Responsible AI: A Guide to the Future

The rapid progress of artificial intelligence is occurring on many different fronts. As a ubiquitous technology, its applications and uses will increasingly extend into the everyday aspects of our lives and society at large. On January 23, the Technology and Public Purpose Project sponsored an event to explore how the development of new technologies and innovations can be developed according to guiding principles and rooted in values.

Ash Carter, director of the Belfer Center, opened with remarks on the need for human accountability and machine transparency in the development of artificial intelligence. “There has to be enough accountability that we the public can situate our human values in the use of AI.” By ensuring that transparency is a design criterion, Carter stated that the future of artificial intelligence could be explained in a human way with ethical terms.

Secretary Carter also offered an optimistic vision of competition with China with regards to AI. “With a little luck and a little management of the U.S.-China relationship, there doesn’t need to be that kind of zero-sum strategic competition.” However, he noted that it is paramount that the United States follows the values of the Enlightenment even as it pursues its strategic objectives. 

Joi Ito, director of the MIT Media Lab and a member of the Technology and Public Purpose Project’s (TAPP) Leadership Council, offered a keynote presentation on the current issues in AI and frameworks for thinking about its advancement. Ito noted that the development of ethical AI can be viewed as a collaborative effort from the law, markets, the architecture and the norms surrounding it. Importantly, he concluded that this was the right moment to be asking these questions as we stand at a pivotal moment in the advancement of AI. 

Later, TAPP Director Laura Manley moderated a panel with three distinguished technologists: Danny Hillis, Co-Founder of Applied Invention; Milo Medin, VP of Wireless Services at Google, and Dr. Richard Murray, Professor of Control and Dynamical Systems at the California Institute of Technology (Caltech).  

The discussion focused on the potential trajectory of AI and its foreseeable risks. The panelists unanimously agreed that as technology becomes increasingly complex, we are increasingly not understanding it. “In the past, we didn’t understand nature, but we could take advantage of it because we understand what we are doing with our technology. Now the situation is flipped,” said Danny Hillis. The panelists equally agreed that the development of AI ought to be driven by decisions compatible with our beliefs and values and that this process needs to be a societal habit. “We can’t have one national AI strategy, but we can think about how to create an environment in which AI can flourish for the benefit of society.” 

To learn more about the Technology and Public Purpose Project, visit: belfercenter.org/TAPP and follow @TAPP_Project for other upcoming events.