If a Bayesian posterior distribution is $\pi(\theta|D) \sim N(\mu_p, \tau_p^2)$, what is the relationship between the posterior precision and the prior/likelihood precisions?