If two variables $X_1$ and $X_2$ are perfectly uncorrelated, what happens to the estimate of $\beta_1$ when adding $X_2$ to a simple regression $Y = \beta_0 + \beta_1 X_1 + \epsilon$?