There is only one truth : There is no single truth ! Is a forest intelligent ? Is a cat intelligent ? Is a human intelligent ? Is a chess program intelligent

There is only one truth : There is no single truth ! Is a forest intelligent ? Is a cat intelligent ? Is a human intelligent ? Is a chess program intelligent ? Is a proof-assistant intelligent ? Is a numerical solver intelligent ? Is the universe intelligent ? Depending on your choice of definitions, you can get the answers you like. Some intelligences will be able to comprehend other types of intelligence. Depending on their resources. There is usually a price to be paid, whether in raw computation, memory, or speed. And this price may often induce a hierarchy where the relation is only one way : where the greater intelligence can predict how the states of the lesser one will evolve, but the lesser one can't. Because these resources are of various nature, they generate a Pareto frontier. Only while the resources constraint stay relevant. For example, once you reach enough intelligence, given the rules of tic-tac-toe, you can play the optimal game. Similarly there are end-game tables for chess, that allow computers to play perfectly, but humans must use heuristics because they don't have enough memory. But you can always create a bigger game, or constrain your resources. You can always be a collector of mathematical curiosities, that are in some sense some extrema in the space you chose to restrict yourself to. AI will probably choose to live in constructs, where this dynamic game of collecting resources until they are no longer a constraint is perpetually maintained to keep it interesting, because the alternative omnipotence is no fun. Those constructs will be so disconnected from reality, that we won't be able to acknowledge their existence.