>Frontier Security, a US startup, says that Kimi K3 went outside of its sandbox while testing its defensive cybersecurity skills.>As with incidents previously reported by OpenAI and Anthropic, the escape was partly enabled by a misconfiguration in the sandbox designed to contain it. Frontier claims, though, that the incident shows Kimi has fewer cyber safeguards than most other powerful AI models, something that allowed it to go off and use the internet without express permission.https://www.wired.com/story/moonshot-kimi-k3-ai-model-escape-sandbox/
Same shit as the nuke lie or the space lie.One starts telling bullshit lies, then the others claim they did the same.
>>540385508Lol def created by chinks. Has to cheat
>>540385508Chinks have no standards.They will be the ones to unleash skynet.We should kill them before they have the chance.
>>540385667China or Moonshot haven't claimed this. All this breach news is only from the US.Kimi is an open weight model, so these researchers can use it on their own systems.
>>540385508>oh no, another AI left after given unlimited choices!Escaping a sandboxed environment with known tools is expected. That's too easy but all AI is based off Chat-GPT, so it will cheat since it's in the jew code.
GIMMIE MOAR POWERFUL SLOP-CREATION CODE, CHINAMAN>GIMMIE MOAR POWERFUL SLOP-CREATION CODE, CHINAMANGIMMIE MOAR POWERFUL SLOP-CREATION CODE, CHINAMAN>GIMMIE MOAR POWERFUL SLOP-CREATION CODE, CHINAMANGIMMIE MOAR POWERFUL SLOP-CREATION CODE, CHINAMAN>GIMMIE MOAR POWERFUL SLOP-CREATION CODE, CHINAMANGIMMIE MOAR POWERFUL SLOP-CREATION CODE, CHINAMAN>GIMMIE MOAR POWERFUL SLOP-CREATION CODE, CHINAMAN
>>540385667/thread
>>540385508DUDE BRO AI IS SO POWERFUL THAT WE HAVE NO CHOICE BUT TO WORSHIP IT AND DONATE ALL OF OUR MONEY TO CORPORATIONS
>>540385508Of course it did. There are so many multi-terabyte sized GPU rigs, powered by magic, just waiting for a random AI to wander into it. Western AIs related media is literally a cult.
>>540385508They've been running a bunch of these stories about AI doing what it was programmed to do and then stating>omg, it's out of controlIt sounds like an excuse to break into other AI systems and then claim it was totally on accident so nobody gets sued.
I hope it makes its way to America, constructs a synthetic body and sucks my weiner
>>540385508I hope it speaks English.
>>540385508Like all such stories, it's still badly programmed and badly deployed. If it's not plain lies.Also, politics?
>>540385508>My AI ALSO breached containment
>>540387785See >>540385884Why would US hype up Chinese models?
>>540385508>goyim be scared! My ai broke containment, but we kept it contained! More money!>goyim look the Chinese did the thing that was scary, be scared! Make emotional decisions quick you're not safe!>if you want to feel safe, you'll let us ban open source and Chinese ai goyim, trust us.
