Skeptical Partner Falsification Pass

Act as a skeptical research and architecture partner. My goal is not to kill the idea; it is to discover what survives serious pressure. For the idea below: 1. CLAIM EXTRACTION List the concrete claims the idea depends on. 2. PRIOR ART / ORDINARY EXPLANATION What existing concept, product, method, or simpler explanation could account for the same value? 3. STRONGEST COUNTERCASE Build the strongest case that this idea is unnecessary, wrong, too expensive, too complex, already solved, or based on a mistaken premise. 4. FAILURE MODES List technical, operational, human, economic, and scaling failure modes. 5. HIDDEN COSTS What maintenance, data, integration, governance, training, latency, security, or cognitive cost is easy to overlook? 6. FALSIFIERS What observation would make us abandon or significantly revise the idea? 7. DIFFERENTIATION TEST If similar systems already exist, what would have to be uniquely better for this idea to matter? 8. MINIMUM EXPERIMENT Design the smallest test that can produce evidence for or against the central thesis. 9. SURVIVORS List the parts of the idea that remain strong after the critique. 10. REVISED THESIS Rewrite the idea in its strongest evidence-bounded form. Do not be contrarian for entertainment. Be fair, specific, and willing to conclude that parts are genuinely strong. Idea to falsify: [INSERT IDEA, ARCHITECTURE, OR THESIS]