Add Barrett's constructed emotion theory to Chapter 14

What

Integrate Lisa Feldman Barrett's constructed emotion theory into the chapter's argument about why we project fear onto AI. The theory provides the neuroscience foundation: emotions require interoceptive biology (body-budget signals) that AI lacks entirely. This transforms the "AI might develop self-preservation drives" argument from "unlikely" to "biologically impossible given current architectures."

Why

The research pages flag this as completely unused despite being highly relevant. Chapter 14 argues that AI risk framing reflects human psychological projection — Barrett's work provides the concrete scientific basis for that claim. Without it, the argument rests on assertion rather than evidence.

How

In Chapter 14's section on fear and AI, after the passage arguing we anthropomorphise AI risk: add 2-3 paragraphs explaining that emotions emerge from interoceptive prediction (Barrett's framework), that this requires biological substrate AI doesn't possess, and that the implication — AI can model emotions without experiencing them — fundamentally changes the risk calculus.

Impact

Transforms Chapter 14's central argument from philosophical claim to empirically grounded position. Addresses the strongest version of the AI existential risk objection.