One Step of Gradient Descent

You are fitting the linear model $\hat{y} = wx + b$ to this two-point training set:

The loss is mean squared error over the full batch:

$L(w, b) = \frac{1}{n}\sum_{i=1}^{n}\bigl(\hat{y}_i - y_i\bigr)^2, \qquad n = 2$

You initialise $w_0 = 0$, $b_0 = 0$ and take one full-batch gradient descent step with learning rate $\eta = 0.1$, updating $w$ and $b$ simultaneously from the gradients computed at $(w_0, b_0)$.

What are $w_1$, $b_1$, and the loss $L(w_1, b_1)$ after that single step?