Instead of "confidence interval," let's say "uncertainty interval" (stat.columbia.edu)