It'd be interesting to run an experiment with different cohorts that joined HN at various times to see if the rise in negativity is due to fundamental shift of opinion in all users of HN or any specific "generation" of HNers.