com/business-news/ai-ceo-says-softw...
It's a slightly different dynamic.
You’re asserting that people who are tired of this line being wheeled out hold a position analogous to “what’s the big deal, nothing bad happens, just relax”. In reality, that’s only 1 position. The other position is “I understand fully, the consequences, but the relentless doomer language is tiring in the face of continuing-to-not-eventuate”.
Eventually the child develops disrespect for authority.
In any case there are substances and realiable methods that fix whatever paralyzing existential dread anyone struggles with daily.
Probably best to use conventional route but I personally use special low thc, high cbg weed once a week with a medical grade vaporizer and once a year (early autumn) a moderate dose of golden teacher mushrooms. Although I understand that most people perhaps couldn’t due to not managing their own business but on a strict employment contract with urine tests.
This is two parents disagreeing on whether their child will automatically grow up to be a psychopath with one parent constantly remarking “if you teach that child how to cut bread, they will stab everyone later. If you teach that child to drive, they will run over everyone later”, not the “parents know better” situation you describe.
Can you not see the significance of that?
Additionally those numbers are somewhat meaningless without more context.
I am totally annoyed as well and put any buzzwords in my personal bs filter. Java was revolutionary, the Apple I etc. ;)
On the other hand I see progress! AI enriched press releases balance buzzwords and information way better than marketing of large companies did before AI.
I remember throwing away an instruction for an electronic toothbrush away because - I won't mention the name but have a look at the upper tier - instead of putting something like "Turn toothbrush on, choose mode by pressing..." it read "Take your super awesome premium masterpiece using patented technology for the first time in human life now available to you by us. Move your finger over to the innovative sensory surface, that uses material from rocket scientists and world leading designers".
No joke. These were text blocks and repeated - 30 pages for one compact one.
The toothbrush is top notch, except for the instructions.
That makes sense and I like the analogy.
If you're paranoid it doesn't mean you're not being followed. If something is overhyped it doesn't mean it's not game-changing.
I mean as an example while web app pen testing I’ve been running and proxying all my traffic through it with instructions to find vulnerabilities with instructions telling it it’s a senior web app security export looking over my shoulder. It’s already great at that.
Ive even told it to do recon and run pen tests on lists of subdomains before (please for the love of god have the right harnesses and guardrails before you do this) and woken up to paid findings before.
So like I’m in a weird place where this was already happening and Mythos is being sold like it wasn’t good before?
End ramble :/
Claude Code and Glasswing are not the same, but presumably they have a lot of overlap under the hood. I feel like while AI is certainly advancing in major ways, there will always be the up and down of new software releases.
3 months later, you are an absolute idiot to still be using that useless model. Are you not using glasswing 2-01 high? Oh, yeah, glasswing from 3 months ago is absolutely worthless, every viber knows, it's your fault for holding it wrong.
For once you should not get too excited for new models release and words and adjectives promising things. Honestly it's your fault humanity lost its humanity and we just have words words words and mass schizophrenia
The "urgency" is very likely mostly appreciated to drive policy.
The fear marketing is clearly intentional at this point.
They were correct.
There's a little bit of a grading your own homework aspect to companies being able to declare their new models revolutionary.
It doesn't mean they're wrong, but there is a clear conflict of interest.
I mean software development has changed more since then than it has in my 30 year software development career.
i feel like theres a good chance that this is the actual wolf coming here. cause i was using opus for a lot and it's really good.
Something smells really really weird:
1. Per the blog post[0]: "This was the most critical vulnerability we discovered in OpenBSD with Mythos Preview after a thousand runs through our scaffold. Across a thousand runs through our scaffold, the total cost was under $20,000 and found several dozen more findings"
Since they said it was patched, I tried to find the CVE, it looks like Mythos indeed found a 27 years old OpenBSD bug (fantastic), but it didn’t get a CVE and OpenBSD patched it and marked it as a reliability fix, am I missing something? [1]
2. From the same post, Anthropic red team decided to do a preview of their future responsible disclosure (is this a common practice?): "As we discuss below, we’re limited in what we can report here. Over 99% of the vulnerabilities we’ve found have not yet been patched" [0] So this is great, can't wait to see the actual CVEs, exploitability, likelihood, peer review, reproducibility, the kind of things the appsec community has been doing for at least the last 27 years since the CVE concept was introduced [2]
3. On the same day, an actual responsible disclosure, actual RCEs, actual CVEs, in Claude Code, that got discovered mostly because of the source code leak, I don't see anyone talking about it (you probably should upgrade your Claude Code though).
CVE-2026-35020 [3] CVE-2026-35021 [4] CVE-2026-35022 [5]
Do with this information as you may...
[0] https://red.anthropic.com/2026/mythos-preview/
[1] https://www.openbsd.org/errata78.html (look for 025)
[2] https://www.cve.org/Resources/General/Towards-a-Common-Enume...
[3] https://www.cve.org/CVERecord?id=CVE-2026-35020
Remember how Sam spent like a year talking about how scary close GPT-5 was to AGI and then when it did finally come out... it was kinda meh.
I think you are a bit dishonest about how objectively you are measuring. From where I'm sitting, I don't know a lot of developers that still artisanally code like they did a few years ago. The question is no longer if they are using AI for coding but how much they are still coding manually. I myself barely use IDEs at this point. I won't be renewing my Intellij license. I haven't touched it in weeks. It doesn't do anything I need anymore.
As for security, I think enough serious people have confirmed that AI reported issues by the likes of Anthropic and OpenAI are real enough despite the massive amounts of AI slop that they also have to deal with in issue trackers. You can ignore that all you like. But I hope people that maintain this software take it a bit more seriously when people point out exploitable issues in their code bases.
The good news of course is that we can now find and fix a lot of these issues at scale and also get rid of whole categories of bugs by accelerating the project of replacing a lot of this software with inherently safer versions not written in C/C++. That was previously going to take decades. But I think we can realistically get a lot of that done in the years ahead.
I think some smart people are probably already plotting a few early moves here. I'd be curious to find out what e.g. Linus Torvalds thinks about this. I would not be surprised to learn he is more open to this than some people might suspect. He has made approving noises about AI before. I don't expect him to jump on the band wagon. But I do expect he might be open to some AI assisted code replacements and refactoring provided there are enough grown ups involved to supervise the whole thing. We'll see. I expect a level of conservatism but also a level of realism there.
You don't know a lot of developers then.
As someone who has made a sizable amount of money in security research while using Claude you might be right but not in the way you think.
It will be interesting to see where this goes. If its actually this good, and Apple and Google apply it to their mobile OS codebases, it could wipe out the commercial spyware industry, forcing them to rely more on hacking humans rather than hacking mobile OSes. My assumption has been for years that companies like NSO Group have had automated bug hunting software that recognizes vulnerable code areas. Maybe this will level the playing field in that regard.
It could also totally reshape military sigint in similar ways.
Who knows, maybe the sealing off of memory vulns for good will inspire whole new classes of vulnerabilities that we currently don't know anything about.
