A researcher fits a simple linear regression model with a single predictor $x$ and obtains $R^2 = 0.50$. She then adds a second predictor $w$ which is completely uncorrelated with both $x$ and $y$ in the sample. What is the expected behavior of the multiple $R^2$ and the Adjusted $R^2$ in the new model?