Intelligence has to happen somewhere.
Biologically inspired architectures make the substrate part of the question. Neural computation, continuous systems, and hardware/software co-design all suggest ways to think beyond ordinary discrete software.
The interest is in the relationship between intelligence and physical systems: how dynamics, constraints, and structure participate in computation.
A direction, not a conclusion
This territory meets dynamical systems and control. It also leaves a question open: which aspects of intelligence are properties of a model, and which arise from the system in which the model exists?
A working entry. The connections may change.