Currently when I review my generated content for the blog posts, I try to have satisfy 2 objectives at the same time. The first one is usefulness, meaning that it has to be comprehensive, well structured, easy to extract for a crawler or an AI agent, etc. The second is to sound human. If I don't try to optimize for human sounding content, I'm thinking people will not want to read further because people dislike content that sounds like it was written by a clanker. Even when useful. But does that second objective still matter now? Because I assume people read blog articles themselves less and less, as they go to chat GPT and wait for the AI agent to serve them content instead. So in the end only the robot reads it, and probably doesn't care about whether it sounds human or not. It just cares about whether it is useful. Dropping the second objective would make the job easier, but maybe I would be shooting myself in the foot?