News for hackers. Stuff that matters.
Re:Conversational AI will be used to Manipulate Us in real-time by roywiggins on Thursday February 23, 2023 @10:31AM (#34911840)
Doesn't that basically devolve into my AI agent and your AI agent conferring and trying to mutually convince the other that they're a human before we actually communicate? Not sure that's a great future to live in.
Re:Conversational AI will be used to Manipulate Us in real-time by throwaway4aday on Friday February 24, 2023 @10:47AM (#34925659)
I don't see an alternative, at least if processing power and developments in ML continue to improve. Eventually you'll get AI clickfarms where people are just running hundreds or thousands of AIs out of a warehouse trying to scam people day and night. Arming everyone with their own AI bodyguard would prevent a lot of that from succeeding. The other options aren't too appealing e.g. only trusting face to face communication which basically means throwing away the internet and telephones and even hand written letters. Trying to regulate your way out will either lead to a war on AI where you're knocking down doors and arresting people who are trying to do legitimate independent ML research or you are just bombarded by offshore agents in countries with no regulation.