A remark on the positivity of K2