Confidence intervals you're allowed to peek at anytime
Asked at Netflix
Product teams hate waiting for a fixed sample size and want to watch experiments live, stopping as soon as there's a clear answer. But peeking at ordinary p-values and confidence intervals inflates the error rate.
Explain what an always-valid (anytime-valid) confidence sequence is, how it differs from a fixed-sample interval, and the tradeoff it makes.
Your answer
This one is open-ended. Work it through, then check your reasoning against the full solution.