The issue is that people will use votes for if they like the thing or not instead of if it’s in good faith, even if you tell them not to, both on purpose to harm opposing views, and unintentionally because they’re more likely to notice a bad faith tactic coming from someone disagreeing than from someone agreeing with them.
that huge fruit slice is a bit interesting, because it seems like the AI cant decide if its generating a slice of watermelon or red pomelo