(Many) of us want home robots. What's the holdup?

MIT CSAIL's Improbable AI Lab Director Pulkit Agrawal explains his "SoftMimic" approach to making robots safer around humans. Paper: https://arxiv.org/abs/2510.17792 Director: Rachel Gordon Preditor: Tim Malieckal Script Supervisor: Alex Shipps Participants: Pulkit Agrawal, Principal Investigator, MIT CSAIL Dr. Gill Pratt, Chief Scientist for Research, Toyota Motor Corp (TMC) Neville Hogan, Prof. Mechanical Engineering, Director of Newman Lab Gabriel Margolis, Improbable AI Lab Michelle Wang, Improbable AI Lab Venki Parsuram, Improbable AI Lab Gabe the Robot Special Thanks: Kendall Penington, Head of Communications @ 1x Technologies Wil Norton, Electrical Engineering and Computer Science @ MIT