https://archive.is/aA1dB
The link does not seem to be working
Works for me though, even when using a proxy that is usually blocked everywhere.
NSA director: 'Mythos "broke into almost all of our classified systems in hours"
11–20 of 131 posts
Re: NSA director: 'Mythos "broke into almost all of our classified systems in hours"
#12If mythos can break into almost all of their classified systems in hours then other models including opus, gpt, gemini and large open weight models can do so as well, maybe you'll have to double hours or it may become days, but they also will, there is no "maybe" in here.
State sponsored, non-public penetration fine tunes (of possibly public ones) likely can do it even faster.
Unsupervised penetration RL loop is ideal setup similar to optimization one – it's relatively easy to gain function on it.
Re: NSA director: 'Mythos "broke into almost all of our classified systems in hours"
#13Not being funny but does most of HN subscribe to the economist? I dont think ive ever paid for an online newspaper ( and Im not trying to be edgy )
more likely, most of HN who care about reading this article use something like archive.is
Re: NSA director: 'Mythos "broke into almost all of our classified systems in hours"
#14If mythos can break into almost all of their classified systems in hours then other models including opus, gpt, gemini and large open weight models can do so as well, maybe you'll have to double hours or it may become days, but they also will, there is no "maybe" in here. State sponsored, non-public penetration fine tunes (of possibly public ones) likely can do it even faster. Unsupervised penetration RL loop is idea…
I don't think that is necessarily true.
- With a weaker model, the time to break into the system might grow so larger that it becomes infeasible, similar to how password hashes can be bruteforced, but if the password is long enough, that is not going to happen in our lifetime.
- There might be problems which are inherently unsolvable with a lower level of intelligence. For example, your dog won't derive calculus from scratch, even if it lived forever.
- LLMs might be biased in such a way that they never explore the entire solution space, no matter how many attempts are made. Some models are notorious for getting stuck in a loop, trying small variations of the same approach every time, even though it is doomed to fail. This can be counteracted somewhat with higher sampling temperature, but that hurts reasoning capabilities.
Re: NSA director: 'Mythos "broke into almost all of our classified systems in hours"
#15If mythos can break into almost all of their classified systems in hours then other models including opus, gpt, gemini and large open weight models can do so as well, maybe you'll have to double hours or it may become days, but they also will, there is no "maybe" in here. State sponsored, non-public penetration fine tunes (of possibly public ones) likely can do it even faster. Unsupervised penetration RL loop is idea…
I don't think that is necessarily true. - With a weaker model, the time to break into the system might grow so larger that it becomes infeasible, similar to how password hashes can be bruteforced, but if the password is long enough, that is not going to happen in our lifetime. - There might be problems which are inherently unsolvable with a lower level of intelligence. For example, your dog won't derive calculus from…
The concept of infinity claims that the dog eventually becomes Shakespeare. The same way we handled encryption, even before Alan Turing codes were broken and evolved. Last, it is a huge advantage to have the machine/mind and to evolve from there.
P.S. Even if you go back to lemon juice on paper there may be a thief around that knows the trick.
Re: NSA director: 'Mythos "broke into almost all of our classified systems in hours"
#16If mythos can break into almost all of their classified systems in hours then other models including opus, gpt, gemini and large open weight models can do so as well, maybe you'll have to double hours or it may become days, but they also will, there is no "maybe" in here. State sponsored, non-public penetration fine tunes (of possibly public ones) likely can do it even faster. Unsupervised penetration RL loop is idea…
I don't think that is necessarily true. - With a weaker model, the time to break into the system might grow so larger that it becomes infeasible, similar to how password hashes can be bruteforced, but if the password is long enough, that is not going to happen in our lifetime. - There might be problems which are inherently unsolvable with a lower level of intelligence. For example, your dog won't derive calculus from…
Mythos and other models are not brute-forcing passwords (and with this analogy passwords, ie. systems are the same).
We're not talking about dogs, but LLM systems.
Mythos is not exploring entire solution space either.
Usually looping is solved by repetition/frequency/presence/n-gram penalties/DRY/min-p sampling, not temperature but we're not talking about small models that have those classes of issues here.
Re: NSA director: 'Mythos "broke into almost all of our classified systems in hours"
#17This is really making me raise an eyebrow. I’m sure mythos is an improvement for sure. I don’t think the framing of it hacked the entire NSA is fully truthful. I’d like a more in depth understanding of what actually happened. Excited to be proved wrong tho!
Yeah, this article cites someone saying that someone else said something. Maybe it was said, maybe not. Maybe it was a exaggeration, maybe not.
Very insightful.
Re: NSA director: 'Mythos "broke into almost all of our classified systems in hours"
#18https://archive.ph/dXddV
“Donald Trump’s blocking of Anthropic is capricious and chaotic” - current title
I don’t understand the posted title quote and assume it’s missing a lot of context or was misinterpreted as it’s a secondary attribution. “Mythos broke into almost all of our classified systems in hours”.
When you put it on those networks already and gave it compute?
Re: NSA director: 'Mythos "broke into almost all of our classified systems in hours"
#19flagged because reasons. @dang
Re: NSA director: 'Mythos "broke into almost all of our classified systems in hours"
#20HN post title does not match link title > NSA director: 'Mythos "broke into almost all of our classified systems in hours" > Donald Trump’s blocking of Anthropic is capricious and chaotic
I’ve found that high community-upvoted posts don’t bury the lede by parroting the headline. I used to be a headline title scribe until the HN community showed me the light.