‹ BackHN Continuity

Thread

There are no "rogue" AI agents

396 points · 269 comments · zzzeek

  1. cesarsk · · focus · HN ↗
    I think this promotes the narrative that these companies possess agents so powerful that even humans struggle to control them, which ultimately makes their products look sexier. Additionally, I believe AI leaders may have called for regulations to slow down development simply to catch their breath for the sake of our 'survival': agents are going rogue, help us stop them.

    Some recent hacks done with the support of agents occurred simply because endpoints were unprotected. With traditional hacking, the blame falls on the person who didn't secure their system. But with AI, the immediate reaction is to fear we are on the verge of extinction!

Open on Hacker News to reply ↗

Unofficial Hacker News client; not affiliated with Y Combinator.