improving pitching

SETUP.md

@@ -26,7 +26,9 @@ The formal model assumes each session is generated by a latent **actor class** $
 
 ## 2. Business fit framing
 
-**What PHANTOM is for:** Studying how **automated browsing and transaction orchestration** interact with **session-based pricing**: behavior generates a demand proxy $\hat{q}$; pricing policies map interaction history to prices; **Cost of Information (COI)** is the premium the platform can sustain above a floor when information is scarce. Agent-mediated **reconnaissance in one session** and **purchase in another** undermines that asymmetry; the thesis proves a **COI erosion** mechanism under many independent price queries.
+**The problem PHANTOM addresses:** Session-based pricing accumulates demand signals across a user's browsing history and raises quoted prices accordingly—the **Cost of Information (COI)** premium. LLM agents undercut this by separating reconnaissance (many isolated sessions, no signal accumulation) from execution (a clean session that quotes a floor price). The thesis proves that as the number of independent querying agents grows, the realizable price collapses to a minimum order statistic and COI approaches zero.
+
+**What PHANTOM gives you:** A controlled platform to measure how much COI is at risk under real agent traffic, simulate that risk across contamination levels $\alpha \in [0,1]$, and train pricing policies that remain robust. The pipeline runs from raw interaction logs through behavioral kernel estimation and a contamination generator to a DR-RL gym.
 
 **What you must supply:**