{"type":"rich","version":"1.0","provider_name":"Transistor","provider_url":"https://transistor.fm","author_name":"START","title":"🎧 START pod: Lucas Ngoo, Co-founder & CEO, Cortex AI \"The Real World Is the Next Training Ground for Embodied AI\"","html":"","width":"100%","height":180,"duration":646,"description":"The internet was the training set for intelligence\nNobody has built the equivalent for the physical world\nPreviously, Lucas Ngoo co-founded Carousell, scaled it past $1B \nNow at Cortex AI he's collecting the data robotics labs need to train foundation models\nCameras, VR headsets, glasses on factory workers, retail workers, everyday people. Recording real-world manipulation work (Maybe tens of millions, even hundreds of millions of hours)\nNot building the robot. Not training the model. Collecting what goes in\n2026: scale data in a big way \n2027+: start rolling robots out, 90% teleoperation, 10% autonomy\nWhile humans step in, the system keeps learning\nThat's the flywheel toward full autonomy\n01:38 Cortex as the data layer for general-purpose robotics\n02:19 Why text data doesn’t transfer cleanly to the real world\n02:51 Recording day-to-day human manipulation work\n03:23 “Tens of millions” to “hundreds of millions of hours” of data\n06:37 Cortex’s role: “We are really the data piece”\n07:28 When humanoids become part of everyday life\n08:00 2026 as the year data scales\n08:13 90% teleoperation, 10% autonomy\n08:37 Data flywheel toward full autonomy\n08:55 Synthetic data + real-world data reinforcing each other","thumbnail_url":"https://img.transistorcdn.com/q-E1hh7K6IS4AfZiNy2p4MYVGcUOO8lQP92h8QbOEOA/rs:fill:0:0:1/w:400/h:400/q:60/mb:500000/aHR0cHM6Ly9pbWct/dXBsb2FkLXByb2R1/Y3Rpb24udHJhbnNp/c3Rvci5mbS80NzVj/MDEzNjkxNjU1N2Uy/NDFhMDQ3M2ZhNWI3/NWY0MS5wbmc.webp","thumbnail_width":300,"thumbnail_height":300}