Research Program Draft
The corpus rewrites folk-psychological phenomena as equilibria of mechanisms with thresholds. Two questions sit underneath that method.
Why this scale
Classical science has measured the scales below the human (atoms, cells, neurons) and above (climate, markets as aggregate) in great depth. The scale between two or N agents interacting through an architecture is methodologically almost empty. Most of what now feels like modern subjective experience is decided on that scale. The marginal return on understanding one more electron is near zero for daily life; the marginal return on formalising one previously unmodelled inter-agent mechanism is large.
Modeling under observer effects
Modeling agents perturbs them. Publishing the model of σ_R changes how people ghost; publishing the platform food chain changes how creators read their own role. This is real. It is not grounds to refuse the work. Impossibility has facets, and each facet allows a different attack.
- Information-theoretic. Capturing the full state would need infinite data. Attack: reduce to a sufficient statistic. Cartel or Schumpeter separates two regimes that look identical under any concentration index by reading a single eigenvector asymmetry.
- Reflexive. The model perturbs the system being modeled. Attack: publish the model together with the architectural change that defines the new equilibrium. Ghosting Isn't Cowardice proposes Hinge's "Your Turn" as exactly that move.
- Conceptual. No agreed-upon target object exists. Attack: show that the missing scalar ground truth was never the right object. Personality Is Not Who You Are argues "personality" is a cost tensor under a context operator.
Status: working draft, written in conversation with the corpus. Will be hardened into a standalone essay (sources/research-program.tex) once the framing stabilizes.