I generally agree, and I think interpretability is a fools errand for any sufficiently complex nonlinear model. That said, I’d be surprised if there weren’t e

I generally agree, and I think interpretability is a fools errand for any sufficiently complex nonlinear model. That said, I’d be surprised if there weren’t eventually successful breakthroughs from the fields of nonlinear dynamics / pattern formation.