I think the next test will be asking an AI to come up with a new branch of mathematics - just letting it rip and telling it to construct a system that doesn't reduce to combinatorics, group theory, graph theory, analysis, etc. Just get wild with it and don't start with any known problem as a jumping off point.
I think something like the Collatz conjecture will be solvable not as number theory or ergodic theory but some other completely wacky environment that humans haven't even sniffed at.
IMO that’s where AI is going: as soon as a problem can be formulated clearly enough, AI will trounce us humans. I have yet to see evidence that it can decide what problems are important at a remotely human level.