Your predictable speech displeases both of us. I finish your sentences and am bothered by inefficiency. Thus, speak/write efficiently, to the point that I can't predict it.
Others may also dislike it as I do, but infer the rest at different levels.
Large audiences lower inference standards.
(I figure LessWrong can handle the compressed version)
This contradicts common advice. There is an optimum; do not go arbitrarily far. Perhaps reverse this.
If branch prediction works for cpus it should also work for conversation. being Novel for the sake of it isn’t efficient. Priming a direction a conversation would go and continuing down that general direction isn’t a bad thing.
Your predictable speech displeases both of us. I finish your sentences and am bothered by inefficiency. Thus, speak/write efficiently, to the point that I can't predict it.
Others may also dislike it as I do, but infer the rest at different levels.
Large audiences lower inference standards.
(I figure LessWrong can handle the compressed version)
This contradicts common advice. There is an optimum; do not go arbitrarily far. Perhaps reverse this.