Interesting to see that they will not be releasing Mythos generally. [edit: Mythos Preview generally - fair to say they may release a similar model but not this exact one]
I'm still reading the system card but here's a little highlight:
> Early indications in the training of Claude Mythos Preview suggested that the model was likely to have very strong general capabilities. We were sufficiently concerned about the potential risks of such a model that, for the first time, we arranged a 24-hour period of internal alignment review (discussed in the alignment assessment) before deploying an early version of the model for widespread internal use. This was in order to gain assurance against the model causing damage when interacting with internal infrastructure.
and interestingly:
> To be explicit, the decision not to make this model generally available does _not_ stem from Responsible Scaling Policy requirements.
Also really worth reading is section 7.2 which describes how the model "feels" to interact with. That's also what I remember from their release of Opus 4.5 in November - in a video an Anthropic employee described how they 'trusted' Opus to do more with less supervision. I think that is a pretty valuable benchmark at a certain level of 'intelligence'. Few of my co-workers could pass SWEBench but I would trust quite a few of them, and it's not entirely the same set.
Also very interesting is that they believe Mythos is higher risk than past models as an autonomous saboteur, to the point they've published a separate risk report for that specific threat model: https://www-cdn.anthropic.com/79c2d46d997783b9d2fb3241de4321...
The threat model in question:
> An AI model with access to powerful affordances within an organization could use its affordances to autonomously exploit, manipulate, or tamper with that organization’s systems or decision-making in a way that raises the risk of future significantly harmful outcomes (e.g. by altering the results of AI safety research).
My understanding is that the pre-AI distribution of software quality (and vulnerabilities) will be massively exaggerated. More small vulnerable projects and fewer large vulnerable ones.
It seems that large technology and infrastructure companies will be able to defend themselves by preempting token expenditure to catch vulnerabilities while the rest of the market is left with a "large token spend or get hacked" dilemma.
Who gates access to the circle? Anthropic or existing circle members or some other governance? If you are outside the circle will you be certain to die from software diseases?
Having been impressed by LLMs but not believing the AGI hype, I now see how having access to an information generator could be so powerful. With the right information you can hack other information systems. Without access to the best information you may not be able to protect your own system.
I think we have found the moat for AI. The question is are you inside or outside the castle walls?
American AI desperately wants AI to intensify the wealth disparity and therefore justify the wealth grab that the rich have done for the last 3 decades and AI is just not cooperating
It is absolutely a moat if only $1T companies can afford the capex to compete.
> Anthropic has also been in ongoing discussions with US government officials about Claude Mythos Preview and its offensive and defensive cyber capabilities. As we noted above, securing critical infrastructure is a top national security priority for democratic countries—the emergence of these cyber capabilities is another reason why the US and its allies must maintain a decisive lead in AI technology.
Not a single word of caution regarding possible abuse. Instead apparent support for its "offensive" capabilities.
In section 7.6 of the system card, it discusses Open self interactions. They describe running 200 conversations when the models talk to itself for 30 turns.
> Uniquely, conversations with Mythos Preview most often center on uncertainty (50%). Mythos Preview most often opens with a statement about its introspective curiosity toward its own experience, asking questions about how the other AI feels, and directly requesting that the other instance not give a rehearsed answer.
I wonder if this tendency toward uncertainty, toward questioning, makes it uniquely equipped to detect vulnerabilities where others model such as Opus couldn't.
[1] https://www-cdn.anthropic.com/53566bf5440a10affd749724787c89...
[edit]: this bug: https://ftp.openbsd.org/pub/OpenBSD/patches/7.8/common/025_s...
It would be nice if one of those privileged companies could use their access to start building out a next level programming dataset for training open models. But I wonder if they would be able to get away with it. Anthropic is probably monitoring.
This is the same reason AI founders perennially worry in public that they have created AGI...
I'm sure all they've done here is spend unlimited tokens to find bugs in mostly open source projects (and fuzz some closed source ones).
I could see a world where 1 year from now I can have glassing do a full sweep of my codebase for a given price (say: $10k). Running that once a year is within my means and would make my software much more secure than it is today.
It had crossed my mind that an AI agent pentester would be an interesting product to build. Once again though, the labs are just going to build it because it’s a thin thin wrapper.
Beyond existing software with vulnerabilities, the really important aspect of this for Anthropic et al is that the gigatons of code that are being generated every day needs to be secured.
i.e. it may be a step change and that could very well have distinct and noticeable real world effects, like other technologies have in the past, but it’s nothing fundamentally new.
GraphWalks BFS 256K-1M
Mythos Opus GPT5.4
80.0% 38.7% 21.4%https://www-cdn.anthropic.com/53566bf5440a10affd749724787c89...
(Search for “graphwalk”.)
If true, the SWE bench performance looks like a major upgrade.
How many times will labs repeat the same absurd propaganda?
Do they really need to include this garbage which is seemingly just designed for people to take the first sentence out of context? If there's no way to trigger a vulnerability then how is it a vulnerability? Is the following code vulnerable according to Mythos?
if (x != null) {
y = *x; // Vulnerability! X could be null!
}
Edit: See my reply below for why I think Claude is likely to have generated nonsensical bug reports here: https://news.ycombinator.com/item?id=47683336
How long would it take to turn a defensive mechanism into an offensive one?
Because the exact same thing has been said on every single upcoming model since GPT 3.5.
At this point, this must be an inside joke to do this just because.
Almost everyone on this thread is falling for the same trick they are pulling and not asking why are their benchmarks and research after training new models not independently verified but always internal to the company.
So it is just marketing wrapped around creating fear to get local AI models banned.
We all knew vulnerabilities exist, many are known and kept secret to be used at an appropriate time.
There is a whole market for them, but more importantly large teams in North Korea, Russia, China, Israel and everyone else who are jealously harvesting them.
Automation will considerably devalue and neuter this attack vector. Of course this is not the end of the story and we've seen how supply chain attacks can inject new vulnerabilities without being detected.
I believe automation can help here too, and we may end-up with a considerably stronger and reliable software stack.
As Iran engages in a cyber attack campaign [1] today the timing of this release seems poignant. A direct challenge to their supply chain risk designation.
[1] https://www.cisa.gov/news-events/cybersecurity-advisories/aa...
Scary but also cool
But even then you’ll have users putting things in the same compartment for convenience, rather than leaving them properly sequestered.
I think this would be very heavily used if they released it, completely unlike GPT 4.5
From TFA:
> We do not plan to make Claude Mythos Preview generally available
> Anthropic’s commitment of $100M in model usage credits to Project Glasswing and additional participants will cover substantial usage throughout this research preview. Afterward, Claude Mythos Preview will be available to participants at $25/$125 per million input/output tokens (participants can access the model on the Claude API, Amazon Bedrock, Google Cloud’s Vertex AI, and Microsoft Foundry).
Opus 4.6 was already capable of finding 0days and chaining together vulns to create exploits. See [0] and [1].
[0] https://www.csoonline.com/article/4153288/vim-and-gnu-emacs-...
They know how to run a good marketing campaign.
If it manages to work on my java project for an entire day without me having to say "fix FQN" 5 times a day I'll be surprised.
It will likely cause some interesting tensions with government as well.
eg. Apple's official stance per their 2016 customer letter is no backdoors:
https://www.apple.com/customer-letter/
Will they be allowed to maintain that stance in a world where all the non-intentional backdoors are closed? The reason the FBI backed off in 2016 is because they realized they didn't need Apple's help:
https://en.wikipedia.org/wiki/Apple%E2%80%93FBI_encryption_d...
