Loophole: Adversarial Agents To Stress Test Your Morality — Brendan Rappazzo, Morgan Stanley
Adversarial agents can test and reveal contradictions in personal morals.
“I could sort of say yes or no.”
Brendan Rappazzo discussed a project at Morgan Stanley involving adversarial agents that challenge individuals' moral beliefs by identifying contradictions. This approach taps into legal system analogies, aiming to better understand and codify personal morals, which has important implications for ethical AI applications.