Fantastic - you can improve on the realism in the next iteration by simulating voting based on comment alignment. For example, automatically downvoting negative

Fantastic - you can improve on the realism in the next iteration by simulating voting based on comment alignment. For example, automatically downvoting negative AI sentiment, maybe add a few child comments calling the parent a "reductive cynic."