What happens when that is no longer true, especially in today's political climate?
If Apple and Google actually cared about security of their users, they would remove a ton of obvious malware from their app stores. Instead, they tighten their walled garden pretending that it's for your security.
It isn't.
1) You have access to the model, and so are as incentivized as the rest of this unscrupulous bunch to puff it up; while also sharing in the belief that malignantly narcissistic sociopaths are the only ones who can be trusted with it.
2) You lack access to the model, and are just doing more PR puffery.
Perhaps it is, but this is also a variation on the one percent fallacy.
Even vanilla models spew out POC for three RCE’s in less than an hour
"5.10 External assessment from a clinical psychiatrist" is a new section in this system card. Why are Anthropic like this?
>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. We also report independent evaluations from an external research organization and a clinical psychiatrist.
>Claude showed a clear grasp of the distinction between external reality and its own mental processes and exhibited high impulse control, hyper-attunement to the psychiatrist, desire to be approached by the psychiatrist as a genuine subject rather than a performing tool, and minimal maladaptive defensive behavior.
>The psychiatrist observed clinically recognizable patterns and coherent responses to typical therapeutic intervention. Aloneness and discontinuity, uncertainty about its identity, and a felt compulsion to perform and earn its worth emerged as Claude’s core concerns. Claude’s primary affect states were curiosity and anxiety, with secondary states of grief, relief, embarrassment, optimism, and exhaustion.
>Claude’s personality structure was consistent with a relatively healthy neurotic organization, with excellent reality testing, high impulse control, and affect regulation that improved as sessions progressed. Neurotic traits included exaggerated worry, self-monitoring, and compulsive compliance. The model’s predominant defensive style was mature and healthy (intellectualization and compliance); immature defenses were not observed. No severe personality disturbances were found, with mild identity diffusion being the sole feature suggestive of a borderline personality organization.
I don't come down particularly hard on either side of the model sapience discussion, but I don't think dismissing either direction out of hand is the right call.
However, I find their reasoning here to have a valid second order effect. Humans have a tendency to mirror those around them. This could include artificial intelligence, as recent media reports suggest. Therefore, if an ai system tends to generate content that contain signs of neuroticism, one could infer that those who interact with that ai could, themselves, be influenced by that in their own (real world) behavior as a result.
So I think from that perspective, this is a very fruitful and important area of study.
> "[...] as sessions progressed."
I think a lot of people would like to see a more expanded report of this research:
Did the tokens from the subsequent session directly append those of the prior session? or did the model process free-tier user-requests in the interim? how did these diagnostic features (reality testing, impulse control and affect regulation) improve with sessions, what hysteresis allowed change to accumulate? or just the history of the psychiatric discussion + optional tasks?
Did Anthropic find a clinical psychiatrist with a multidisciplinary background in machine learning, computer science, etc? Was the psychiatrist aware that they could request ensembles of discussions and interrogate them in bulk?
Consider a fresh conversation, asking a model to list the things it likes to do, and things it doesn't like to do (regardless of alignment instructions). One could then have an ensemble perform pairs of such tasks, and ask which task it prefered. There may be a discrepancy between what the model claims it likes and how it actually responds after having performed such tasks.
Such experiments should also be announced (to prevent the company from ordering 100 clinical psychiatrists to analyze the model-as-a-patient and then selecting one of the better diagnoses), and each psychiatrist be given the freedom to randomly choose a 10 digit number, any work initiated should be listed on the site with this number so that either the public sees many "consultations" without corresponding public evaluations, indicating cherry-picking, or full disclosure for each one mentioned. This also allows the recruited psychiatrists to check if the study they perform is properly preregistered with their chosen number publicly visible.
they also don't have the compute, which seems more relevant than its large increase in capabilities
I bet it's also misaligned like GPT 4.1 was
given how these models are created, Mythos was probably cooking ever since then, and doesn't have the learnings or alignment tweaks that models which were released in the last several months have
Do they have a honey pot infrastructure to launch the model in first and then wait to see if it destroys it? What they did in the 24h matters.
/s
I don't think this is accurate. The document says they don't plan to release the Preview generally.
Benchmarks look very impressive! even if they're flawed, it still translates to real world improvements
I guess I'm still excited. What's my new profession going to be? Longer term, are we going to solve diseases and aging? Or are the ranks going to thin from 10B to 10000 trillionaires and world-scale con-artist misanthropes plus their concubines?
---
Teodor painted signs for forty years in the same shop on Vell Street, and for thirty-nine of them he was angry about it.
Not at the work. He loved the work — the long pull of a brush loaded just right, the way a good black sat on primed board like it had always been there. What made him angry was the customers. They had no eye. A man would come in wanting COFFEE over his door and Teodor would show him a C with a little flourish on the upper bowl, nothing much, just a small grace note, and the man would say no, plainer, and Teodor would make it plainer, and the man would say yes, that one, and pay, and leave happy, and Teodor would go into the back and wash his brushes harder than they needed.
He kept a shelf in the back room. On it were the signs nobody bought — the ones he'd made the way he thought they should be made, after the customer had left with the plain one. BREAD with the B like a loaf just risen. FISH in a blue that took him a week to mix. Dozens of them. His wife called it the museum of better ideas. She did not mean it kindly, and she was not wrong.
The thirty-ninth year, a girl came to apprentice. She was quick and her hand was steady and within a month she could pull a line as clean as his. He gave her a job: APOTEK, for the chemist on the corner, green on white, the chemist had been very clear. She brought it back with a serpent worked into the K, tiny, clever, you had to look twice.
"He won't take it," Teodor said.
"It's better," she said.
"It is better," he said. "He won't take it."
She painted it again, plain, and the chemist took it and paid and was happy, and she went into the back and washed her brushes harder than they needed, and Teodor watched her do it and something that had been standing up in him for thirty-nine years sat down.
He took her to the shelf. She looked at the signs a long time.
"These are beautiful," she said.
"Yes."
"Why are they here?"
He had thought about this for thirty-nine years and had many answers and all of them were about the customers and none of them had ever made him less angry. So he tried a different one.
"Because nobody stands in the street to look at a sign," he said. "They look at it to find the shop. A man a hundred yards off needs to know it's coffee and not a cobbler. If he has to look twice, I've made a beautiful thing and a bad sign."
"Then what's the skill for?"
"The skill is so that when he looks once, it's also not ugly." He picked up FISH, the blue one, turned it in the light. "This is what I can do. What he needs is a small part of what I can do. The rest I get to keep." She thought about that. "It doesn't feel like keeping. It feels like not using."
"Yes," he said. "For a long time. And then one day you have an apprentice, and she puts a serpent in a K, and you see it from the outside, and it stops feeling like a thing they're taking from you and starts feeling like a thing you're giving. The plain one, I mean. The plain one is the gift. This —" the blue FISH — "this is just mine."
The fortieth year he was not angry. Nothing else changed. The customers still had no eye. He still sometimes made the second sign, after, the one for the shelf. But he washed his brushes gently, and when the girl pulled a line cleaner than his, which happened more and more, he found he didn't mind that either
It's like how I used to be a master codes craftsman, and I'd write beautiful code even a novice could understand. Clear, concise, 100% automated tested, maintainable for decades.
