AI & robotics

Clever clumsiness: Self-taught walking robot

01/09/19 Wired — Researchers from the Berkeley AI Research Lab and Google Brain have taught a robot to teach itself how to walk, through trial and error, in just two hours; the results "are as awkward as they are magical."

Artificial intelligence opens health data privacy to attack

12/21/18 — Current privacy laws and regulations are nowhere near sufficient to keep an individual's health data private in the face of advances in artificial intelligence, according to a new study from IEOR professor Anil Aswani and his team.

Robots in the Berkeley Artificial Intelligence Research (BAIR) Lab

These robots are learning the old-fashioned way—by playing

12/19/18 California Magazine — Unlike most robots, the ones in the Berkeley Artificial Intelligence Research (BAIR) Lab haven't been programmed to perform a specific task. Instead, they've been programmed to learn new stuff by observation or through physical trial and error.

Watch just a few self-driving cars stop traffic jams

11/19/18 Science — Anyone can start a traffic jam - just by tapping the brakes. Now, scientists at Berkeley have shown that a few self-driving cars can prevent such jams - and in some cases double the average speed of surrounding vehicles.

Doctored photo of shark on a Houston highway

Fighting fake news

11/14/18 — Berkeley students have created SurfSafe, a machine learning tool that identifies when an online photo has been doctored.

Assembling an ionocraft microrobot under a microscope

Microrobots fly, walk and jump into the future

11/14/18 — Researchers led by Kris Pister are overcoming significant technical hurdles to push the boundaries of robotic miniaturization.

Turning cars into robot traffic managers

10/29/18 — Berkeley transportation researchers are addressing the emerging era of smart vehicles with a project that uses machine learning to manage traffic where autonomous, semi-autonomous and manned vehicles share the road. They presented their project, called Flow, at the Conference on Robotic Learning.

Computer science students waiting to enter a job fair at UC Berkeley

As companies embrace AI, it’s a job-seeker’s market

10/16/18 Reuters — With artificial intelligence expanding into ever more applications, the number of students trained in AI hasn't kept pace with the demand for workers - a mismatch reflected at a recent UC Berkeley career fair.

Cartoon of worker relaxing at desk while computer does work.

AI may not be bad news for workers

09/14/18 The Economist — Countering fears that artificial intelligence will eliminate millions of jobs, a new report co-authored by Berkeley Engineering roboticist Ken Goldberg says AI may instead enhance job satisfaction for ordinary workers by taking over mundane tasks.

AI-altered video makes it look like you can dance

08/27/18 Engadget — Experimenting with deep learning systems, a team of Berkeley computer science researchers is using artificial intelligence to make dancers look more skillful. The project uses algorithms to pass dance moves via video from a skilled dancer to a virtual skeleton to an amateur subject.

How robot hands are evolving

07/30/18 New York Times — Drawing on the latest in machine learning, Autolab and BAIR Lab researchers are developing robots that can pick up or move all manner of objects, and even learn to make a bed.

Roadway image showing bounding boxes over objects of note, like cars and traffic signals

Download the world’s largest self-driving dataset

06/07/18 Interesting Engineering — The Berkeley DeepDrive Industry Consortium has released an open-source dataset of 100,000 self-driving video sequences - largest of its kind - to the public in hopes of helping engineers develop autonomous vehicles.

Robot’s play

06/02/18 — Robots can manipulate objects they have never encountered before by building on previous play and observations.

Kiwi robot transporting food in downtown Berkeley

Those four-wheeled robots on campus, explained

05/31/18 — The presence of talented EECS student engineers, along with plenty of hungry students and a gourmet local food scene, goes a long way to explain the decision by Felipe Chavez to launch his robot-powered meal delivery startup Kiwi in Berkeley.

Stephanie Tena teaches middle schooler Bianca Castro coding and AI during an after- school club

The future of AI depends on high school girls

05/23/18 The Atlantic — From a coding summer camp for high schoolers at Stanford to a project showcase for teens and their mentors at Berkeley, AI4All is working to widen the pipeline of women entering the field of artificial intelligence.

Computer illustration of human-looking bot

Microsoft acquires Berkeley AI company to humanize its bots

05/21/18 The Verge — AI start-up Semantic Machines, co-founded and co-led by EECS professor Dan Klein, is being acquired by Microsoft, which hopes to make its Cortana digital assistant and other bots sound more naturally human by establishing "a conversational AI center of excellence in Berkeley."

How to interact with robots

05/04/18 Wired — EECS professor Anca Dragan discusses one of the big challenges in her field of human-robot interaction: getting robots to anticipate human behavior. “It's not enough to know what people are currently doing. They need to know what's going to happen in the future.”

AI: The revolution hasn’t happened yet

04/25/18 Medium — EECS and statistics professor Michael Jordan writes about the opportunity and imperative of developing a "human-centric engineering discipline" to address the ethical concerns of artificial intelligence.

Nimble robot makes strides in dexterity

03/26/18 MIT Tech Review — The latest iteration of Dex-Net, from the lab of robotics professor Ken Goldberg, could sort through your junk drawer with unrivaled speed and skill, using machine learning to determine how to pick up even odd-looking objects with incredible efficiency.

Robotic surgical arm picking up grains of rice from a moving platform

How flight simulation tech can help turn robots into surgeons

03/02/18 Wired — Robotics researchers from Berkeley's AUTOLab, led by IEOR and EECS professor Ken Goldberg, have built a heaving robotic platform - mimicking the motion of a breathing, heart-beating human patient - to help develop algorithms that robotic surgical assistants can use to guide their cutting.