{"type":"video","version":"1.0","html":"<iframe src=\"https://www.loom.com/embed/fb16e5424cd34f1798f7f0eb04c26f9f\" frameborder=\"0\" width=\"1506\" height=\"1129\" webkitallowfullscreen mozallowfullscreen allowfullscreen></iframe>","height":1129,"width":1506,"provider_name":"Loom","provider_url":"https://www.loom.com","thumbnail_height":1129,"thumbnail_width":1506,"thumbnail_url":"https://cdn.loom.com/sessions/thumbnails/fb16e5424cd34f1798f7f0eb04c26f9f-e0abffc2b1984e49.gif","duration":246.9927,"title":"Revolutionizing Liquidity for Semi-Fungible Assets in AI Markets 🚀","description":"Hi everyone, in this video, I demonstrated a new market mechanism designed to address liquidity issues for semi-fungible assets, particularly in the context of an AI compute market. We explored how our system allows for automatic upgrades of assets, consolidating fragmented liquidity into a single fluid pool while ensuring fair pricing through incentive-compatible payments. I showcased how different buyer preferences are modeled to maximize welfare, and we successfully cleared the market using convex optimization. This architecture can be applied to various asset classes with quality tiers, and I encourage you to think about how this could impact your own markets."}