Or consider the algorithmic complexity classes underpinning all of this, not just P vs NP, but also whether quantum computers are actually a distinct class, since that is not proven yet either afaik. The reason we're moving to quantum resistant algos in the first place.