RSA 1024 is post quantum if you want to ignore progress in cryptography and use current algorithms. (We have no quantum computers that can crack it right now)

It's about preparing for quantum computers by using algorithms that are secure against conventional and future quantum computers. If you assume that a quantum computer will exist that can crack RSA 2048/4096, then all data that gets send right now can be decrypted at that time. If we get working quantum computers in 20 years then in 20 years all banking data, chat messages, emails,... send with RSA today can be compromised.
If we switch to algorithms that don't get easier to crack with quantum computers then even when they get strong enough nothing will change and only data send with older algorithms can be decrypted.

See also the rest of the Wikipedia article, here a continuation of my previous snippet:

Most widely used public-key algorithms rely on the difficulty of one of three mathematical problems: the integer factorization problem, the discrete logarithm problem or the elliptic-curve discrete logarithm problem. All of these problems could be easily solved on a sufficiently powerful quantum computer running Shor's algorithm or possibly alternatives.

As of 2024, quantum computers lack the processing power to break widely used cryptographic algorithms; however, because of the length of time required for migration to quantum-safe cryptography, cryptographers are already designing new algorithms to prepare for Y2Q or Q-Day, the day when current algorithms will be vulnerable to quantum computing attacks.

Algorithms. Plural. Shor's and Grover's. Nothing public that can break anything in use but progress marches on and governments are always expected to be 5-10 years ahead