The Center on Global Economic Governance hosted a panel discussion on “Artificial Intelligence: Implications for Governance and Public Policy" on February 23, 2018.
Robin J. Lewis, chair of the Worldview Global Culture Alliance and former SIPA associate dean, Ronaldo Lemos, a visiting professor at Columbia SIPA, and Liangang Sun, a professor at Peking University touched on the wide-ranging consequences of artificial intelligence (AI) in areas such as urbanization, jobs and employment, and proper governance. They also discussed ethical and legal questions related to ownership of data, privacy rights, and the potential for AI to replace politicians in the policymaking process.
In opening remarks, Lewis noted the current hype surrounding AI and how there is little consensus of its long-term significance. He highlighted the long tradition of imaginative works in popular culture--from Frankenstein to 2001: A Space Odyssey--related to AI. As these works illustrate, machines can make human lives better or worse. For this reason, it is important to begin a global dialogue on AI, which will only grow in importance and impact industries ranging from retail to finance to transportation.
Government is another sector ripe for influence. Lewis gave examples of AI concepts that involve the public sector, such as smart cities, facial recognition technology, and crime prediction algorithms. Each subject has exciting potential as well as headwinds in terms of its development and implementation. Lewis underlined two important AI-related questions for the public sector: how does government evaluate and regulate technology without stifling innovation? And) how can technology best be used to improve lives? In his conclusion, Lewis discussed the discrepancies between the two global AI powerhouses—the United States and China. In the U.S., most research on AI is completed in the private sector, by firms such as Google and Amazon, whereas China has adopted a more state-centric approach.
Ronaldo Lemos focused his talk on the developing world and public policy. He noted how AI has become very good at playing games, such as Chess, Go, and potentially even soccer, though it is the potential influence on public policy that is critical. An important policy question for Lemos is how to keep AI in the public interest. Will it create or erase more jobs? While some studies estimate 47 percent of U.S. jobs will be automated in the next 20-30 years, other researchers argue this figure is much lower, around 9 percent. Whether one ascribes to the optimistic or pessimistic projection, it appears certain that some sectors and jobs are more at risk than others. Taxi drivers, for example, are more at risk than occupations such as nursing.
Lemos noted that artificial intelligence often transcends political preferences. He gave the example of former Brazilian President Dilma Rousseff, a leader of Brazil’s leftist Workers Party. During a visit to Silicon Valley Rousseff was impressed by the potential of driverless cars and wanted them brought to Brazil, despite the fact that such technology could lead to millions of job losses for Brazilian taxi drivers
The final panelist, Professor Liangang Sun, offered a radical departure from “traditional” thoughts on the implications of artificial intelligence. He argued that technology can be used to uproot conventional politics and a government of people can be replaced by a government of data. Sun believes greater trust should be placed in technology and algorithms to improve society. Belief in traditional politicians is currently low, and big data and AI can replace any imperfections in the legal system created by humans. Sun said the advent of Bitcoin reflects this major shift, and how people no longer have to trust money printed by national governments.
A lively Q&A session followed the panel discussion.
This event was organized in partnership with the Worldview Global Culture Alliance (WGCA) and Columbia SIPA Entrepreneurship and Policy Initiative.
- Kevin Gilmartin MPA ’18
