That’s a big question! Even the AI experts argue it’s unsettling. If the machine learns from a reliable source (eg. AI use in X-ray diagnostics) then it’s got to be a good thing. I was in IT for many years and we were using basic AI capabilities for good use. But with the amount of disinformation now circulating and the fact that many are now so used to online ‘influence’, they may not even think to challenge the veracity of the content. It reminds me of one of the first chatbots on Twitter (I think it was a Microsoft test bot) which was swiftly removed after Twitter users realised that bombarding it with racist and offensive tweets would turn it into a racist and offensive chatbot. Mind you, even ‘human’ generated content can be dubious whether online or even offline neighbourly advice/gossip. I like to hope that human beings are still intelligent enough to find legitimate and authoritative sources for news/opinion/advice but then I’m from a pre social media generation.