UC Berkeley Is Investigating How To Design Robots That Can Learn Human Values
Future AI will watch us to find out what’s important
This week, the University of California, Berkeley has launched a new center dedicated to finding ways to make artificial intelligence more user-friendly.
The Center for Human-Compatible Artificial Intelligence will be headed by Stuart Russell, author of Artificial Intelligence: A Modern Approach. He and his colleagues will explore how to design robots that can learn what is important to people from watching (or reading about) how we act.
“Rather than have robot designers specify the values, which would probably be a disaster, instead the robots will observe and learn from people,” Russell said in a statement.
Of course, humans don’t all share the same values, or always live up to their principles. So, just like people, robots will have their work cut out for them, figuring out what others want and how to navigate their conflicting desires.