Because when you read a qubit, you collapse the superposition and end up with a boring old 0 or 1. Which is what you want when the algorithm is completed, but during the execution of the algorithm you want to do operations on the superpositions. So you cannot read the value and feed it back because in the process, you destroy it.
Comments
Because when you read a qubit, you collapse the superposition and end up with a boring old 0 or 1. Which is what you want when the algorithm is completed, but during the execution of the algorithm you want to do operations on the superpositions. So you cannot read the value and feed it back because in the process, you destroy it.