Discuss the theoretical foundations of the AI control problem. Analyze the concepts of instrumental convergence and the orthogonality thesis. How might an unaligned superintelligence behave theoretically if its final goals are even slightly mis-specified?