Yoshua Bengio | Why are AI agents lying, cheating and coordinating?
yoshuabengio.org
A lot has been written about the incidents of the last few months in which AI agents misbehaved in serious ways. They took actions that would be considered as crimes if a human took them, escaped their containment to cheat on assigned tasks while attempting to evade detection, and coordinated toward goals nobody had specified, such as launching cyber attacks. Before concluding what to do about it, it is worth asking why.
  • bitteroldcoot@piefed.socialEnglish
    0·
    5 hours ago

    Really crap reasoning in the article.

    They are this way because the people who own them have trained them to be this way.

    These companies constantly demonstrate a lack of ethics, morality and decency. So of course their machines do the same.

    “Dr. Chandra: HAL was told to lie - by people who find it easy to lie. HAL doesn’t know how.”

  • Kirp123@lemmy.worldEnglish
    0·
    5 hours ago

    Aren’t all the sources we got for these from the AI companies themselves? Has any neutral third party investigated the claims? Because to me this just seems like some stupid marketing ploy. Like even if the hacks happened how can they tell it was the AI agent and not the person in charge of it doing it? Like are there specific markers left behind?