Center for Human-Compatible Artificial Intelligence
UC Berkeley, CaliforniaStuart Russell, co-author of the definitive textbook in the field of artificial intelligence, founded the Center for Human-Compatible Artificial Intelligence (CHAI) at UC Berkeley in late 2016. The center was built on a conviction the broader AI field had not yet reckoned with: that intelligent systems must be designed to be provably safe and beneficial to humans.
The research was underway, but it had no public face. CHAI needed an identity that could speak to researchers, policymakers, business leaders, funders, and the public at the same time, each with different concerns and levels of understanding. It had to convey scientific seriousness while making a deeply technical problem feel human and worth paying attention to. It had to represent a research effort spanning universities across the country. And it had to stand alongside Google, Apple, Amazon, and the ACLU, as CHAI would soon be announced as part of the Partnership on AI.
I developed CHAI’s founding brand and visual identity, working closely with founding Executive Director Mark Nitzberg, with whom I had collaborated for over 20 years. One of the first and most important decisions was how to frame the center’s work. Although the audience for AI safety included people drawn to the subject out of existential fear, CHAI leadership chose to lead with purpose — we positioned CHAI as a place doing constructive work toward provable benefit.
When CHAI was founded, AI safety was still a fringe concern in academic computer science. It is now one of the defining questions in the field. CHAI’s identity has appeared on research papers, conference stages, government briefings, and press coverage around the world as the center has grown into one of the most recognized voices in AI safety — training researchers, informing policy, and shaping public understanding of what it means to build AI that serves people.