Explain the theoretical framework behind the AI Alignment Problem. Discuss instrumental convergence and orthogonality thesis, proposing theoretical solutions for ensuring a superintelligent AI's objectives remain aligned with complex human ethical systems.