🎧 START pod: Lucas Ngoo, Co-founder & CEO, Cortex AI "The Real World Is the Next Training Ground for Embodied AI"

The internet was the training set for intelligence

Nobody has built the equivalent for the physical world

Previously, Lucas Ngoo co-founded Carousell, scaled it past $1B 

Now at Cortex AI he's collecting the data robotics labs need to train foundation models

Cameras, 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)

Not building the robot. Not training the model. Collecting what goes in

2026: scale data in a big way 
2027+: start rolling robots out, 90% teleoperation, 10% autonomy

While humans step in, the system keeps learning

That's the flywheel toward full autonomy

01:38 Cortex as the data layer for general-purpose robotics
02:19 Why text data doesn’t transfer cleanly to the real world
02:51 Recording day-to-day human manipulation work
03:23 “Tens of millions” to “hundreds of millions of hours” of data
06:37 Cortex’s role: “We are really the data piece”
07:28 When humanoids become part of everyday life
08:00 2026 as the year data scales
08:13 90% teleoperation, 10% autonomy
08:37 Data flywheel toward full autonomy
08:55 Synthetic data + real-world data reinforcing each other