Robotics & AI

Is spaghetti the key to building a better robot?

1/22/2020 - Studying how spaghetti reacts to water might offer clues to how robots built from flexible materials can better mimic human movement, according to Oliver O'Reilly, professor of mechanical engineering.

Building a world where data privacy exists

Dawn Song in her office

11/18/2019 New York Times - Electrical engineering and computer sciences professor Dawn Song, a leading expert in computer security and trustworthy artificial intelligence, is building a platform in which people control their own data online and are compensated for its use by corporations.

Roach-inspired robot

10/25/2019 Engineers have created an insect-sized robot that can move as fast as a cockroach and withstand the weight of the average human.

Moments of untruth

Photo illustration of Barack Obama, Mark Zuckerberg and Elizabeth Warren

10/25/2019 Computer science researchers are using groundbreaking machine learning technologies to expose deepfake videos, manipulated images and other types of digital deception.

With AI, machines become expert at reading brain scans

Two brain scans: One as normally seen by a radiologist, and one with hemorrhaging areas highlighted by AI technology

10/22/2019 - A computer algorithm developed by scientists at UCSF and UC Berkeley bested two out of four expert radiologists at finding tiny brain hemorrhages in head scans — an advance that one day may help doctors treat patients with traumatic brain injuries, strokes and aneurysms.

U.S. Military researchers work to fix easily fooled AI

Dawn Song

9/19/2019 NPR - NPR's All Things Considered talked to EECS professor Dawn Song about her AI work with traffic signs to explain how U.S. Military researchers are working to combat what they call "adversarial artificial intelligence." That's when someone hacks into an AI system to transmit the wrong information.

Warning! AI is heading for a cliff

Stuart Russell

9/19/2019 California Magazine - Electrical engineering and computer sciences professor Stuart Russell proposes a solution to AI’s fundamental "design error."