I saw an IBM commercial that depicted an AI personal assistant informing a user that their credit card had been used for a fraudulent charge. It asked the user if a particular charge was legit and they said no. Then the AI informed the user that the transaction had been canceled and a new card had been issued.
I have a couple problems with this. First, what if the AI was hallucinating parts of this interaction? Secondly, at some point the user’s AI will be interfacing with the bank’s AI, then we will effectively be subservient to a bunch of algorithms automatically running the world and we will basically be children with our AI parents taking care of us.