From Artificial Intelligence: A Modern Approach, Third Edition, Chapter 26:
Note that the concept of ultraintelligent machines assumes that intelligence is an especially important attribute, and if you have enough of it, all problems can be solved. But we know there are limits on computability and computational complexity. If the problem of defining ultraintelligent machines (or even approximations to them) happens to fall in the class of, say, NEXPTIME-complete problems, and if there are no heuristic shortcuts, then even exponential progress in technology won't help—the speed of light puts a strict upper bound on how much computing can be done; problems beyond that limit will not be solved. We still don't know where those upper bounds are.
If the textbook's argument is correct, then there may be a strict upper bound to intelligence, meaning that the potential (or damage) of super-intelligence is limited. However, it is contingent on there actually being a theoretical maximum for intelligence.
Is there any literature that suggests that we know for sure whether such a maximum exist? Is the existence of that maximum dependent on our definition of intelligence (so adopting a vague and hand-wavey definition would imply no theoretical maximum, while adopting a strict and formalized definition would imply a theoretical maximum)?