## A7. Evolutionary Psychology and Meta-Insights **Supplementary finding:** ADHD traits — rapid environmental scanning, novelty-seeking, relational cognition — were highly adapted to high-stimulation ancestral environments. Barack et al. (2024) confirmed this experimentally: ADHD individuals depart resource patches sooner in foraging tasks, consistent with an exploration-biased strategy. Modern low-stimulation contexts cause "G Collapse" (emotional volatility, burnout, profound executive dysfunction). Generative AI providing rapid-fire stimulation, dialogue, and novelty satisfies the dopaminergic requirements that modern environments fail to meet. **Meta-insight across all domains:** The populations who need these tools most benefit from them the most. Toli et al. found implementation intention effects of d = 0.99 in clinical populations vs. d = 0.65 in general populations. Joo et al. found spacing interventions specifically help those with elevated visual crowding. Kofler et al. found 75–81% of ADHD cases show the WM deficits that make local-first architecture necessary. A well-designed tool's efficacy curve is steepest for the most impaired users. **Implication for Kon:** The app should feel alive, not static. The convergence of voice-first interaction (reduces navigation complexity), local-first architecture (eliminates latency), and AI presence (provides external regulation) addresses different links in the same causal chain. Each feature amplifies the others.