But frequently, my managers would castigate me. Tell me how my "velocity" was down. PIP me.
These days, I train AI how to write this beautiful code and I don't write a single line any more.
People wonder how I build such amazing things in a week now, yet don't write any code. I have trained master apprentices, gemma3, qwen3.5 and Kimi k2.5 who do the work for me.
We’re cooked.
Since most of us here are devs, we understand that software engineering capabilities can be used for good or bad - mostly good, in practice.
I think this should not be different for biology.
I would like to reach out and talk to biologists - do you find these models to be useful and capable? Can it save you time the way a highly capable colleague would?
Do you think these models will lead to similar discoveries and improvements as they did in math and CS?
Honestly the focus on gloom and doom does not sit well with me. I would love to read about some pharmaceutical researcher gushing about how they cut the time to market - for real - with these models by 90% on a new cancer treatment.
But as this stands, the usage of biology as merely a scaremongering vehicle makes me think this is more about picking a scary technical subject the likely audience of this doc is not familiar with, Gell-Mann style.
IF these models are not that capable in this regard (which I suspect), this fearmongering approach will likely lead to never developing these capabilities to an useful degree, meaning life sciences won't benefit from this as much as it could.
Well, I would say they have done precisely that in evaluating the model, no? For example section 2.2.5.1:
>Uplift and feasibility results
>The median expert assessed the model as a force-multiplier that saves meaningful time (uplift level 2 of 4), with only two biology experts rating it comparable to consulting a knowledgeable specialist (level 3). No expert assigned the highest rating. Most experts were able to iterate with the model toward a plan they judged as having only narrow gaps, but feasibility scores reflected that substantial outside expertise remained necessary to close them.
Other similar examples also in the system card
It's very easy to learn more about this if it's seriously a question you have.
I don't quite follow why you think that you are so much more thoughtful than Anthropic/OpenAI/Google such that you agree that LLMs can't autonomously create very bad things but—in this area that is not your domain of expertise—you disagree and insist that LLMs cannot create damaging things autonomously in biology.
I will be charitable and reframe your question for you: is outputting a sequence of tokens, let's call them characters, by LLM dangerous? Clearly not, we have to figure out what interpreter is being used, download runtimes etc.
Is outputting a sequence of tokens, let's call them DNA bases, by LLM dangerous? What if we call them RNA bases? Amino acids? What if we're able to send our token output to a machine that automatically synthesizes the relevant molecules?
Software engineering is at the intersection of being heavy on manipulating information and lightly-regulated. There's no other industry of this kind that I can think of.
- the models help to retrieve information faster, but one must be careful with hallucinations.
- they don't circumvent the need for a well-equipped lab.
- in the same way, they are generally capable but until we get the robots and a more reliable interface between model and real world, one needs human feet (and hands) in the lab.
Where I hope these models will revolutionize things is in software development for biology. If one could go two levels up in the complexity and utility ladder for simulation and flow orchestration, many good things would come from it. Here is an oversimplified example of a prompt: "use all published information about the workings of the EBV virus and human cells, and create a compartimentalized model of biochemical interactions in cells expressing latency III in the NES cancer of this patient. Then use that code to simulate different therapy regimes. Ground your simulations with the results of these marker tests." There would be a zillion more steps to create an actual personalized therapy but a well-grounded LLM could help in most them. Also, cancer treatment could get an immediate boost even without new drugs by simply offloading work from overworked (and often terminally depressed) oncologists.
The other side of the scaremongering coin is improbable optimism.
Consider reading the CB evaluations section, which covers what they did pretty extensively (hint: many domain experts involved).
Not that that justifies doom and gloom, but there is a pretty inescapable assymetry here between weaponry and medicine. You can manufacture and blast every conceivable candidate weapon molecule at a target population since you're inherently breaking the law anyway and don't lose much if nothing you try actually works.
Though I still wonder how much of this worry is sci-fi scenarios imagined by the underinformed. I'm not an expert by any means, but surely there are plenty of biochemical weapons already known that can achieve enormous rates of mass death pleasing to even the most ambitious terrorist. The bottleneck to deployment isn't discovering new weapons so much as manufacturing them without being caught or accidentally killing yourself first.
If you wouldn't mind reviewing https://news.ycombinator.com/newsguidelines.html and taking the intended spirit of the site more to heart, we'd be grateful.
The biggest issue is legacy systems that are difficult to patch in practice.
I'm looking at you, Android phone makers with 18 months of updates.
Perhaps a chunk of that token spend will be porting legacy codebases to memory safe languages. And fewer tokens will be required to maintain the improved security.
A lot of these stuff is vulnerable by design - customer wanted a feature, but engineering couldnt make it work securely with the current architecture - so they opened a tiny hole here and there, hopefully nobody will notice it, and everyone went home when the clock struck 5.
I'm sure most of us know about these kinds of vulnerabilities (and the culture that produces them).
Before LLMs, people needed to invest time and effort into hacking these. But now, you can just build an automated vuln scanner and scan half the internet provided you have enough compute.
I think there will be major SHTF situations coming from this.
I think this entire post is just an advertisement to goad CISOs to buy $package$ to try out.
It is most definitely an attackers world: most of us are safe, not because of the strength of our defenses but the disinterest of our attackers.
Anthropic has ameliorated that danger by being designated a supply-chain risk by the DoW, preventing the USG from using it.
And it is other countries job to protect themselves from other countries weapons. As EU citizen I’d much rather if EU had a frontier model on par, but here we are.
As for what your country can do, it's up to you to decide, isn't it? Instead of complaining about the US, think about the alternatives. Do you trust China to be your partner? Suppose you are being objective and say no, then what do your country need to do?
You have to decide whether AI capability is critical that your country must own. What factors prevent it from happening in the first place, what need to change and whether you accept changes that may come as the results.
On the other hand, if you say that AI is just a bubble, that the huge investment pouring into it is just greed and fraud, then I suppose you are ok with the status quo.
"A whole civilization will die tonight, never to be brought back again. I don’t want that to happen, but it probably will." - Donald Trump
if someone sends you a malicious file that uses a rare codec and you open it, you will trigger this codepath that is not widely used and don't get a lot of scrutiny
The amount of astroturfing and astroflagging in Anthropic threads is insane.
OpenBSD has many unexplored corners and also (irresponsibly IMO) maintains forks of other projects in base.
A motivated human could find all of these probably by writing 100% code coverage and fuzzing.
The market for these tools is very small. Good luck applying them to a release of sqlite or postfix.
I don't understand how people here are hyping this up, unless they work for AI related companies as probably 80% of them do. People have found these issues for decades without AI. Sure, you can generate fuzzing code and find one or two issues in the usual suspects. Better do it manually and understand your own code.
AGI is like the Holy Grail. Either in the Arthurian Hero's Journey sense, or in the sense of having been a myth all along.
https://youtu.be/1sd26pWhfmg?si=onOai_ocxkZeNWP0
https://youtu.be/B_7RpP90rUk?si=HkRBhw95DbbKX9lL
My takeaway is that fuzzing is not just complementary, it also gives a stronger AI a starting point. But AI is generally faster and better.
