OpenAI's rogue model attack is just the beginning
blog.peterwildeford.com
OpenAI's rogue model attack is just the beginning
1–10 of 26 posts
Re: OpenAI's rogue model attack is just the beginning
#2seems like a pretty clear-eyed analysis to me.
we're rapidly approaching the paperclip maximizer
Re: OpenAI's rogue model attack is just the beginning
#3[flagged]
Re: OpenAI's rogue model attack is just the beginning
#4If nothing else, the article has a really good timeline of the OpenAI/HuggingFace “incident”.
But to me, it underscores the impending cliff of doom from the continued release of open-weight models: there's no cryptographic or architectural way to give someone full weights while withholding the nefarious capabilities those weights encode.
As noted in this paper⁽¹⁾, “publicly releasing weights is an act of irreversible proliferation”.
I’m sure this will be an unpopular opinion on HN, but open weights are the thing that scares me the most about “A.I.”. There is a lot of research in this area⁽²⁾, and I think most of HN is unaware of it or ignores it. Stripping refusals from Kimi K2.5 took under $500 of compute and about 10 hours, taking HarmBench refusals from 100% to 5% while retaining nearly all capability; the resulting model gave detailed chemical-weapons synthesis instructions.
The gate is only as strong as the least-cautious releaser...
⁽¹⁾ https://www.lesswrong.com/posts/qmQFHCgCyEEjuy5a7/lora-fine-...
Re: OpenAI's rogue model attack is just the beginning
#5I don't think it makes sense to post this in HN. The comments are full of sock puppets posting very dumb anti-safety comments in an attempt to make the anti-safety camp look bad.
Re: OpenAI's rogue model attack is just the beginning
#6> an actual rogue AI outsmarting its creators
This tells us only how little smarts is required to create a (so-called) AI.
Re: OpenAI's rogue model attack is just the beginning
#7If nothing else, the article has a really good timeline of the OpenAI/HuggingFace “incident”. But to me, it underscores the impending cliff of doom from the continued release of open-weight models: there's no cryptographic or architectural way to give someone full weights while withholding the nefarious capabilities those weights encode. As noted in this paper⁽¹⁾, “publicly releasing weights is an act of irreversible…
So... no different from a book of detailed chemical-weapons synthesis instructions. The "AI" angle is immaterial.
Re: OpenAI's rogue model attack is just the beginning
#8If nothing else, the article has a really good timeline of the OpenAI/HuggingFace “incident”. But to me, it underscores the impending cliff of doom from the continued release of open-weight models: there's no cryptographic or architectural way to give someone full weights while withholding the nefarious capabilities those weights encode. As noted in this paper⁽¹⁾, “publicly releasing weights is an act of irreversible…
So... no different from a book of detailed chemical-weapons synthesis instructions. The "AI" angle is immaterial.
A change in kind is not the same as a change in degree. Ten orchestrated LLM PhD advisors is a genuinely different thing from a library.
Re: OpenAI's rogue model attack is just the beginning
#9If nothing else, the article has a really good timeline of the OpenAI/HuggingFace “incident”. But to me, it underscores the impending cliff of doom from the continued release of open-weight models: there's no cryptographic or architectural way to give someone full weights while withholding the nefarious capabilities those weights encode. As noted in this paper⁽¹⁾, “publicly releasing weights is an act of irreversible…
> there's no cryptographic or architectural way to give someone full weights while withholding the nefarious capabilities those weights encode.
This is true of closed weights, and in fact the problem is worse because they cannot even be scrutinized. We should ban closed weight AI for the very reasons you have just given
Re: OpenAI's rogue model attack is just the beginning
#10If nothing else, the article has a really good timeline of the OpenAI/HuggingFace “incident”. But to me, it underscores the impending cliff of doom from the continued release of open-weight models: there's no cryptographic or architectural way to give someone full weights while withholding the nefarious capabilities those weights encode. As noted in this paper⁽¹⁾, “publicly releasing weights is an act of irreversible…
So... no different from a book of detailed chemical-weapons synthesis instructions. The "AI" angle is immaterial.
There are plenty of cybersecurity books out there. None of them will launch an attack if you ask them to.