Berkeley Launches Center for Human-Compatible Artificial Intelligence

Berkeley Launches Center for Human-Compatible Artificial Intelligence

Funded in part by a $5.5 million grant from the Open Philanthropy Project, the University of California, Berkeley has announced the launch of a Center for Human-Compatible Artificial Intelligence.

Led by electrical engineering professor Stuart Russell, the center will work to ensure that AI systems are beneficial to humans and act in a manner aligned with human values. In addition to the grant from  OPP — a collaboration between Good Ventures, a foundation established by Facebook co-founder Dustin Moskovitz and his wife, Cari Tuna, and charity research nonprofit GiveWell — the center is supported by grants from the UK-based Leverhulme Trust and the Future of Life Institute, which was co-founded by Skype co-founder Jaan Tallinn.

One approach to AI that Russell and others are exploring is inverse reinforcement learning, through which robots learn about human values by observing human behavior. "Rather than have robot designers specify the values, which would probably be a disaster, instead the robots will observe and learn from people," said Russell. "Not just by watching, but also by reading. Almost everything ever written down is about people doing things, and other people having opinions about it. All of that is useful evidence."

Still, Russell and his colleagues expect to encounter challenges. "People are highly varied in their values and far from perfect in putting them into practice," said Russell. "These aspects cause problems for a robot trying to learn what it is that we want and to navigate the often conflicting desires of different individuals."

"UC Berkeley Launches Center for Human-Compatible Artificial Intelligence." University of California, Berkeley Press Release 08/29/2016.