In movie "Interstellar": Cooper stated that he was going to dock. Then Case calculated a possibilty for which the ship would successfully dock with the endurance which was Not possible. I assume, that Case included Cooper into its algorithm, and yet predicted no success rate. Maybe I'm missing the connection. "Case knows how to dock with the endurance, but not when it's spinning, that's where Cooper comes in". I think for an advance AI it should be able to match its vector b,(Cooper's ship) with vector a (Endurance). Pretty much all were doing it matching Cooper's ship velocity with the endurance, and matching the endurance's rotation speed as it declines, with Cooper's ship. And if an AI CAN do this, its answer should not be not possible. It should have a higher possibility rate.
In our life, some times we can't say it is not possible, but it is necessary
