Should the U.S. government spend research money to develop beerbots? It’s a question Sen. Jeff Flake (R-Ariz.) has asked, in aiming to prohibit the Department of Defense from spending money on robotic bartenders: “There are beerbots in the private sector already, so why would we devote resources for this?” This is not an idle thought. Flake has proposed an amendment to H.R. 6157, the Department of Defense and Labor, Health and Human Services, and Education Appropriations Act, 2019, to “prohibit the use of funds for the development of beerbots or other robot bartenders.”
The proposed law apparently is aimed at an MIT Computer Science and Artificial Intelligence Lab project, which is developing a bartender robot to hand beer to waiter robots. The project’s goal is “to control a group of robots interacting with an environment in order to cooperatively solve a problem.”
{mosads}Although dressed up ostensibly as an amendment to save taxpayer money from being “wasted” on beerbots, the Flake amendment targets the wrong problem. The United States should be spending more — not less — on robotics research. This is a strategic imperative, considering the massive advances in this domain being made by America’s chief rival, China.
At the World Robot Conference in Beijing this month, China showcased its advances: miniature robot tanks equipped with laser weapons; robot surgeons; the DaVinci robotic arm with a 540-degree rotation range; a gliding fox built by mimicking a bat’s actual wings; and a robot basketball player. China even has developed robots to serve as teachers’ aides and medical assistants. China’s robotics push is part of its “Made in China 2025” strategy designed to make the country the world’s leader in technology and artificial intelligence (AI).
A report by Tsinghua University notes, “From 2013 to the first quarter of 2018, China’s investment and financing in the AI industry accounted for 60 percent of the world’s total.” According to the report, China published “the largest number of AI-related research papers, as well as highly cited papers,” and “ranked first in the number of AI-related patents, most of which focus on AI application.”
Aside from China, research on robotic technology is valuable for numerous reasons. And although it has received less media hype, the use of robots won’t be restricted to a bar or restaurant. Robots can perform many jobs that are difficult and dangerous for humans, ranging from mining to disposing of bombs and landmines to cleaning windows in high-rise buildings. If humans can be spared from having to do dangerous tasks they currently do because of economic compulsions, they can divert their labor to tasks that are more fulfilling and cannot be done by machines.
To be sure, the increasing sophistication of tasks that robots are capable of performing will adversely impact employment opportunities for humans even in desirable areas. For example, if a robot bartender, costing a mere $430, can mix drinks as well as a human, businesses likely would swap bartenders with the robot. Critics will cavil that a robot bartender could never substitute for the real thing — arguing that a bartender does more than merely satisfy drink orders. Bartenders, after all, make conversation, provide a service experience and sometimes even serve as matchmakers.
These claims probably are overstated; many customers just need a drink, and younger customers are more interested in checking their mobile devices than in making conversation with humans. Some people avoid human interaction, outside their immediate social circles, moving through life with noses buried in their smartphones or tuning out their surroundings by wearing headphones.
And robots likely are here to stay. They are cheaper, more efficient, tireless and easier to manage than human workers. They will grow in sophistication, learning to communicate and make jokes. If bartenders are replaceable, so are waiters — Chili’s and Applebee’s already use devices in lieu of waiters and expect the trend to escalate. Robots likely will evolve to take higher-paying work, as AI technology advances and many routine office functions exhibiting repeat patterns may be automated out of human hands. Accountants, lawyers, bank workers, secretaries, et. al. could find their domains threatened. And as the population ages and needs more support, an array of robot maids, nurses, mobility aides, and medical assistants may be expected to meet social needs.
In addition to these benign domains, robots also will impact human conflict — robotic weapons systems and robot fighters could alter risk calculations and generate new headaches for the laws of war. For example, who bears responsibility for targets killed by autonomous weapons systems? What happens when these machines make mistakes and kill innocent children? Similarly, terrorists could turn self-driving cars and robot fighters into weapons — posing new security challenges. Finally, in a doomsday scenario, autonomous robots may demand rights, or turn against their human masters.
In this environment, we should support research that helps us to gain understanding of how robots can cooperate with each other and with human beings. Sen. Flake’s focus on beerbots misses the bigger picture. America’s leadership in technology is vital to its future and needs adequate investment to continue. So what if it also has the incidental benefit of serving up a cold one?
Sandeep Gopalan is a professor of law and pro vice chancellor for academic innovation at Deakin University in Melbourne, Australia. He previously was co-chairman or vice chairman of American Bar Association committees on aerospace/defense and international transactions, a member of the ABA’s immigration commission, and dean of three law schools in Ireland and Australia. He has taught law in four countries and served as a visiting scholar at universities in France and Germany.