51–60 of 131 posts
In the end I got to help write up the issue but to my knowledge they never patched it as it would have caused major issues with maintenance by closing off access needed for some legacy software patches.
What happens when open source models achieve Mythos level capabilities in six months' time?
What matters isn't the power of the tool, but whether defenders have had time to secure against. Today's cyberweapon is tomorrow's laughably obsolete.
Stuxnet used to be a national security threat, now I'm not sure it would be useful for anything.
From outside? Or did you have a shit ton of unpatched systems that only internal users could access?
Not surprised, our security systems are 95% security through obscurity these days. Mythos didn't find new ways to break security, it just went down the list of common security exploits and exposed them for being common even among government agencies.
Not a surprise. I got in a LOT of trouble for identifying and outlining a trivial privilege escalation attack that worked on both NIPR and SIPR. In the end I got to help write up the issue but to my knowledge they never patched it as it would have caused major issues with maintenance by closing off access needed for some legacy software patches.
If 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…
Let's just take GPT 5.5 and Opus 4.8 as an example. Both are worse than Mythos 5, but they're capable of quite a bit when the guardrails are lifted and they're paired with a skilled human operator. They more than "good enough" to reach the same result with the addition of some human effort.
Earlier quoted context omitted.
I think it's probably more like a year or a year and a half. I don't want to say two years, but it's what I'm actually thinking.
GLM 5.2 is already between 4.6 Opus 4.7 Opus level based on Artificial Analysis aggregation. 4.6 Opus is about 4 months old at this point, so seems like open source is maybe 4-6 months behind. It could still take a year but seems closer to 6 months.
There's simply no replacement for training on more, better tokens, with more parameters. Mythos/Fable was estimated to be closer to 10T parameters than the 800B like GLM 5.2 is.