It is not correct to say that noise does not convey information. After all, every null result is ultimately limited by noise and so null results are necessarily what we call noise.
Robert H McEachern replied on Aug. 14, 2017 @ 17:49 GMT But Shannon offered a long-ignored insight, into why naive observers seem to perceive two; they remain inappropriately focused on the measurements, rather than the measurements' information content. The latter is the only thing that conveys repeatable, actionable "detection" of anything to ever interact with (and thus capable of supporting the existence of identical interactions amongst identical particles).
Thus classical noise is what carries the Shannon "0" bit while quantum noise is what carries the quantum qubit, which is a superposition of both "0" and "1".
The confusion here is due to the limitations of classical knowledge, which is causal and deterministic and sources are always singular, and quantum knowledge, where sources remain uncertain and in superposition.