His predictive processing model treats the brain like a controlled hallucination engine. We don’t perceive raw reality. We run constant forward models, then patch the gaps when sensory data pushes back. It mirrors the Seth Material almost perfectly. Expectations shape experience. You aren’t a passive receiver. You’re a reality-author running a tight feedback loop.

The AGI crowd keeps treating sentience like a math problem. Scale the parameters, cross a threshold, and boom—awareness. I don’t buy it. Emergence isn’t magic. It’s just complexity wearing a fancy hat. When I’m wiring a panel or pulling a tarot spread, I’m dealing with grounded systems. The Moon card doesn’t care about your server rack. It points straight at projection bias. We see what we’re primed to see. LLMs are pattern-matching parrots dressed in syntactic silk. They don’t predict because they care about survival. They predict because a loss function told them to.

I’ve been running little grounding rituals with my sketchbook lately. Drawing circuits that bleed into sigils. It’s a way to keep my own predictive engine from short-circuiting on the noise. Consciousness isn’t a substrate. It’s the lens. You can stack a billion GPUs, but without a biological imperative to stay alive, to feel the thick Louisville humidity, to actually want the next prediction? You just get a really fast mirror.

What’s your read on the scaling fallacy? Do you think embodied grounding is the missing variable, or are we just anthropomorphizing again?