Your question is Python Memory and GIL. 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).
Explain how memory management and the Global Interpreter Lock work in Python, and how they impact multi-threaded machine learning pipelines. When do threads help, when do they not, and what alternatives would you consider?