Embodied Intelligence & Robotics Center

We explore how intelligent systems learn and adapt through physical interaction with the real world, studying the fundamental principles of embodied intelligence.

Based at UC Berkeley, EMBER Center investigates the integration of generative AI with physical systems operating in unstructured environments. Our goal is to develop systems that can perceive, learn, and safely act in dynamic, unpredictable real-world settings.

Our Publications

EMBER x HIC Seminar

Talks from leading researchers in robotics and embodied intelligence.

Fall 2026

Most Wednesdays, 11 a.m. – noon Banatao Auditorium (Floor 3), Sutardja Dai Hall, UC Berkeley

No registration required. Hosted in person; recordings available afterwards.

Next up Wed, November 4, 2026

Matt Mason

Carnegie Mellon University

Why Is Robotics Hard?

Location change: Gateway 4355

Abstract

Is robotics hard? Moravec's paradox suggests that it is. Moravec's paradox says that the things we expected to be hard, like playing champion-level chess, turned out to be easy. And the things we expected to be easy, like moving the chess pieces, turned out to be hard. This seems to be born out in recent experience — after 50 years of work our robots still cannot move chess pieces as well as any eight-year-old child. One explanation is that manipulation is fundamentally hard. But there is another explanation: roboticists are dense. This talk will explore both possibilities.

Coming up

Nov 18

Karen Adolph · NYU
Dec 2

Stephanie Gil · Harvard University
Dec 4

Luca Carlone · MIT

Full schedule & past recordings

Our Sponsors

We're grateful for the support of these organizations that make our research possible.