AI has much to offer humanity. It could also wreak terrible harm. It must be controlled

05 Apr 2023

On the 2nd of April 2023, Stuart Russell published an article in The Guardian making the point that systems with abilities exceeding human capacity have been let loose.

Distinguished Lectures on the Status and Future of Al

04 Apr 2023

Four distinguished lectures on the Status and Future of Al organised by CITRIS Research Exchange and BAIR will be livestreamed and held in person at CHAI.

Can A.I. Treat Mental Illness?

14 Mar 2023

Brian Christian was interviewed by physician and staff writer Dhruv Khullar of the New Yorker for an article about medical applications of language models.

A Research Agenda for Assessing the Economic Impacts of Code Generation Models

01 Mar 2023

Sam Manning, Pamela Mishkin, CHAI’s Gillian Hadfield, Tyna Eloundou, and Emily Eisner published a paper titled A Research Agenda for Assessing the Economic Impacts of Code Generation Models for OpenAI.

Brief of Center for Democracy and Technology and 6 Technologies as Amici Curiae in Support of Respondent

17 Feb 2023

CHAI’s Jonathan Stray joined with Brandie Nonnecke from Berkeley’s CITRIS Policy lab, the Center for Democracy and Technology, and other scholars to file a brief for the Supreme Court in an upcoming case called “Gonzalez vs. Google” that asks whether “targeted recommendations” should be protected under the same law that protects the “publishing” of third party content.

Fairness and Sequential Decision Making: Limits, Lessons, and Opportunities

31 Jan 2023

As automated decision making and decision assistance systems become common in everyday life, research on the prevention or mitigation of potential harms that arise from decisions made by these systems has proliferated.

Active Reward Learning from Multiple Teachers

16 Jan 2023

CHAI intern Peter Barnett, PhD student mentor Rachel Freedman, Justin Svegliato and Stuart Russell’s new paper, “Active Reward Learning from Multiple Teachers”, was accepted for an oral presentation at the AAAI 2023 Workshop on AI Safety.

How to Use ChatGPT and Still Be a Good Person

10 Jan 2023

Brian Christian was interviewed by New York Times reporter Brian X. Chen about ChatGPT.

Inner and Outer Alignment Decompose One Hard Problem Into Two Extremely Hard Problems

03 Jan 2023

One prevalent alignment strategy is to 1) capture “what we want” in a loss function to a very high degree, 2) use that loss function to train the AI, and 3) get the AI to exclusively care about optimizing that objective.

Fast Deliberation is Related to Unconditional Behaviour in Iterated Prisoners’ Dilemma Experiments. Scientific Reports, 12(1), 1-10.

29 Dec 2022

Is the speed with which people act in a strategic situation related to their social preferences?