There's a qualitative difference between fundamental and incremental innovation.
An example of a fundamental innovation would be radio. Before radio there was... no radio. Nobody could communicate across vast distances invisibly and without sound. After radio, they could. It's an absolute, sharp discontinuity.
An example of an incremental one would be multi-core processors, self-driving cars, or reusable rockets that land themselves. These are awesome don't get me wrong but they're refinements of things we already have. We already had cars, CPUs, and rockets, and now we have better ones.
The line is not perfectly sharp. Were ICs fundamental or just a refinement of transistors? (I'd argue that the photographic down-projection of circuitry in that way was fundamental.) But I think it's fairly clear to me that nothing of even the magnitude of the IC has come about since 1970. Certainly there hasn't been anything of the magnitude of radio, electromagnetism, powered flight, or rocketry in a long time. The last was probably understanding DNA and the DNS->RNA->protein synthesis pathway.
At this point I'm fairly convinced that this doesn't work, but if it had worked it would have qualified:
https://en.wikipedia.org/wiki/RF_resonant_cavity_thruster
"Cold" fusion, had it actually worked, would also have qualified. Something like that would convince me we have not entered a new dark age.
A possibility I didn't mention in my OP is that we've hit the limits of human cognition. I find it hard to believe that we understand "all of science," but maybe we are close to understanding all that the brain of homo sapiens can process.
- CRISPR
- Machine Learning (the unsupervised variety, specifically)
- Quantum computing (not there yet)
I think these are the three we are "on the cusp" on. They're the ones we'll likely look back on 20 years from now and say "yeah, that sure changed everything, didn't it?"
CRISPR-Cas9 is a nuclease, it cuts both strands of the DNA. Doing that to your DNA is a Very Bad Thing and usually turns out poorly regardless of whether you're trying to introduce new DNA into that area. If you've got mouse cells, then cool you only care if you get a few cells get the mutation you want, and if the others are screwed who cares? You get to move on with your experiment radically more quickly than you used to be able to. But that same equation doesn't apply to humans.
As for fundamental new things:
1. There's some research in fmri neurofeedback and magnetic stimulation on how to train/affect highly targeted brain regions to achieve variety of psychological effects including spiritual effects that take experienced meditators years to achieve, achieved in short periods of time.
2. lab on chip technology enables you to measure "everything" in the blood .Could lead to complete and fully accurate diagnostics. See healthtell.com