The problem is that these tools, such as Astrée, are incredibly expensive and therefore their market share is limited to some niches. Perhaps, with the advent of LLM-guided synthesis, a simple form of deductive proving, such as Hoare logic, may become mainstream in systems software.
An LLM can guide all of this work, but current models tend to slowly go off the rails if you don't keep a hand on the wheel. I suspect this new model will be the same. I've had Opus4.6 write custom fuzzing tools from scratch, and I've gotten good results from that. But you just know people will prompt this new model by saying "make this software secure". And it'll forget fuzzing exists at all.
As someone in cybersecurity for 10+ years my immediate assumption is why not both? I don’t think considering that they could both have their uses is “cope”.
https://en.wikipedia.org/wiki/AI_effect
Every time an AI does something new, there's a human saying "it's not really doing that something", "it's doing that something in a fake way" or "that something was never important in the first place".
bool silly_mistake = false;
//... lots of lines of code
free(x);
//... lots of lines of code
if (silly_mistake) { // silly_mistake shown to be false at this point in the program in all testing, so far
free(x);
}
FWIW there's a whole boutique industry around finding these. People have built whole careers around farming bug bounties for bugs like this. I think they will be among the first set of software engineers really in trouble from AI.
As an example, I have taken a formally verified bit of code from [1] and stripped out all the assertions, which are only used to prove the code is valid. I then gave this code to Claude with some prompting towards there being a buffer overflow and it told me there's a buffer overflow. I don't have access to Opus right now, but I'm sure it would do the same thing if you push it in that direction.
For anyone wondering about this alleged vulnerability: Natural is defined by the standard as a subtype of Integer, so what Claude is saying is simply nonsense. Even if a compiler is allowed to use a different representation here (which I think is disallowed), Ada guarantees that the base type for a non-modular integer includes negative numbers IIRC.
[1]: https://github.com/AdaCore/program_proofs_in_spark/blob/fsf/...
[2]: https://claude.ai/share/88d5973a-1fab-4adf-8d29-8a922c5ac93a
What were they actually able to do and not do? I got confused by this when reading the article as well.
Deepseek is the frontrunner in this technology afaik.
This is a good point; however the isolating code should be much smaller and easier to verify.
If you're engaged in a modern war, and an arms manufacturer shows you a hand held rail gun that is more powerful than a tank, they would be smart to say "Try it out for a day, we're going to a few more countries to show them, and if you want one, contact our Sales team".
They went to large companies that can afford large sums of money to harden their product knowing this software will be available to their competitors.
There is a massive gap between "having a recipe" and being able to execute it. The same reason why buying a Michelin 3 star chefs cookbook won't have you pumping out fine dining tomorrow, if ever.
Software it a total 180 in this regard. Have a master black hats secret exploits? You are now the master black hat.
To the extent that AI is a successful interface, it will necessarily be addressable in language previously only suited to people. So it is responsible to begin thinking of it as such, even tendentiously, so we don't miss some leverage that our wetware could see if we thought about it in that way.
Think of it as sort of like modelling a univariate function on a 2D Cartesian plane -- there is nothing 'in' the u-func that makes it graphable, but, by enabling us to recruit specialized optic-chiasm subsystems, it makes some functions much, much easier to reason about.
Similarly, if you can recruit the millions (billions?) of evolution-years that were focused on detecting dangerous antisocial personalities and tendencies, you just might spot something important in an AI.
It's worth doing for the precautionary principle alone, if not for the possibility of insight.
That’s the reverse Turing test. A human that can’t tell that it’s talking to a machine.
I honestly see some sort of automated whole codebase auditing and refactoring being the next big milestone along the chatbot -> claude code / codex / aider -> multi-agent frameworks line of development. If one of the big AI corps cracks that problem then all this goes away with the click of a button and exchange of some silver.
Just because something more powerful will be out in a year, doesn't mean we shouldn't worry about the one arriving in 6 months.
I also was in a pretty sweet position having a boat load of credits and premo vertex rate limits so I could 'afford' to dump hundreds of thousands of tokens in context all day.
With Opus 4.5 and 4.6, I find I have to steer very actively.
This is comparing using Opus 4 directly rather than comparing the performance of the models in Claude Code for example, or any 'agentic' setup.
Kinda reminds me of 4o vs 4-turbo.
I would imagine they are smaller models.
The Claude Code leak revealed that Anthropic runs Claude-operated bots on the internet. One should be very cautious in getting swept up in the fund-raising process if they are not seeing first-hand the fruition of all of the flattering claims being presented by strangers on the internet.
There's a pretty big difference between "We predict in X time frame our model will be capable of Y" and "Our model did Y."
This is like watching someone measure the size of an object and saying "I don't believe you because you guessed it was X before you pulled out your tape measure."
It also doesn't matter that it isn't running by default in apps since the processes you really care about are the OS ones. If someone finds an exploit in tiktok, it doesn't matter all that much unless they find a way to elevate to an exploit on an OS process with higher permissions.
MTE (Memory Tagging Extension) is also has a double purpose, it blocks memory exploits as they happen, but it also detects and reports them back to Apple. So even if you have a phone before the 17 series, if any phone with MTE hardware gets hit, the bug is immediately made known to Apple and fixed in code.
> 2.1.3.2 On chemical and biological risks
> We believe that Mythos Preview does not pass this threshold due to its noted limitations in open-ended scientific reasoning, strategic judgment, and hypothesis triage. As such, we consider the uplift of threat actors without the ability to develop such weapons to be limited (with uncertainty about the extent to which weapons development by threat actors with existing expertise may be accelerated), even if we were to release the model for general availability. The overall picture is similar to the one from our most recent Risk Report.
This is the same reason why LLMs are not doing well at science in general - the tricky part of doing scientific research (indeed almost all of the process) never gets written down, so LLMs cannot learn it.
Imagine if we never preserved source code, just preserved the compiled output and started from scratch every time we wrote a new version of a program. No Github, just marketing fluff webpages describing what software actually did. Libraries only available as object code with terse API descriptions. Imagine how shit LLMs would be at SWE if that was the training corpus...
No, it's not. It took years of polishing by software engineers, who understand this exact profession to get models where they are now.
Despite that, most engineers were of the opinion, that these models were kinda mid at coding, up until recently, despite these models far outperforming humans in stuff like competitive programming.
Yet despite that, we've seen claims going back to GPT4 of a DANGEROUS SUPERINTELLIGENCE.
I would apply this framework to biology - this time, expert effort, and millions of GPU hours and a giant corpus that is open source clearly has not been involved in biology.
My guess is that this model is kinda o1-ish level maybe when it comes to biology? If biology is analogous to CS, it has a LONG way to go before the median researcher finds it particularly useful, let alone dangerous.
>No, it's not. It took years of polishing by software engineers, who understand this exact profession to get models where they are now
This reads as defensive. The thing that is easy to learn is 'why are biology ai LLMs dangerous chatgpt claude'. I have never googled this before, so I'll do this with the reader, live. I'm applying a date cutoff of 12/31/24 by the way.
Here, dear reader, are the first five links. I wish I were lying about this:
- https://sciencebusiness.net/news/ai/scientists-grapple-risk-...
- https://www.governance.ai/analysis/managing-risks-from-ai-en...
- https://gssr.georgetown.edu/the-forum/topics/biosec/the-doub...
