Tech Week / San Francisco
Learning From Human Video & Scaling VLA Training
A discussion of robotics learning, VLA and world models, focusing on collecting and scaling human demonstration video data to train physical AI.
Bright Data
SOMA
Video here is training data, not AI video generation.
Organizer event page ↗Source · checked 2026-10-12
AFTER THE EVENT
Videos, recaps & related posts
Matched to this event, organizer and edition. Pre-event announcements are labeled separately; follow the publisher for the original content.
Recaps & recordings
-
Attendee report: Human video and VLA panel
Attendee report and photos from the October 7 session.
Lars-Fredrik Forberg / Humanoid.guide · Checked 2026-10-12