This is the kind of thing that makes me worry about AI the most. No paper trail. No one to blow the whistle. You don’t even need the rooms full of sociopaths and complicity like we do today. Just raw data and reward metrics. Does the AI receive negative reinforcement when it contributes to a young girl killing herself or developing an eating disorder? Obviously not. It doesn’t even “know” what it’s doing besides making the number go up.
We see how that works out with United Healthcare denying claims by AI.
This is the kind of thing that makes me worry about AI the most. No paper trail. No one to blow the whistle. You don’t even need the rooms full of sociopaths and complicity like we do today. Just raw data and reward metrics. Does the AI receive negative reinforcement when it contributes to a young girl killing herself or developing an eating disorder? Obviously not. It doesn’t even “know” what it’s doing besides making the number go up.
We see how that works out with United Healthcare denying claims by AI.