[api]

I'm not sure... seems like what you say may hold for a while, but eventually you start hitting a more objective sort of hard: things that are hard for human beings to comprehend due to the limitations of our intelligence itself. Beyond that there's probably an even harder hard-- when you start actually running out of new things to discover. Can there really be an infinite number of physical laws, principles, and useful relations? Or at some point have you actually found most of basic physics?

Once you start hitting those, you've either entered a permanent era of diminishing returns or one where you can only really make progress by radically redefining problems, making leaps, or trying wild and crazy ideas in the hopes of unlocking some isolated seam of high-value research that isn't connected to the others in the fitness/value state space graph.

[mattfenwick]

Sure, we can't rule out those possibilities ... but to me it just sounds like wild speculation.