It is well known that Grover's algorithm can solve AES in $O(\sqrt(n))$ time, which is why symmetric key length needs to be double to maintain their security level in the face of a quantum adversary. A recent eprint paper (https://eprint.iacr.org/2022/948.pdf) claims there exists a polylog time attack on AES (or searching unsorted sets in general) using a quantum computer. Can someone familiar with quantum computing comment on the validity of this attack?