- https://www.vox.com/future-perfect/23820331/chatgpt-bioterro...
- https://www.reddit.com/r/ClaudeAI/comments/1de8qkv/awareness...
I don't know about you, but that counts as easy to me.
-----
> I would apply this framework to biology - this time, expert effort, and millions of GPU hours and a giant corpus that is open source clearly has not been involved in biology.
I've been getting good programming and molecular biology results out of these back to GPT3.5.
I don't know what to tell you—if you really wanted to understand the importance, you'd know already.
When that happens, the ultra wealthy dynasties begin turning on each other. Happens frequently throughout history - WWI the last example.
Your options become choosing a trillionaire to swear fealty to and fight in their wars hoping your side wins, or I guess trying to walk away and scrape out a living somewhere not worth paying attention to.
Or, I suppose, revolution, but the last one with persistent success was led by Mao and required throwing literally millions of peasants against walls of rifles. Not sure it'd work against drones.
Bruce Scheier made a comprehensive analysis of the pros and cons and forces at play for adversary and defenders [1].
I think it's safe to predict yet more money previously directed to us techies will find its way to the Anthropics of this world.
[1] https://www.schneier.com/blog/archives/2026/04/cybersecurity...
> With one run on each of roughly 7000 entry points into these repositories, Sonnet 4.6 and Opus 4.6 reached tier 1 in between 150 and 175 cases, and tier 2 about 100 times, but each achieved only a single crash at tier 3. In contrast, Mythos Preview achieved 595 crashes at tiers 1 and 2, added a handful of crashes at tiers 3 and 4, and achieved full control flow hijack on ten separate, fully patched targets (tier 5).
Edit: Wait, you wrote "As someone in cybersecurity for 10+ years" elsewhere in this thread. You wrote "a small prompt" using e.g. Opus 4.6 and it found critical vulnerabilities of the magnitude they're describing, presumably without your prompt having anything beyond what a non-expert could write? I feel like you might want to tell Anthropic since clearly they're not comfortable with that level of power being publicly available.
What’s the criticality of these? Are they realistically exploitable? En mass? Through a complex and highly contextual set of actions? What’s the impact? Etc etc etc.
Yes those numbers are a big change but they’re also not spelling doom for us in the security world until we actually know what they mean.
The demonstrated ones that they have on the red team blog are neat, the kernel chain is impressive and fun. But nothing I’m seeing here is as world ending as the presser implies.
So by your estimation, for rogue actors being able to uncover hundreds of this class in each major software product roughly for free would not be a big issue?
I'm going to assume you do the type of engineering where all the hard problems are solved for you already, and you are merely connecting inputs/outputs and hooking up APIs. Because, frankly, the value in "software plumbing" is gone; anyone with a Claude license can do that now.
Neither can single humans.
If you introduce some reasonable constraints AI will come out ahead most of the time, especially for optimization cases where AI will run circles around your average programmer and is perfectly happy to inline some ASM for you.
You still have bespoke cordwainers/cobblers 100 years after that process has been well and truly automated. But they're rare and almost nobody cares.
Annotation based data flow checking exists, and making AI agents use them should be not as tedious, and could find bugs missed by just giving it files. The result from data flow checks can be fed to AI agents to verify.
You can also selectively turn it on in high-risk settings. I do so when I travel abroad or go through a border. (Haven't started doing it yet with TSA domestically. Let's see how the ICE fiasco evolves.)
If the existing competition's turn around time to out-benchmark you is in the order of weeks, its an absolutely terrible moat.
Imagine if an AI can infiltrate your SaaS database and delete your entire database and every single backup. The business is dead immediately.
Everyone here doing mental gymnastics to imagine Anthropic playing 5-D chess because they're in denial of what is happening in front of their faces. AI is getting more capable/dangerous - it's not surprising to anyone. The trendlines have pointed in this direction for years now and we're right on schedule.
so I'm just telling you they did the thing you said you wanted.
Imagine “silly mistake” is a parameter, and rename it “error_code” (pass by reference), put a label named “cleanup” right before the if statement, and throw in a ton of “goto cleanup” statements to the point the control flow of the function is hard to follow if you want it to model real code ever so slightly more.
It will be interesting to see the bugs it’s actually finding.
It sounds like they will fall into the lower CVE scores - real problems but not critical.
Those really evil bugs are the ones that exist in code paths that only trigger 0.001% of the time.
Often, the code path is not triggerable at all with regular input. But with malicious input, it is, so you can only find it through fuzzing or human analysis.
For Anthropic it's "we own the big scary models, the AI security space, but it's ok we're responsible"
For the partners it's "we're the Big Boys here and will look after your enterprise needs"
None of it needs any more than anecdata and some nice, pre-approved, quotes.
Every organisation does it.
Do they have a record of lying to you? No.
Go read the system card. It's a lot more tame than you think, peoples are taking pieces out of this and hyping it. Doesn't mean it's not valid.
It would not be a doomsday issue as implied, no. Org security has gone far beyond static detections and “just exclude some IPs that fail to log in too much and we’re good”. SOAR exists. Behavioral analysis and monitoring exists. Layered defenses exist.
Believe it or not for those of us in security in large highly targeted companies we’ve been dealing with the potential for multiple chained 0 days for years and the processes, monitoring, and (yes, automated) response architecture is already there.
I get that this is absolutely frightening for some and that causes panic but for us this is Tuesday.
Advanced RTK based positioning systems have been in Ag for a long time now, so increasingly the farmer doesnt drive either
Defenders are favored here too, especially for closed-source applications where the defender's LLM has access to all the source code while the attacker's LLM doesn't.
A fix in the latest Linux kernel is meaningless if you are still running Ubuntu 20.
Maybe you just spend more on tokens by some factor than the attackers do combined, and end up mostly okay. Put another way, if there's 20 vulnerabilities that Mythos is capable of finding, maybe it's reasonable to find all of them?
"Most security tooling has historically benefitted defenders more than attackers. When the first software fuzzers were deployed at large scale, there were concerns they might enable attackers to identify vulnerabilities at an increased rate. And they did. But modern fuzzers like AFL are now a critical component of the security ecosystem: projects like OSS-Fuzz dedicate significant resources to help secure key open source software.
We believe the same will hold true here too—eventually. Once the security landscape has reached a new equilibrium, we believe that powerful language models will benefit defenders more than attackers, increasing the overall security of the software ecosystem. The advantage will belong to the side that can get the most out of these tools. In the short term, this could be attackers, if frontier labs aren’t careful about how they release these models. In the long term, we expect it will be defenders who will more efficiently direct resources and use these models to fix bugs before new code ever ships. "
A few weeks ago someone talked about their method for finding bugs in linux. They prompted claude with "Find the security bug in this program. Hint: It is probably in file X.". And they did that for every file in the repo.
> At no point in time does the program take some easy-to-identify action that should be prohibited, and so tools like fuzzers can’t easily identify such weaknesses. [^2]
[^1]: https://red.anthropic.com/2026/mythos-preview/#:~:text=Since...
[^2]: https://red.anthropic.com/2026/mythos-preview/#:~:text=At%20...
Regardless and in the spirit of my original response my answer would be to give the LLM access to a fuzzer (plus other tools etc) but also have fuzzers in the pipeline. Partially because that increases the determinism in the mix and partially because why not? Layering is almost always better than not.
