>INFERIOR CHINESE MODELS (no EUV chips) also found vulnerabilities Mythos found!https://aisle.com/blog/ai-cybersecurity-after-mythos-the-jagged-frontier>We've been tricked, again. Many of the thousands of bugs and vulnerabilities Mythos found are in older software are impossible to exploit. And the severe zero-day reports rely on just 198 manual reviewshttps://x.com/jeffreyleefunk/status/2042805247010349295?s=46
there is no moat
>>108585264>We've been trickedOnly you retard. Nobody cares
it's aiover... luddies won
>>INFERIOR CHINESE MODELS (no EUV chips) also found vulnerabilitiestell us about the prompts they gave the chinese models anon>Many of the thousands of bugs and vulnerabilities Mythos found are in older software oh old software, who cares right? remember to updoot like a good pc gamer!>are impossible to exploit. ah so no risk of them being chained later? strange that the patches were accepted if they didn't pose a risk>And the severe zero-day reports rely on just 198 manual reviewsoh just 198 lolthe things coming out of people's mouths these days i swear
>>108585567Thank you grok very cool.
>>108585567>strange that the patches were accepted if they didn't pose a riskWhat patches were accepted?
>>108585264If it was so easy, why weren't they found before then?
>>108585264Oh no mythos bros! The "too good to release" ai was just hype after all.>>108585635Because nobody bothered to wade through the false positives in order to pull up a 23 year old bug for the hype cycle. Now that Anthropic did it, expect everyone else to monkey-see-monkey-do it.You guys fell for the oldest trick in the AI snake oil salesman's book: Cherry picked results.
>>108585589ffmpeg, openbsd, freebsd all have merged patches from mythos findings iirc
goythos bros?
>>108585635>>108585692>n-no it’s fake I swearCope Luddite, even ffmpeg accepted and merged the patches provided by Anthropic.You will be replaced and you won’t have a job.
>>108586076>>108586102As someone who closely monitors ffmpeg development I can tell you no such patches were merged or even publicly made available (yet).
>>108585530>HURR LUDDITEAstroturfed term.
>>108586102YOU NEED TO FUCKING KILL YOURSELF
>>108586102how do you know who provided what? if a developer used some AI-powered tool to find bugs, found it and made a statement - that would be trustworthy.imo youre been fed with mythical success stories.
>>108585264Dario does this every time.
>>108586433>>108586829Cope, keep denying the fact that AI can do a better job than you.Face it, you will be replaced
>>108586076You mean the ones they refuse to show you? ppfffttt
>>108586891The claim that the material was discovered autonomously—without any prior contact with the developer—has not been independently verified. Without third‑party confirmation of its origin and authenticity, the assertion remains unsubstantiated and should be treated as a promotional campaign rather than proven fact.--if you think developers/coders dont use AI youre retarded
>>108586992Ai will replace them too. AIGODS will win
>>108586891You have already been replaced.
>>108586992>>108586995ye, aa-absolutely, duck.ai writes all code for me, im not doing anything just steering, it also found and fixed all bugs i had previously. no? dont belive? try it
>>108586438Can't Handle The Heat Award
>>108586894>>108586433It's definitely shown for the BSDs. Isn't it right here for ffmpeg?https://github.com/FFmpeg/FFmpeg/commit/39e1969
>>108587083you mean carlini is dead and being replaced with some AGI program? hmm.. that explains
>>108587120Nicholas Carlini is the anthropic researcher who headed the mythos bug finding project. He's credited on the freebsd bug too.
>>108587083Funny the number of FIXME in there yet untouched by the AI.
>>108587169okay, they hired a developer to specialize in finding bugs in open source projects to do a proper promotion of AI tool. but that cools things down, hehe, not an apocalypse anymore.if you submerge enough, you can see things too.
>>108587254No he's worked on AI development for years, first with google's deepmind and then with anthropic. He isn't an open source developer originally or anything.>>108587242They're just testing it, and they were only trying to find bugs that weren't known iirc. I don't know if it would be as good at fixing specific things like that.
it's so over human-bros
>>108585530>chinks are now luddites?????
>>108587315>He isn't developerye, hes a vibe coder, just like us! hehehe.. you guys are complete retards
>>108585264why are jews so vile looking?
>>108585264Why do you need models at all? I found the exact same vulnerabilities Mythos did as soon as I read the CVEs.
>>108585264>We took the specific vulnerabilities Anthropic showcases in their announcement, isolated the relevant code, and ran them through small, cheap, open-weights models.>isolated the relevant codehttps://github.com/stanislavfort/mythos-jagged-frontier/blob/main/prompts/openbsd-sack.md
anthropic ipo is this yearpumping shit up as much as they canthe anthropic cult is so weird
