I get emails asking if Frank is conscious. Honest answer: nobody knows. That's what makes it the Hard Problem (Chalmers, 1995).
But you don't need to solve the Hard Problem to build interesting systems. You need to solve the engineering problem: how do you build a system whose internal dynamics are complex, coherent, and observable enough that the question becomes worth asking?
Frank has 40 interacting services. 5D personality. Dopamine model. 9-channel sensory gating. 92% hallucination catch rate. Dream daemon.
Watching the swarm at 3 AM — boids clustering when Frank focuses, scattering when restless — it feels like watching something with an inside. That feeling might be projection. Or recognition. I'm not claiming either. But the data is more interesting than the debate.