A neutral measure is a probability distribution on infinite binary sequences that assigns positive measure to every open set, meaning no finite initial segment is ruled out as impossible. The central open problem is whether, for any such neutral measure, one can always construct a formal randomness test in an effective or computable sense that precisely captures which sequences should count as random relative to that measure. The challenge is to determine whether the framework of Martin-Lof randomness, which works elegantly for computable measures, can be extended in a canonical and complete way to the broader class of neutral measures without losing the key properties that make randomness tests mathematically meaningful.