Berkeley AI Research - BAIR

AI can feel sudden. Disruptive. Everywhere at once.

For Cal Professor Stuart Russell, the questions began long before the headlines.

As co-author of Artificial Intelligence: A Modern Approach, he helped define how generations of students learn the field. In Human Compatible, he brought the challenge of AI alignment into public debate. In Provably Beneficial Artificial Intelligence, he examined how to design systems that are safe by construction.

Last year, EECS Professor Russell was awarded the AAAI Award for Artificial Intelligence for the Benefit of Humanity, honoring his work on the conceptual foundations of provably beneficial AI and his leadership in building the field of AI safety.

His research challenges a long-standing assumption in AI: that machines should optimize fixed objectives. Instead, he proposes treating AI as an “assistance game,” systems designed to support human interests while remaining uncertain about them.

As conversations around regulation and safety accelerate, the thinking behind them has been unfolding at Cal for decades.

The technology may feel new. The questions guiding it are not.

As part of the BAIR Computer Vision Workshop, a panel of alumni spanning 30 years of computer vision history at Berkeley, moderated by Lisa Dunlap.

In one hour, Prof. Anca Dragan will present the third distinguished lecture on the status and future of AI, on "AI Agents That Do What We Want: Progress and Challenges."

Livestream:

Anca Dragan - AI Agents That Do What We Want: Progress and Open Challenges Researchers used to define objectives for artificial intelligence (AI) agents by hand, but with progress in optimization and reinforcement learning, it becam...

Join the first Distinguished Lecture on the Status and Future of AI on Wednesday at 12pm PT.

Livestream links:

CITRIS, CDSS and BAIR Present: Jaron Lanier on Data Dignity in AI - CITRIS and the Banatao Institute Microsoft's Jaron Lanier presents "Data Dignity and the Inversion of AI" on Sept. 13, in the Banatao Auditorium. This seminar will be livestreamed.

Join us in an exciting series of distinguished lectures on The Status and Future of AI featuring Jaron Lanier, Alison Gopnik, Anca Dragan and Timnit Gebru.

Register for attending in-person or online: https://citris-uc.org/news-events/citris-research-exchange

Berkeley AI researchers developed a novel brain implant that just enabled a paralyzed woman speak after 18 years!

A Stroke Stole Her Ability to Speak at 30. A.I. Is Helping to Restore It Years Later. The brain activity of a paralyzed woman is being translated into words spoken by an avatar. This milestone could help others who have lost speech.

Berkeley joins the NSF-backed AI institute for cybersecurity, with five BAIR faculty members joining the executive committee

UC Berkeley joins NSF-backed AI institute for cybersecurity - Berkeley Engineering Five faculty members part of executive committee

Exclusive sneak-peek into the life of ChatGPT architect and OpenAI co-founder John Schulman, check out the full interview here:

ChatGPT architect, Berkeley alum John Schulman on his journey with AI In an interview, Schulman discussed his time at Berkeley, why he cofounded OpenAI and the future of artificial general intelligence

Wondering what to watch this weekend? Check out this thought-provoking and timely lecture by Prof. Stuart Russell, co-author of the most popular AI textbook in the world, on "How Not to Destroy the World with AI":

Stuart Russell calls for new approach for AI, a ‘civilization-ending’ technology “Intelligence really means the power to shape the world in your interests,” professor says in campus talk

The first of the UC Berkeley lectures on the status and future of Artificial Intelligence is *today* at 4 PM PT! Tune in to Prof. Jitendra Malik's lecture for free on YouTube and preview the talks that are up next:

ArtificiaI intelligence - University of California, Berkeley ChatGPT is catalyzing a paradigm shift in artificial and natural intelligence. How does it work and why is it surprising experts? And what are the implications for the future? As a world leader in artificial intelligence with a history of challenging convention, UC Berkeley is shaping the future of....

Woody was a Berkeley AI PhD advised by Prof. Pieter Abbeel. Congrats on the successful launch, Woody! 🚀

Congratulations to Berkeley Engineering alum Warren “Woody” Hoburg (M.S.’11, Ph.D.’13 EECS) and the successful launch of the SpaceX Crew-6 Mission to the International Space Station.

We got to know more about Woody and how he became a spacecraft pilot last year, https://bit.ly/3IwLpXJ.