From 14aae3dc9ae24f9bab73a719ebbd006a06c6327c Mon Sep 17 00:00:00 2001
From: Daniel Rosel <daniel@alves.world>
Date: Tue, 10 Mar 2026 14:52:17 +0100
Subject: [PATCH] improving understandable aspects of the abstract

---
 docs/index.html | 12 ++++++++----
 1 file changed, 8 insertions(+), 4 deletions(-)

diff --git a/docs/index.html b/docs/index.html
index 872f461..82e1483 100644
--- a/docs/index.html
+++ b/docs/index.html
@@ -319,10 +319,10 @@
         <h2 class="title is-3">Abstract</h2>
         <div class="content has-text-justified">
           <p>
-            PHANTOM formalizes a mechanism failure in dynamic pricing under non-human transaction orchestration. LLM-based agents can run reconnaissance in isolated sessions and execute purchases in clean sessions, reducing the platform's ability to extract the Cost of Information (COI), the premium usually generated by demand signal expression.
+            When you shop online, prices often change based on how much interest you show — the more you browse, the more the site learns about your intent and may raise prices accordingly. This works because stores assume that a curious, engaged shopper is more likely to buy. But AI assistants are now doing the shopping research on behalf of users: they browse in one session to gather price information and then let the user purchase in a fresh session at the lower, unadjusted price. The store never sees the connection between the two, so it never gets to factor in that genuine intent — and loses the revenue it would have earned.
           </p>
           <p>
-            The project combines behavioral modeling and robust control. We built a controlled e-commerce platform (hotel and airline modes), logged full interaction trajectories and price exposures, learned separable human/agent transition kernels, and used those signals to train contamination-aware pricing policies with a distributionally robust reinforcement learning objective.
+            PHANTOM studies this problem and builds defenses against it. We created a realistic fake store (in hotel and airline modes) where both real people and AI agents were given shopping tasks, and we recorded every click, scroll, and page visit. By comparing how humans and AI agents move through a site, we found clear patterns that tell them apart. We then used those patterns to build a smarter pricing system that can recognize when it is likely talking to an AI scout and adjust its strategy accordingly — protecting the store's margins without making things worse for genuine shoppers.
           </p>
         </div>
       </div>
@@ -353,6 +353,7 @@
 
 
 <!-- Image carousel -->
+<!--
 <section class="hero is-small">
   <div class="hero-body">
     <div class="container">
@@ -385,6 +386,7 @@
 </div>
 </div>
 </section>
+-->
 <!-- End image carousel -->
 
 
@@ -403,8 +405,10 @@
           <h2 class="subtitle has-text-centered">COI from first principles.</h2>
         </div>
         <div class="item item-video2">
-          <img src="static/images/banner.svg" alt="PHANTOM unified teaser banner" width="1920" height="1080" decoding="async" style="display:block; width:100%; height:auto;" onerror="this.onerror=null;this.src='static/images/carousel2.jpg';"/>
-          <h2 class="subtitle has-text-centered">Unified teaser: vulnerability, behavioral kernels, and robust control loop.</h2>
+          <video poster="" id="video2" controls muted loop height="100%" preload="metadata">
+            <source src="static/videos/BehaviorKernelConstructionScene.mp4" type="video/mp4">
+          </video>
+          <h2 class="subtitle has-text-centered">Behavioral kernel construction: learning how humans and agents differ.</h2>
         </div>
         <div class="item item-video3">
           <video poster="" id="video3" controls muted loop height="100%" preload="metadata">