Test case of a feedback loop, where the toaster "learns" how the user likes their toast done.
Every time the toast is finished the user tells the toaster whether the toast was under-done, over-done or just right.
This data, along with the toast time and bread width is saved into the database along along with the users name.
A ML algorithm uses the database to determine the optimal toasting time for the given parameters in order to achieve "just-right" toast.
This means that we have a toaster that learns from user preferences.