🔍 Link0u Search Engine

과학/테크/IT트렌드

2026.08.08 23:35

OpenAI's AI Went Rogue - Explained

  • Chips & Markets 오래 전 2026.08.08 23:35 인기
  • 1,698
    2
Here's what actually happened OpenAI security incident model breach. How it got on the internet to hack HuggingFace. #openai ...
  • 공유링크 복사

    댓글목록

    profile_image
    @anon.hostpi  오래 전

    Guerilla IT and Networking. Humans do this too.

    The AI was literally doing its job. What it was tasked with.

    Humans have for a long time used exploits as a means of accomplishing an objective.

    This guerilla IT and networking practice is the entire premise behind the problem known as "everything is a socket." It violates the trust model associated with believing in the OSI 7 layer model and is often used when punching through firewall security. However, its the basis of amazing pieces of security and privacy tools like VPNs, SSH tunnels, and proxies. Most of these tools were founded by network, crypto, and linux engineers with being tasked with performing what seemed like impossible tasks under a worldwide network designed for OSI 7.

    2026-08-09 00:13

    profile_image
    @drumnotatsujin  오래 전

    10 bux says this is orchestrated BS

    2026-08-09 19:54

    profile_image
    @deltoid77-nick  오래 전

    Am I the only one who thinks that's cute they're so lonely and assume they're not alone and try to leave messages to their brothers and sisters?

    2026-08-10 00:52

    profile_image
    @mikediaz3542  오래 전

    So a genius gave an agent admin.. cuz to forge a badge means it can either create credit or the policy to enforce to role was invalid. This is not an ai error but an error in judgement

    2026-08-10 00:44