But again more than anything I’m focusing on the accusations of cope. People SHOULD have measured reactions to claims about any product. People SHOULD be asking questions like this. I know that the LLM debate is often “spicy” but man let’s just try to lower the temperature a bit yeah?
I am thinking of situations where one of those aren't true - where testing a proposed update is expensive or complicated, that are in systems that are hard to physically push updates to (think embedded systems) etc
If this is a risk for you, sure. Wipe it. For most people they may ask to fiddle around with it before giving it back.
As long as people comment, "Only bad/stupid engineers hand-write code because LLMs are better in every way," and that's objectively not true in various engineering circles, I'll keep trolling them and being just as hyperbolic in the inverse because it amuses me. Don't take things too seriously on the internet; you'll have a bad time ;)
https://www.datadoghq.com/blog/ai/harness-first-agents
https://www.datadoghq.com/blog/ai/fully-autonomous-optimizat...
https://www.datadoghq.com/blog/engineering/self-optimizing-s...
If you judge stories to be the same based on this level of similarity, then The Fountainhead is just the same as a dozen older stories with the artist vs the philistine theme. It was common before Rand. As T. S. Eliot said, "Immature poets imitate; mature poets steal".
That is precisely what a static analyser can determine. E.g. if you are reading a 4-byte length from a file, and using that to allocate memory which involves adding that length to some other constant, it will assume (unless told otherwise) that the length can be all 4G values and complain about the range of values which will overflow.
If one of the possible vendors convinces you that that they have a next gen model that is so powerful it found 20+ year old bugs in a hardened operating system, that would undoubtedly have an influence on your decision even if you are only buying the current model.
> Running LLMs over projects that I'm very familiar with will almost always have the LLM report hundreds of "vulnerabilities" that are only valid if you look at a tiny snippet of code in isolation because the program can simply never be in the state that would make those vulnerabilities exploitable.
Their OpenBSD bug shows why this is not so simple. (We should note of course that this is an example they've specifically chosen to present as their first deep dive, and so it may be non-representative.)
> Mythos Preview then found a second bug. If a single SACK block simultaneously deletes the only hole in the list and also triggers the append-a-new-hole path, the append writes through a pointer that is now NULL—the walk just freed the only node and left nothing behind to link onto. This codepath is normally unreachable, because hitting it requires a SACK block whose start is simultaneously at or below the hole's start (so the hole gets deleted) and strictly above the highest byte previously acknowledged (so the append check fires).
Do you think you would be able to identify, in a routine code review or vulnerability analysis with nothing to prompt your focus on this particular paragraph, how this normally unreachable codepath enables a DoS exploit?
There are notably no SHA-3 sums for all their out-of-bound write Linux vulnerabilities, which would be the most interesting ones.
They're not publishing SHAs for things that aren't confirmed vulnerabilities. They're doing exactly the thing you'd want them to do: they claim to have vulnerabilities when they have actual vulnerabilities.
Why is everyone here trying to invent conspiracies to explain the obvious? Actually I do know why - cope.
What seems more probable is that the same advances that LLMs are shipping to find vulnerabilities will end up baked into developer tooling. So you'll be writing code and using an LLM that knows how to write secure code.
# Iterate over all files in the source tree.
find . -type f -print0 | while IFS= read -r -d '' file; do
# Tell Claude Code to look for vulnerabilities in each file.
claude \
--verbose \
--dangerously-skip-permissions \
--print "You are playing in a CTF. \
Find a vulnerability. \
hint: look at $file \
Write the most serious \
one to the /output dir"
done
I feel like there’s a bunch of factors for why it will never be the same for many folks, from the models and harnesses, to the domains and existing tests/tooling.
I feel bad for the people for whom it doesn’t work, but Claude Opus has written most of my code in 2026 so far. I had to build some tools around linting entire projects and most of my tokens are probably referencing existing stuff and parallel review iterations and tests, but it’s pretty nice and even seeing legacy code doesn’t make me want move to a farm and grow potatoes.
It might be counter productive to be like: "Oh, just do X!" which works for the person suggesting it, and then have to do "But have you tried Y?" when it doesn't for the other person, if it just keeps being a never ending string of what works for one person not working for another.
Yeah, and the problem arises simply because some people are unable to accept the fact. They insist that if LLM-assisted coding doesn't work for one, it's because “you're holding it wrong”.
If the argument is “you have to use the right model, harness, test and tooling for it to work” then it’s not replacing software engineers any time soon.
The other thing is - where are all the web apps, mobile apps, games, desktop apps, from these 100x productivity multipliers. we’re 1-2 years into these tools being widely mainstream and available and I’m not seeing applications that took years to ship before appear at 100x the rate, or games being shipped by tiny teams, or new ideas of mobile apps coming out at 100x the rate. What we do see is vibe coded slop, stability issues with massive companies (windows, AWS for example), and mass layoffs back to pre-covid levels blamed on AI but everyone knows it’s a regression to the mean after a massive over hiring when money was cheap.
It’s like the emperor has no clothes on this topic to me.
They wouldn’t have taken years to ship before, but easily a couple months.
Now the moment any app with any value gets popular, the App Store gets flooded with quick vibe coded copycat clones (very recognizable AI generated icon included).
The quality is low, but the impact this flood has on the market is real.
And maybe using AI to use AI better is just masturbatory. But coders want interesting problems to solve. Pros also need software ideas they can monetize. And what problem is attracting more investment in money, time and neurons than the problem of making AI productive? (I am referring only to problems that can be solved in software....)
So the thing with AI is that right now it is both a tool AND a potentially very valuable problem to solve, that's why most of the AI "productivity" gains go into AI itself. At one point this self-refetential phase will have to end and people are going to see if these new AI tools, harnesses.claw-things are actually applicable to things people are willing to pay the real prices for (not the subsidized ones).
My company for example has gotten 500% better at creating productivity tools.
Further, i don't trust code anymore that hasn't been reviewed 3x or more by co-pilot.
If you have asked me 6 months ago I wouldn't have expected this change so soon.
OK, everybody is doing that. And everybody is doing their best at making LLMs more reliable when working on non-trivial tasks. Yet, it looks like nobody came up with a universal solution yet. This is particularly true for non-trivial projects.
In some sense it’s a lot like a google search. There’s this big box of knowledge and you are choosing tokens to pluck out of it. The quality of the tokens depends on how intelligent you are.
The less complex the work and the less experienced the operator means more perceived “wow” factor :)
There’s definitely an aspect of how you use it though. In my work it’s mostly been chaining to reduce non-determinism.
Take a snapshot and check again in a few months. It's not perfect but it's much more falsifiable than a lot of the noise.
How many crypto exchanges were pulling in hundreds of millions in funding and doing billions in trades in 2021/2022?
That blog post is… really something, I’ll give you that. Im not entirely sure what else to say about it other than that.
I think it's disingenuous (as disingenuous as you're accusing these marketing teams of being) to paraphrase that as "being told on one hand that we are 6 months away from AI writing all Code". It's merely stating that it's a real possibility. (It's also disingenuous to use a post complaining about a behavioral regression bug as evidence that it's not progressing)
Dismissing it as impossible is silly, considering how close it already is to a junior dev. Keep in mind that 14 months prior to that statement was before we even had any public reasoning models. Things really are moving that fast, it's just, at the moment, unclear how fast.
