Your question is Bias-Variance Error Decomposition. Take a moment with it on the right.
Talk me through your thinking if you like. When you're confident, submit your answer and I'll grade it like a real screen (7/10 or better passes).
How would you decompose errors into variance and bias for an ML model?