A study is designed to test if a new drug lowers blood pressure. The null hypothesis is $H_0: \mu_{new} \geq \mu_{old}$. What is the alternative hypothesis?