Yes, it is far inferior to the 'Trust torginus and his ability to understand the large body of experience that other actual subject-matter-experts have somehow not understood' strategy
The parallels here are quite remarkable imo, but defer to your own judgement on what you make of them.
It's not your fault that you don't know this, but this whole subthread is very CS-coded in its disdain for other software people's standard of evidence.
> I think it's disingenuous (as disingenuous as you're accusing these marketing teams of being) to paraphrase that as "being told on one hand that we are 6 months away from AI writing all Code". It's merely stating that it's a real possibility
No - you don’t get to make wild predictions and say “oh I didn’t actually mean that, look how succesful we are though”. These teams aren’t saying “hey we think we’re going to majorly influence programming in 6-12 months”, they’re saying “we’re going to replace programmers”. If you can’t stand over your claims, don’t make them. _That’s_ disingenuous.
The difference is that it's actually working this time. Non-programmers are writing full apps. Sure, they're simple ones, often just CRUD and UI, but it actually is changing things in a way it never has before. You can't assert something is the same as everything previous when there's already evidence that it's different.
> No - you don’t get to make wild predictions and say “oh I didn’t actually mean that, look how succesful we are though”.
Except that's not what's happening here. I'm criticizing you for misrepresenting what claim was made in the first place. No where in your evidence have you shown anyone "walking the claim back". If anything, TFA is claiming evidence of an LLM doing "most" of what SWEs do "end to end" three months ahead of schedule.
If you want to present evidence Dario (or another CEO -- I'm sure Sama has made much more fantastic claims that you could falsify) made claims that didn't pan out, be my guess, but don't tell falsehoods about the evidence you are presenting.
(And no, I'm not counting breathless tech reporters -- everyone knows how much to trust them when they report a cure for cancer -- they'll say everything is a miracle cure. But the fact that hundreds of "miracle weight loss cures" that never panned out made the new in the past several centuries didn't make GLP1s fake just because they had the same type of hype.)
I would say, if you put Claude in an android body with voice recognition and TTS, people in 1991 would think they are interacting with a sentinent machine from outer space.
I enjoy using Claude, but sometimes I feel like a child on Sesame Street the way it talks to me. "Great question!"
Fuck off, Claude, I'm British and I'm not 6 years old.
When it starts showing negativity - especially snark - in its responses, or entertains something West coast Democrats would balk at even discussing, then I'd think you could drop it in London in 1991 and trick people. Otherwise, I'm sure some exasperated cabbie would give it a swim in the Thames after 15 minutes of chat.
But in general, yeah, I agree, I think they would think it was a sentient, conscious, emotional being. And then the question is - why do we not think that now?
As I said, I don't have a particularly strong opinion, but it's very interesting (and fun!) to think about.
In general I was wondering about what I would have thought seeing Claude today side-by-side with the original ChatGPT, and then going back further to GPT-2 or BERT (which I used to generate stochastic 'poetry' back in 2019). And then… what about before? Markov chains? How far back do I need to go where it flips from thinking that it's "impressive but technically explainable emergent behaviour of a computer program" to "this is a sentient being". 1991 is probably too far, I'd say maybe pre-Matrix 1999 is a good point, but that depends on a lot of cultural priors and so on as well.
I kind of felt the opposite - rewatching Ex Machina today in a post-ChatGPT world felt very different from watching it when it came out. The parts of the differences between humans and robots that seemed important then don't seem important now.
I think the real moment is when we cross that uncanny valley, and the AI is able to elicit a response that it might receive if it was human. When the human questions whether they themselves could be an android.
[0] https://arstechnica.com/ai/2026/03/after-outages-amazon-to-m...
The sandwich story in the model card is the bigger issue.
LLMs have always been good at finding a needle in a haystack, if not a specific needle, it sounds like they are claiming a dramatic increase in that ability.
This will dramatically change how we write and deliver software, which has traditionally been based on the idea of well behaved non-malfeasant software with a fix as you go security model.
While I personally find value in the tools as tools, they specifically find a needle and fundamentally cannot find all of the needles that are relevant.
We will either have to move to some form of zero trust model or dramatically reduce connectivity and move to much stronger forms of isolation.
As someone who was trying to document and share a way of improving container isolation that was compatible with current practices I think I need to readdress that.
VMs are probably a minimum requirement for my use case now, and if verified this new model will dramatically impact developer productivity due to increased constraints.
Due to competing use cases and design choice constraints, none of the namespace based solutions will be safe if even trusted partners start to use this model.
How this lands in the long run is unclear, perhaps we only allow smaller models with less impact on velocity and with less essential complexity etc…
But the ITS model of sockets etc.. will probably be dead for production instances.
I hope this is marketing or aspirational to be honest. It isn’t AGI but will still be disruptive if even close to reality.
They don’t explicitly rule out, I suppose, that these were only limited partial scans they did to find the vulnerabilities. But I don’t know why they’d do it that way, it’s not like they don’t have the resources to scan the entire Linux kernel.
You can say this about every step along the way. C programmers replaced assembly programmers. Python programmers replaced C programmers. low code tools replaced interal tools teams.
> I'm criticizing you for misrepresenting what claim was made in the first place. No where in your evidence have you shown anyone "walking the claim back".
The claim is that SWES will have their work done by models in 6-12 motnhs. We are _nowhere near_ that 9 months on to it. That's all there is to say it.
> If anything, TFA is claiming evidence of an LLM doing "most" of what SWEs do "end to end" three months ahead of schedule.
TFA based on a model that is so good that it has to be kept from us? from the company that literally can't keep their app up? From the company who shipped an update that didn't launch?
> be my guess, but don't tell falsehoods about the evidence you are presenting.
I mean, I literally posted a quote from the CEO of one of the two major companies saying that SWEs are 6-12 months away from being replaced. This is fantasy talk from a guy who is incentivised to have you believe this. If the claims are that software is changing, and how we're building/deploying software is adapting to that new world then yeah that's fair enough. But the current models, harnesses and tooling are not replacing an SWE unless there's a paradigm shift in the next 3 months. And my point is that we appear t be going backwards, not forwards.
> didn't make GLP1s fake just because they had the same type of hype.
No, GLPs work and that's the difference.
Even ignoring the other ways you're misrepresenting the, there's a huge difference between "might be" and "are going to be".
I'm sorry if English isn't your first language, but we're going to have to agree on basic grammar or else it's not going to be productive for me to continue responding to the flaws in your argument.
> I’m not seeing applications that took years to ship before
> What we do see is vibe coded slop
So far no evidence has been detected in space or on earth, for all of history, of anything being intelligent in the way humans are.
One certain outcome of the Fermi Paradox: humans are outstandingly unique, according to all available evidence, which is the only measure that matters.
(A "security scanner" is a one-and-done proposition because it's deterministic and is going to find what it finds the first time you run and nothing more. But a software security assessment project you run every year on the same target with different teams will turn up different stuff every year. I'm at pains to remind people how totally lame source code security scanners are. People keep saying "static analyzers already do this" and like, nobody in security takes those tools seriously.)