https://video.ut0pia.org/videos/watch/8cc4e59d-73ea-46e1-8da0-2b4b97ff4f54 Feed it 67 videos from the 2022 World Cup and ask for the near misses, the shots that almost scored but did not, each with a reason, and it returns them. Ask it to track Messi across the entire corpus and describe the camera framing, and it finds the moment he slaloms past a sliding defender. James Le's point is that this is unusual because video has no memory. Almost every video AI system answers each query from scratch, and a bigger context window does not fix it, because the real problem is that there is no durable representation to retrieve into. His fix is to stop treating video as a bag of frames and start treating it as a spatial temporal volume, then build a memory layer over it: a context graph of time bounded moments, the entities and where they appear, the relationships between them, and corpus level themes. At TwelveLabs that is an embedding encoder, a context store, and a video language model exposed as an API. The design rules are blunt: ingest once and reason many times, store primitives not answers, ground every claim to a timestamp, and let intent decide what to remember, because brand safety and sports highlights need different things from the same footage. Speaker info: https://x.com/le_james94, https://www.linkedin.com/in/khanhnamle94/, https://jameskle.com/, Timestamps: 0:00 - Video has no memory 0:50 - Video is a spatial temporal volume, not a bag of frames 2:06 - Three problems: wrong context, wrong memory, weak reasoning 3:36 - Five properties that make video memory hard 4:53 - The TwelveLabs stack: Marengo, the context store, and Pegasus 5:56 - Search versus memory 7:48 - The context graph 9:04 - Five design principles for a video memory layer 10:45 - From a static model to a video worker 12:51 - Demo: sports and tracking Messi across the World Cup 15:37 - Demos: traffic security and ad placement Thu, 23 Jul 2026 09:53:54 GMT https://validator.w3.org/feed/docs/rss2.html PeerTube - https://video.ut0pia.org https://video.ut0pia.org/lazy-static/avatars/0287a09a-aae7-4840-9843-b416426e7046.webp https://video.ut0pia.org/videos/watch/8cc4e59d-73ea-46e1-8da0-2b4b97ff4f54 All rights reserved, unless otherwise specified in the terms specified at https://video.ut0pia.org/about and potential licenses granted by each content's rightholder.